Is there anyone out there who's used both Harlequin and Allegro, and
can talk about some differences, and opinions? I'm about to get a PC
running LINUX *just* so I can program in CL, but I'd like to know
why people choose one or the other.
If there's ever been a thread in this vein, then please lead me to it
in lieu of starting a new one. I do not wish to ask about religion;
just some simple stats and history (and opinions too) for a someone
who's new to CL.
dave
David Bakhash <·····@bu.edu> writes:
> Is there anyone out there who's used both Harlequin and Allegro, and
> can talk about some differences, and opinions? I'm about to get a PC
> running LINUX *just* so I can program in CL, but I'd like to know
> why people choose one or the other.
I don't think LispWorks is available for Linux.
Have you checked with Harlequin?
ACL 4.3 and ACL 5.0 beta are freely available
for non-commercial use under Linux. The
development environment for ACL under Linux
is Emacs/Xemacs. Some like that, some don't.
Rainer Joswig <······@lavielle.com> writes:
> David Bakhash <·····@bu.edu> writes:
>
> development environment for ACL under Linux
> is Emacs/Xemacs. Some like that, some don't.
Personally, not only do I like that, but I require it.
didn't know that Harlequin did it differently. Thanks for that info.
dave
David Bakhash <·····@bu.edu> writes:
> Rainer Joswig <······@lavielle.com> writes:
>
> > David Bakhash <·····@bu.edu> writes:
> >
> > development environment for ACL under Linux
> > is Emacs/Xemacs. Some like that, some don't.
>
> Personally, not only do I like that, but I require it.
>
> didn't know that Harlequin did it differently. Thanks for that info.
>
> dave
If I remember right, you can use LispWorks from Emacs, too.
But it also includes a full blown development environment
(multitude of browsers, editor, user interface designer,
documentation tool, etc.). Actually I prefer to
use the IDE. Even better when it comes as an OS
(like on the Lispm). ;-)
David Bakhash <·····@bu.edu> writes:
> Personally, not only do I like that, but I require it.
>
> didn't know that Harlequin did it differently. Thanks for that info.
but be aware that the Harlequin editor _is_ an emacs-style editor!
As far as I remember (haven't used it except for evaluation
5 years ago) it's closer to GNUEmacs than e.g. MCL's FRED
(Fred Resembles Emacs Deliberately), which is one of my favourites
(FRED has a meager built-in command set, but then it's programmable
in CLOS - great fun!).
--
regards,
Espen
Espen Vestre <··@nextel.no> writes:
> but be aware that the Harlequin editor _is_ an emacs-style editor!
> As far as I remember (haven't used it except for evaluation
> 5 years ago) it's closer to GNUEmacs than e.g. MCL's FRED
That's true. Does it support fonts, color, etc? I have to check.
> (Fred Resembles Emacs Deliberately), which is one of my favourites
> (FRED has a meager built-in command set, but then it's programmable
> in CLOS - great fun!).
Well, and in MCL each text field in windows usually is a FRED.
This is cool.
On Tue, 16 Jun 1998 16:52:36 GMT, Zeno <····@deltanet.com> wrote:
> I am going to begin programming in Lisp also, and have been looking at
> Allegro. But I am a Windows programmer for Win95 and NT. Why is it
> necessary to get a Linux box?
Allegro is available under Windows NT.
> I think there's a royalty [for Allegro] or other fee to be paid for a runtime
> version when distributing the apps.
Correct.
> Does this mean that I cannot write stand-alone applications in Lisp?
I'm not sure what you mean by this. You can certainly create a standalone
image and distribute it. You just have to pay Franz for that.
> If I write a program in C++ or VB and distribute it, the compiled code is
> machine language, and what I used to write it is unimportant to the user. Is
> Lisp different in this respect?
Lisp is a much larger language, with significant built-in libraries. Every
distribution of your application requires the Lisp run-time environment.
If you use some 3rd-party C-based libraries, such as numerical analysis or
GUI code, you'd also have to pay them every time you distributed the libraries
with your application. Lisp is pretty much the same, except that these
extra libraries are always required in every application.
-- Don
--
Don Geddis ······@tesserae.com
Tesserae Information Systems, Inc. http://tesserae.com
275 Shoreline Drive, Suite 505 (650) 508-7893
Redwood Shores, CA 94065 (650) 508-7891 [fax]
······@meta.Tesserae.COM (Don Geddis) writes:
< > If I write a program in C++ or VB and distribute it, the compiled code is
< > machine language, and what I used to write it is unimportant to the user. Is
< > Lisp different in this respect?
I don't really think so (but what do I know), if you compile the
program of course.
< Lisp is a much larger language, with significant built-in libraries. Every
< distribution of your application requires the Lisp run-time environment.
You can dramatically reduce the size of the runtime image by ripping
out the compiler, debugger, the developer environment (forget what
this includes exactly), and toplevel in the new acl though. You can
have a splash window under windows and link to the shared acl library
so you can do your own main (), as well. I think with windows you even
get an IDE with it (which is not emacs - again, maybe it's useful to
some). Someone should find and tell Martin that there's an `OLE'
interface :)
Under windows if you use `\' as a pathname separator you still have to
escape it though. I can't believe that windows uses the escape
character (at least in C++ as well as every other language I know) for
a pathname separator still. (pathname "\\\\somewhere\\in\\") Very odd.
Oh, well - enough late night lisp advocacy.
On Wed, 17 Jun 1998 13:37:08 GMT, Zeno <····@deltanet.com> wrote:
> When you say I can distribute stand-alone apps, I take this to mean that the
> runtime version consists of a library to which we link the program? I was
> beginning to get the impression that Lisp needed some sort of shell to be
> started that it could run within.
Again, I'm not sure how technically precise you mean your question to be...
However, the answer is basically yes. Allegro Common Lisp is able to dump
a single binary image that doesn't need a shell to get it started. It would
be pretty tough for an end customer to have any idea what source language you
used to create that standalone binary image.
-- Don
--
Don Geddis ······@tesserae.com
Tesserae Information Systems, Inc. http://tesserae.com
275 Shoreline Drive, Suite 505 (650) 508-7893
Redwood Shores, CA 94065 (650) 508-7891 [fax]
Chemistry is physics without thought; mathematics is physics without purpose.
Zeno <····@deltanet.com> wrote:
> I am going to begin programming in Lisp also, and have been looking at
> Allegro. But I am a Windows programmer for Win95 and NT. Why is it
> necessary to get a Linux box?
It isn't. These things are available for NT as well. In fact,
there is a free version on the franz site.
> And this stuff is expensive, isn't it? To get run time, Allegro is 6K
> for the Enterprise version, and then I think there's a royalty or
> other fee to be paid for a runtime version when distributing the apps.
> Does this mean that I cannot write stand-alone applications in Lisp?
> If I write a program in C++ or VB and distribute it, the compiled code
> is machine language, and what I used to write it is unimportant to the
> user. Is Lisp different in this respect?
No, lisp isn't different at all in this regard. C++ or VB could
charge you for distribution, they just don't. Franz has decided that this
is a good way to earn some $$ and they are doing this. People who feel it
is worth the money will pay it. Harlequin charges $600 for their product,
with no additional fees; you can distribute until you are blue in the face
and they won't charge you a dime.
Personally, i feel this is amazingly short-sighted and
counterproductive strategy on the part of franz. We sit back an decry
the success of java ( and it has been successful! ) while lisp,
obstinently superior, has languished. Do we think those free compilers,
enviornments, and tools playes a factor? Just a theory.
> Please excuse my ignorance. I have not yet read one book or written
> one line of code with Lisp.
Not at all. Happy hacking.
> On 10 Jun 1998 10:30:35 -0400, David Bakhash <·····@bu.edu> wrote:
dave
David Hanley wrote:
>
> No, lisp isn't different at all in this regard. C++ or VB could
> charge you for distribution, they just don't. Franz has decided that this
> is a good way to earn some $$ and they are doing this. People who feel it
> is worth the money will pay it. Harlequin charges $600 for their product,
> with no additional fees; you can distribute until you are blue in the face
> and they won't charge you a dime.
>
> Personally, i feel this is amazingly short-sighted and
> counterproductive strategy on the part of franz. We sit back an decry
> the success of java ( and it has been successful! ) while lisp,
> obstinently superior, has languished. Do we think those free compilers,
> enviornments, and tools playes a factor? Just a theory.
>
I am also in Zeno's position of deciding between ACL and LispWorks. In
the
case of ACL I am put off by the price, but the project I have in mind
would
require a sizable body of persistant objects, and I am attracted to the
availability of AllegroStore. I am new to OODBMS's, and have only used
MCL's WOOD, so I'm still reading up in this area. How do users of
LispWorks
typically handle their persistant data (on an NT platform?). How happy
are
ACL user's with Allegrostore?
Thanks,
Eric Scott
Cognitive Ergonomics Research Facility
San Diego State University
>
>I am also in Zeno's position of deciding between ACL and LispWorks. In
>the
>case of ACL I am put off by the price, but the project I have in mind
>would
>require a sizable body of persistant objects, and I am attracted to the
>availability of AllegroStore. I am new to OODBMS's, and have only used
>MCL's WOOD, so I'm still reading up in this area. How do users of
>LispWorks
>typically handle their persistant data (on an NT platform?). How happy
>are
>ACL user's with Allegrostore?
>
>Thanks,
>
If you want the functionality of WOOD (and much much more) in
LispWorks (4.01 on NT), you should have a look at Heiko Kirschke's
PLOB! (Persistent Lisp Objects)
<http://www.lisp.de/software/plob/Welcome.html>. This is a
full-fledged, fast client-server object oriented database (store) for
lisp objects (with b-tree indexing, transactions etc.).
Regards,
Paul
----------------------------------------------------
Paul . Meurer /at/ hit.uib.no
Humanities Information Technology Research Programme
University of Bergen
All�gaten 27
N-5007 Bergen
Norway
Eric Scott <········@schemas.sdsu.edu> writes:
> LispWorks
> typically handle their persistant data (on an NT platform?).
Since LispWorks includes an SQL/ODBC interface, you can
use databases like Oracle 8.
Greetings,
Rainer Joswig
* Zeno the Anonymous Poster
| There is a huge market of people such as myself who run 1-5 person
| programming/consulting shops and cater to businesses with 5-100
| employees. Contrary to the opinions of the larger software vendors,
| Franz must have decided that there is no market for programming languages
| there, at least not for their version of Lisp, because the pricing and
| royalties put their product out of our reach.
I'm in that market. I don't have a problem with Franz Inc's pricing.
I have several friends in similar positions as myself whose employers or
project leaders appear to have little qualms about paying for a full
license, either. I'm somewhat dismayed to hear that we don't exist and
that we constitute "no market", but I think this must go to show that
your imagination is somewhat restricted to your own immediate conditions
and that you appear to think that what you cannot imagine also cannot
exist. this is the same mental illness that afflicts Microsoft victims,
who think their sorry condition extends to the whole of the universe.
the very fortunate fact is that it doesn't.
if, on the other hand, your supposed "huge market" can present itself to
Franz Inc and be profitable for them, don't think for a minute that they
wouldn't cater to it. much to my dismay, they have already decided to
cater to Microsoft victims with a 40% discount on the Professional
Edition and a 25% discount on the Enterprise Edition, which I personally
think is a disgrace -- I don't want to have to argue against using
Microsoft's demented crudware and suffering their criminal conduct based
on the price difference of the development system, and beancounters can
be trusted to bring this issue up. an Intel box can, however, run Linux
and get away with a support license slightly more espensive than a
Professional Edition license, but it still isn't great to see that people
get rewarded by a company that should reward smart choices for making the
really stupid choice that going for Microsoft is in the long run.
my current client uses Franz Inc's ACL 5.0 for Linux offering and has
purchased a service contract, and more licenses may come as this spreads
to more systems. the service contract for Linux is a little cheaper in
the short run, but not in the longer run since it costs the same every
year instead of just a maintenance fee, so the goal is to get onto a
fully supported license once Franz Inc (hopefully) decides that Linux is
worth supporting fully. in any case, the cost of the license accounts
for less than 5% of the budgeted project costs over its (minimum) 4-year
life-time, and less than 10% of the development costs the first year.
this seems to be fairly constant in my projects.
I don't find Franz Inc's pricing to put _anything_ out of reach, neither
for me nor for my clients -- on the contrary, Franz Inc's offerings have
put some very interesting work _within_ reach for me and some fairly
complex systems within reach of relatively small budgets for my clients.
would this have happened regardless of their pricing and ability to make
money and stay healthily in business? I don't think so, and that's why
I'm worrying about their subsidizing Microsoft users, too. knowing what
tremendous costs Microsoft puts over on software developers for their
cruddy "operating systems", I have a hard time understanding the prudence
of rewarding that market with huge discounts. I'm sure those who have
yet to understand what Microsoft does to the software industry appreciate
the lower entrance costs, however.
#:Erik
--
http://www.naggum.no/spam.html is about my spam protection scheme and how
to guarantee that you reach me. in brief: if you reply to a news article
of mine, be sure to include an In-Reply-To or References header with the
message-ID of that message in it. otherwise, you need to read that page.
* Zeno wrote:
> There is a huge market of people such as myself who run 1-5 person
> programming/consulting shops and cater to businesses with 5-100
> employees. Contrary to the opinions of the larger software vendors,
> Franz must have decided that there is no market for programming
> languages there, at least not for their version of Lisp, because the
> pricing and royalties put their product out of our reach.
I think there is a market still -- there are small organisations which
are reasonably comfortable paying high prices for software, because
even those prices are actually rather small compared to salary costs,
so if the SW does something reasonably useful it will pay for itself
quite quickly.
However I do wish Franz would not have a run-time license as it's the
source of endless pain.
--tim
Hi Rainer!
Seems we agree more, than we disagree... ;)
Rainer Joswig <······@lavielle.com> writes:
> > If OTOH you plan to develop a very clever, easy to administer web
> > server (maybe based on CL-HTTP, though I wonder whether the license
> > would allow this),
>
> Why not?
I just had the impression that the CL-HTTP license was somewhat
restrictive when it comes to remarketing, but this might just've
been my imagination (haven't looked to deeply at the license for quite
some time)...
> The Mac is a poor choice for Internet services in certain areas
> due to cooperative multi-tasking, unsafe OS, lack of basic remote
> administration facilities, etc. Still I like the OS for its interface and for MCL.
> I'm about to get one of those superfast new G3 Powerbooks for
> my personal MCL hacking. :-)
Seems I should be getting a cheap PowerMac to get a chance to use
MCL <g>. I sure would like Digitool to enter the Unix/Linux arena
(though there probably would be some loss of functionality, since the
Unix arena has nothing to compete with the MacOS's interface
toolkits). Maybe Rhapsody is the solution? Who knows...
> Actually Harlequin's LispWorks for Windows has that attractive pricing and
> royalty free delivery. Their product has good potential but it also
> has some rough edges.
Yes, LWW looks interesting, but sadly there is no version for Linux
(or FreeBSD, ...) which especially my current client is starting to
use more and more...
In the same vain, ECL also looks very interesting, although it focuses
on a slightly different segment of the Lisp market than LWW does...
> MCL is cheap enough, IMHO. One guy bought an educational copy. A month later
> he sold his C++ books at a local bulletin board. ;-)
Well, for me MCL also seems cheap enough, but then again I never
bought one of those throw away IDEs ;)
> We have been using both. Clisp is very nice. Still it lacks some basic
> things (no CLIM, no CL-HTTP, no threads, debugging is not that
> comfortable, redefinable classes
> with updating objects, ...). SCSH is very useful, too. But we are
> going
Well, I wouldn't call most of these things "basic", but your point
is well taken. OTOH CLISP probably would not sport it's small
memory-footprint, if all of those features were included...
> away from Scheme if possible to concentrate on one language.
Yes, I've often come across very nice Scheme implementations (like
e.g. MzScheme/DrScheme), but since I'm more of a CL guy, I haven't
been able to use them for more than personal use...
> MCL is a Mac-only product specifically targetted at the Mac OS.
Yes, I know, I was just trying to make the point...
> No doubt about it. I just have the impression of the universality
> of the Lispm environment. I would want similar things
> on another OS (or better a modern Lispm). The Lispm
> offers me the small application footprint, TCP/IP services
> and a lot of high-level stuff written in Lisp (well, with
> a lot of source). I would like to see more infrastructure
> in Lisp with a clear software architecture and good
> reusability.
<DREAM>
Yes, this is also something I'd like to see, although I'd be weary of
another monolithic niche product like the LispMs (which weren't
"cheap" either). Also it isn't quite clear whether universality is
attainable nowadays, since the Lisp user community has diversified
further, e.g. I don't see how you'd sell the idea of another LispM to
someone who comes from the Windows world, and wants a better
replacement for VB, Delphi or VC++. Offerings like Harlequins Dylan
or LWW seem to be much more in line with their needs...
I'd rather try to get a "LispOS" environment that is not monolithic,
but permits varying depths of "Lispyness", e.g. you could host it on
top of another OS, with varying degrees of integration, and you could
go down to just above the hardware level (e.g. based on a microkernel,
for easy portability). And all of this would still enable your
applications to run unchanged...
Given this, and some really cheap high-powered hardware (like e.g. the
ARM-based NetWinder products from Corel Computing could become), you
could develop on a pure LispOS machine, and deploy on stock Intel/MS
plattforms. And once your client has seen "the light", you could
switch over to "custom" LispOS hardware, for an increase in
power/buck...
</DREAM>
Well, this would probably require more developers than there are
current Lisp users, so maybe not ...
> Btw., sounds like you have an exciting project.
<g> thanks, though I'm of course very jeallous of you, because you get
to play with LispMs, and your current project seems quite interesting,
too... ;)
Regs, Pierre.
--
Pierre Mai <····@cs.tu-berlin.de> http://home.pages.de/~trillian/
"Such is life." -- Fiona in "Four Weddings and a Funeral" (UK/1994)
From: Rainer Joswig
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <1zsipzx4.fsf@lise.lavielle.com>
Pierre Mai <····@cs.tu-berlin.de> writes:
> Hi Rainer!
>
> Seems we agree more, than we disagree... ;)
Ja. ;-)
> Seems I should be getting a cheap PowerMac to get a chance to use
> MCL <g>. I sure would like Digitool to enter the Unix/Linux arena
> (though there probably would be some loss of functionality, since the
> Unix arena has nothing to compete with the MacOS's interface
> toolkits). Maybe Rhapsody is the solution? Who knows...
Rhapsody is dead. MacOS X is the plan de jour. Sigh.
> Well, I wouldn't call most of these things "basic", but your point
> is well taken. OTOH CLISP probably would not sport it's small
> memory-footprint, if all of those features were included...
Hmm, MCL always has a very small memory footprint (not as small as CLisp,
though). A lot of people were starting programming with early MCL
versions on early Mac 68k machines with 8 MB RAM.
> further, e.g. I don't see how you'd sell the idea of another LispM to
> someone who comes from the Windows world, and wants a better
> replacement for VB, Delphi or VC++. Offerings like Harlequins Dylan
> or LWW seem to be much more in line with their needs...
True. But I think there are other markets. Btw., what will happen when
somebody (DoJ mabe) stops MS? The monopoly of MS is absolutely damaging
and I'm not going to shovel more money than necessary into
the pockets of Bill.
> I'd rather try to get a "LispOS" environment that is not monolithic,
> but permits varying depths of "Lispyness", e.g. you could host it on
> top of another OS, with varying degrees of integration, and you could
> go down to just above the hardware level (e.g. based on a microkernel,
> for easy portability). And all of this would still enable your
> applications to run unchanged...
Sounds good.
> Well, this would probably require more developers than there are
> current Lisp users, so maybe not ...
I guess one would have to reactivate a lot of knowledgeable people that
have been moving away.
Greetings,
Rainer
From: Espen Vestre
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <w6k969apl9.fsf@gromit.nextel.no>
Rainer Joswig <······@lavielle.com> writes:
> Rhapsody is dead. MacOS X is the plan de jour. Sigh.
Steve Jobs wasn't very lucky with his marketing lately.
As I read it, Mac OS X IS Rhapsody 2.0, in fact it has
all the features of Rhapsody 1.0 plus some more. The only important
change is that the now very unsure fate of Rhapsody on Intel.
> Hmm, MCL always has a very small memory footprint (not as small as CLisp,
> though). A lot of people were starting programming with early MCL
> versions on early Mac 68k machines with 8 MB RAM.
8MB RAM was a luxury, I developed experimental natural language software
on a Mac SE (8Mhz 68000) with 2.5MB RAM and (what luxury!) a 16Mhz Mac
IIX with 5MB RAM. The versions before 1.3 would even run, although
with tremendous GCing and room only for toy apps, on 1MB machines :-)
--
(espen)
From: Rainer Joswig
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <af754qhm.fsf@lise.lavielle.com>
Espen Vestre <··@nextel.no> writes:
> Rainer Joswig <······@lavielle.com> writes:
>
> > Rhapsody is dead. MacOS X is the plan de jour. Sigh.
>
> Steve Jobs wasn't very lucky with his marketing lately.
> As I read it, Mac OS X IS Rhapsody 2.0, in fact it has
> all the features of Rhapsody 1.0 plus some more. The only important
> change is that the now very unsure fate of Rhapsody on Intel.
See for example this interview with Ken Bereskin, Apple's director of
operating system strategies (an oxymoron?):
http://www.maccentral.com/news/9806/18.x_part1.shtml
My favorite question is the one about the difference between
Copland and Mac OS X.
One of the strengths of Lisp is that it doesn't have to follow
every short term trend. I guess it is very wise not
to follow Apple to early.
Btw., is there any "official" festivity/conference/workshop remembering
the 40 years of Lisp and its influence? What would be the
official date for this?
> > Hmm, MCL always has a very small memory footprint (not as small as CLisp,
> > though). A lot of people were starting programming with early MCL
> > versions on early Mac 68k machines with 8 MB RAM.
>
> 8MB RAM was a luxury, I developed experimental natural language software
> on a Mac SE (8Mhz 68000) with 2.5MB RAM and (what luxury!) a 16Mhz Mac
> IIX with 5MB RAM. The versions before 1.3 would even run, although
> with tremendous GCing and room only for toy apps, on 1MB machines :-)
I started to use Lisp on Macs OS on a Atari ST 1040 (1 MB)
with a Mac emulator. ;-) I guees, I still have Experlisp on floppy
disks.
From: Pierre Mai
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <m3vhptk261.fsf@torus.cs.tu-berlin.de>
Espen Vestre <··@nextel.no> writes:
> 8MB RAM was a luxury, I developed experimental natural language software
> on a Mac SE (8Mhz 68000) with 2.5MB RAM and (what luxury!) a 16Mhz Mac
> IIX with 5MB RAM. The versions before 1.3 would even run, although
> with tremendous GCing and room only for toy apps, on 1MB machines
> :-)
Since there are 8MB cards for the PalmPilot (16MHz 68000-based
Microcontroller "Dragonball"), this would enable Common Lisp in your
pocket, maybe called PalmLispMachine? Interesting thought... ;)
Regs, Pierre.
PS: There already is a nice Scheme Implementation for the PalmPilot,
which needs around 60KB ...
--
Pierre Mai <····@cs.tu-berlin.de> http://home.pages.de/~trillian/
"Such is life." -- Fiona in "Four Weddings and a Funeral" (UK/1994)
From: David Thornley
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <Aytk1.1432$P8.4554652@ptah.visi.com>
In article <··············@gromit.nextel.no>,
Espen Vestre <··@nextel.no> wrote:
>Rainer Joswig <······@lavielle.com> writes:
>
>> Rhapsody is dead. MacOS X is the plan de jour. Sigh.
>
>Steve Jobs wasn't very lucky with his marketing lately.
>As I read it, Mac OS X IS Rhapsody 2.0, in fact it has
>all the features of Rhapsody 1.0 plus some more. The only important
>change is that the now very unsure fate of Rhapsody on Intel.
>
If I read the stars correctly (Rigel before Betelgeuse except after
Arcturus?) there is no technical reason that Mac OS X could not be
ported to anything capable of running a Mach kernel, but Apple
is not ready to turn into a software company rather than a hardware
company. Jobs is, after all, the guy who pulled the plug on Mac
clones.
To get a little back on track, it shouldn't be all that difficult to
port current Mac apps to OS X, as a large subset of the current API
is being rewritten to sit on Mach. Assuming that MCL is as class
an act internally as externally, it should not be difficult to run it
on OS X, and then it could theoretically be ported freely.
Y'know, I was never very optimistic about Rhapsody on Intel, and I
was very nervous about Digitool porting MCL to Rhapsody. I think
the OS X thing is a very good idea.
>> Hmm, MCL always has a very small memory footprint (not as small as CLisp,
>> though). A lot of people were starting programming with early MCL
>> versions on early Mac 68k machines with 8 MB RAM.
>
>8MB RAM was a luxury, I developed experimental natural language software
>on a Mac SE (8Mhz 68000) with 2.5MB RAM and (what luxury!) a 16Mhz Mac
>IIX with 5MB RAM. The versions before 1.3 would even run, although
>with tremendous GCing and room only for toy apps, on 1MB machines :-)
>
I take it you didn't use Garnet. :-)
--
David H. Thornley | These opinions are mine. I
·····@thornley.net | do give them freely to those
http://www.thornley.net/~thornley/david/ | who run too slowly. O-
From: Rainer Joswig
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <90ml9n2r.fsf@lise.lavielle.com>
········@visi.com (David Thornley) writes:
> To get a little back on track, it shouldn't be all that difficult to
> port current Mac apps to OS X, as a large subset of the current API
> is being rewritten to sit on Mach. Assuming that MCL is as class
> an act internally as externally, it should not be difficult to run it
> on OS X, and then it could theoretically be ported freely.
This is a very interesting idea.
> Y'know, I was never very optimistic about Rhapsody on Intel, and I
> was very nervous about Digitool porting MCL to Rhapsody. I think
> the OS X thing is a very good idea.
For MCL this is a good news. The interfaces are not changing that much.
Porting seems to be possible. Remaining difficulties would be
to support integration into a foreign platform and to write
a new compiler backend. Dependency on assembler has been reduced
since the PowerPC port a lot.
For Rhapsody a Unix-based Lisp (like ACL which was already running on NeXT)
would have had advantages. (NeXToids are already complaining that Mac OS X
isn't NeXT like; different UI, no DPS, ...).
From: Pierre Mai
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <m3pvg1jz0g.fsf@torus.cs.tu-berlin.de>
Rainer Joswig <······@lavielle.com> writes:
> Rhapsody is dead. MacOS X is the plan de jour. Sigh.
If Espen read it right, it's only the Intel port which has died, which
is still quite sad, IMHO...
> True. But I think there are other markets. Btw., what will happen when
> somebody (DoJ mabe) stops MS? The monopoly of MS is absolutely damaging
> and I'm not going to shovel more money than necessary into
> the pockets of Bill.
Well, at some time in the future, MS will go the same way IBM and
most other monopolies have gone, although I don't expect the DoJ will
have any real impact on this. It is rather the emergence of some new
technology/market that will break MS' neck... Maybe to only leave us
with some new monopolist (like the change-over from IBM => MS).
But with the emergence of more open standards and open source
software, I think the market reign of monopolists might get ever less
oppressive...
> [LispOS]
I think the strategy of first developing a portable, free GUI library
(like the LispOS project does/did with CLIM/clinc) is on the right
track. Once this is in place, many interesting _development_ tools
could be implemented fairly easily and -- more important -- portably.
Another key project might be the reimplementation of the GNU (X)Emacs
kernel in Common Lisp/CLOS based on the above-mentioned GUI library,
with support for extensions in Common Lisp, and a fairly complete
compatibility package for elisp[1], allowing this implementation to
leverage most of the code available for GNU Emacs.
Having these things in place, I think a truely useful development
environment could be built, which might foster the development of
other parts of the system, like integration of a repository (maybe
based on PLOB!, with ideas drawn from ShapeTools[2]), documentation
system (based on SGML?), etc.
Another important key factor might be the ability to integrate other
languages into this development environment, allowing for development
of thin clients in other languages, like the <HYPE>Java</HYPE>.
OTOH I think we might be making[3] the same mistakes that were made
whilst developing the LispMachine environment, so it would be very
interesting to solicit input from those involved in that "project"...
So, enough dreaming already, it's time to get some "real work"(TM)
done...
Regs, Pierre.
Footnotes:
[1] There has already been a project (at MIT, IIRC), that implemented
a sufficient elisp emulation in a Scheme-based Emacs to let GNUS run
unchanged! The author estimated having implemented around 70% of
elisp to achieve this (although this was sometimes around 1994, so
Emacs and the old GNUS was somewhat smaller and simpler).
[2] ShapeTools was an interesting project on the integration of
Revision/Configuration Management and the build process, done here at
the Technical University of Berlin. Especially the Attributed
Filesystem idea would mesh quite well with an OO repository, I think.
See http://swt.cs.tu-berlin.de/~shape/index.html for more info...
[3] making is quite a strong word here, since this is currently just
a "Gedankenexperiment"...
--
Pierre Mai <····@cs.tu-berlin.de> http://home.pages.de/~trillian/
"Such is life." -- Fiona in "Four Weddings and a Funeral" (UK/1994)
In article <··············@torus.cs.tu-berlin.de>,
Pierre Mai <····@cs.tu-berlin.de> writes:
> Well, at some time in the future, MS will go the same way IBM and
> most other monopolies have gone, although I don't expect the DoJ will
> have any real impact on this. It is rather the emergence of some new
> technology/market that will break MS' neck... Maybe to only leave us
> with some new monopolist (like the change-over from IBM => MS).
How is the new technology supposed to get developed and entrenched enough
before Bill offers BIG bucks to buy it out? He'd have to be asleep at the
wheel for quite a while for that to happen. (Paying a couple of hundred
million dollars for a startup that hasn't made a buck is cheap insurance to
MS. It'd take a fanatic anti MS guy WITH the "next great thing" to upset MS in
the forseeable future.) IMNSHO, of course.
>> [LispOS]
>
> I think the strategy of first developing a portable, free GUI library
> (like the LispOS project does/did with CLIM/clinc) is on the right
> track. Once this is in place, many interesting _development_ tools
> could be implemented fairly easily and -- more important -- portably.
Everyone has there own idea of what a "portable, free GUI library" should
be. CLIM has some nice properties (presentation types!) while being BIG! I
don't know if there are any viable alternatives though.
(http://www2.cons.org:8000/clim-spec/cover.html)
> Another key project might be the reimplementation of the GNU (X)Emacs
> kernel in Common Lisp/CLOS based on the above-mentioned GUI library,
> with support for extensions in Common Lisp, and a fairly complete
> compatibility package for elisp[1], allowing this implementation to
> leverage most of the code available for GNU Emacs.
I believe this is Erik's project, when he has time. (I've lost his URL. He
posts in c.l.lisp regularly.) I'd think writing a interperter for the Elisp
byte code would be a good first step.
> OTOH I think we might be making[3] the same mistakes that were made
> whilst developing the LispMachine environment, so it would be very
> interesting to solicit input from those involved in that "project"...
As much as I'm a fan of the LispM's, I'm afraid that recreating it won't
mean much to anyone except us old fans. I think there are ideas from the
LispM's that could be recreated and extended that would be of value in today's
world. The tranparent networking comes to mind. Add in support for HTTP and
it'd still be great.
I think the basic problem with the LispM concept in today's world is that
the LispM was based on the idea of a tightly integrated environment whereas
the world has chosen to follow the path of a more loosely coupled style. How
to reconcile those those different approaches is the problem that a "modern
LispM" has to solve.
> So, enough dreaming already, it's time to get some "real work"(TM)
> done...
>
> Regs, Pierre.
Work? Dreaming is a lot more fun!
Mike McDonald
·······@mikemac.com
From: Tim Bradshaw
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <ey3emwgwhp1.fsf@todday.aiai.ed.ac.uk>
* Mike McDonald wrote:
> In article <··············@torus.cs.tu-berlin.de>,
> Pierre Mai <····@cs.tu-berlin.de> writes:
>> Well, at some time in the future, MS will go the same way IBM and
>> most other monopolies have gone, [...]
> How is the new technology supposed to get developed and entrenched
> enough before Bill offers BIG bucks to buy it out? He'd have to be
> asleep at the wheel for quite a while for that to happen. (Paying a
> couple of hundred million dollars for a startup that hasn't made a
> buck is cheap insurance to MS. It'd take a fanatic anti MS guy WITH
> the "next great thing" to upset MS in the forseeable future.)
> IMNSHO, of course.
It's very interesting to look at the IBM position in the late 70s
/ early 80s. There's a book called `Big Blue' written by an economist
who was involved in the anti-trust case against IBM, which describes
(I presume contentiously) IBM's nasty practices. It looked to him
(this is in 198[234]) as if IBM were in a completely dominant position
and would be there for many many years unless they were broken up.
His case looks pretty plausible, except for being totally wrong.
Well, you could look at MS now and see all the same monopolistic
practices, and the same feeble attempts to prevent them crapping all
over the industry, and draw the same conclusions. But they might be
equally wrong. It would be interesting to understand what happened to
IBM and see if you can find analogous mistakes that MS are making.
This is topic drift of the worst kind, I should stop now!
--tim
In article <·················@torus.cs.tu-berlin.de>, Pierre Mai
<····@cs.tu-berlin.de> wrote:
> <DREAM>
> Yes, this is also something I'd like to see, although I'd be weary of
> another monolithic niche product like the LispMs (which weren't
> "cheap" either).
Compaq will offer cheaper Alpha machines with Digital Unix (they
have bought Digital). They recently announced that they will
bring the Alpha to "mass market".
Would you think that a cheap (let's dream of <10k DM/ $6k) Compaq
machine (64 bit Alpha, say, >=500 Mhz, ...) bundled with
Symbolics Open Genera 2.0 would be attractive to Lisp developers?
???
--
http://www.lavielle.com/~joswig/
In article <·······················@194.163.195.66>,
······@lavielle.com (Rainer Joswig) writes:
> In article <·················@torus.cs.tu-berlin.de>, Pierre Mai
> <····@cs.tu-berlin.de> wrote:
>
>> <DREAM>
>> Yes, this is also something I'd like to see, although I'd be weary of
>> another monolithic niche product like the LispMs (which weren't
>> "cheap" either).
> Would you think that a cheap (let's dream of <10k DM/ $6k) Compaq
> machine (64 bit Alpha, say, >=500 Mhz, ...) bundled with
> Symbolics Open Genera 2.0 would be attractive to Lisp developers?
>
> ???
If we're going to dream, let's REALLY dream! Some of the Linux crowd has
claimed that you can get Alpha based motherboards for under $1K. You should be
able to put together a machine for say $2K +/-. Now, the really hard part is
where are you going to get Open Genera?
You still have the delivery problem. Unless you stick to plain CL, in which
case the advantages of Genera are given up. Now, if we had a working version
of Open Genera, maybe some industrious hackers would build compatibility
libraries for generic CL?
Mike McDonald
·······@mikemac.com
From: Craig Brozefsky
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <87btrkg3fc.fsf@duomo.onshore.com>
·······@mikemac.com (Mike McDonald) writes:
> able to put together a machine for say $2K +/-. Now, the really hard part is
> where are you going to get Open Genera?
Out of curiosity, why is it called "Open"?
From: Rainer Joswig
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <yauoqoxf.fsf@lise.lavielle.com>
Craig Brozefsky <·····@onshore.com> writes:
> ·······@mikemac.com (Mike McDonald) writes:
>
> > able to put together a machine for say $2K +/-. Now, the really hard part is
> > where are you going to get Open Genera?
>
> Out of curiosity, why is it called "Open"?
Why is "OpenWindows" called "Open"? No seriously, it is a product name.
Open Genera is a virtual Lisp machine, emulates the Ivory
microprocessor and enables you to run the Genera OS and its
software on a DEC Alpha running Digital Unix.
From: Rainer Joswig
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <358F5262.20A522DD@lavielle.com>
Mike McDonald wrote:
> If we're going to dream, let's REALLY dream! Some of the Linux crowd has
> claimed that you can get Alpha based motherboards for under $1K. You should be
> able to put together a machine for say $2K +/-. Now, the really hard part is
> where are you going to get Open Genera?
From Symbolics, I would think.
Then, where do I get Digital Unix? How much would it cost?
> You still have the delivery problem. Unless you stick to plain CL, in which
> case the advantages of Genera are given up. Now, if we had a working version
> of Open Genera, maybe some industrious hackers would build compatibility
> libraries for generic CL?
One company I have heard of calls their thing "Extended Common Lisp".Also, I
wonder what XPORT is: http://www.ascent.com/tools.htm .
Greetings,
Rainer Joswig
In article <·················@lavielle.com>,
Rainer Joswig <······@lavielle.com> writes:
>
>
> Mike McDonald wrote:
>
>> If we're going to dream, let's REALLY dream! Some of the Linux crowd has
>> claimed that you can get Alpha based motherboards for under $1K. You should be
>> able to put together a machine for say $2K +/-. Now, the really hard part is
>> where are you going to get Open Genera?
>
> From Symbolics, I would think.
From WHO? There is no Symbolics anymore. They got liquidated at the
beginning of the year. It's "assets" were bought by two guys who disappeared
with them.
> Then, where do I get Digital Unix? How much would it cost?
You run Linux instead. Alpha Linux supposedly runs DEC Unix apps.
> One company I have heard of calls their thing "Extended Common Lisp".Also, I
> wonder what XPORT is: http://www.ascent.com/tools.htm .
Hmm. Doesn't seem to say much. I wonder what it really is?
Mike McDonald
·······@mikemac.com
From: Rainer Joswig
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <hg1cnfk0.fsf@lise.lavielle.com>
·······@mikemac.com (Mike McDonald) writes:
> In article <·················@lavielle.com>,
> Rainer Joswig <······@lavielle.com> writes:
> >
> >
> > Mike McDonald wrote:
> >
> >> If we're going to dream, let's REALLY dream! Some of the Linux crowd has
> >> claimed that you can get Alpha based motherboards for under $1K. You should be
> >> able to put together a machine for say $2K +/-. Now, the really hard part is
> >> where are you going to get Open Genera?
> >
> > From Symbolics, I would think.
>
> From WHO? There is no Symbolics anymore. They got liquidated at the
> beginning of the year. It's "assets" were bought by two guys who disappeared
> with them.
Not really. I was able to send a bug report to someone.
> > Then, where do I get Digital Unix? How much would it cost?
>
> You run Linux instead. Alpha Linux supposedly runs DEC Unix apps.
But Open Genera? I'd be surprised.
In article <············@lise.lavielle.com>,
Rainer Joswig <······@lavielle.com> writes:
> ·······@mikemac.com (Mike McDonald) writes:
>
>> In article <·················@lavielle.com>,
>> Rainer Joswig <······@lavielle.com> writes:
>> >
>> >
>> > Mike McDonald wrote:
>> >
>> >> If we're going to dream, let's REALLY dream! Some of the Linux crowd has
>> >> claimed that you can get Alpha based motherboards for under $1K. You should be
>> >> able to put together a machine for say $2K +/-. Now, the really hard part is
>> >> where are you going to get Open Genera?
>> >
>> > From Symbolics, I would think.
>>
>> From WHO? There is no Symbolics anymore. They got liquidated at the
>> beginning of the year. It's "assets" were bought by two guys who disappeared
>> with them.
>
> Not really. I was able to send a bug report to someone.
Did you get a response? My understanding was there was one guy left to
handle repairs down in Chatsworth. Anyone know anything more?
>> > Then, where do I get Digital Unix? How much would it cost?
>>
>> You run Linux instead. Alpha Linux supposedly runs DEC Unix apps.
>
> But Open Genera? I'd be surprised.
I don't know anyone who's tried. Linux support for native apps is usually
pretty good. Besides, since you have the source to Linux, one could always fix
it if it didn't. :-)
Wasn't Symbolics charging big bucks for Open Genera anyway? Well, big bucks
as compared to our "dream" anyway.
Mike McDonald
·······@mikemac.com
From: Peter.VanEynde
Subject: Re: Lisp Implementations and Lisp in Life
Date:
Message-ID: <Ev48vt.9G4@uia.ua.ac.be>
Mike McDonald <·······@mikemac.com> wrote:
...
:>> > Then, where do I get Digital Unix? How much would it cost?
:>>
:>> You run Linux instead. Alpha Linux supposedly runs DEC Unix apps.
:>
:> But Open Genera? I'd be surprised.
: I don't know anyone who's tried. Linux support for native apps is usually
: pretty good. Besides, since you have the source to Linux, one could always fix
: it if it didn't. :-)
In the paper describing the lispm-emulator on Alpha they mention that
the emulated lispm is like a machine on the net. It all looks a lot like
using the mach-features of OSF, is this is true Linux won't be
able to run Open Genera. (The tell-tale signs of mach-calls are
negative system-call numbers it seems)
I just hope it isn't true, Open Genera on a 21264 could be fun :-).
Groetjes, Peter
--
It's logic Jim, but not as we know it. http://hipe.uia.ac.be/~s950045
Look in keyservers for PGP key.
From: Michael Harper
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <35906000.30A8628@alcoa.com>
Hi, at least the support portion of Symbolics is defnitely alive again.
My just renewed our hardware support with them nd I just exchanged a
broken Merlin I/O board this week. Haven't inquired about Genera though.
However, I do know that 8.5 had been ready for shipment earlier this
year right before they went under for a couple of months.
Mike Harper
··············@alcoa.com
Mike McDonald wrote:
>
> In article <·················@lavielle.com>,
> Rainer Joswig <······@lavielle.com> writes:
> >
> >
> > Mike McDonald wrote:
> >
> >> If we're going to dream, let's REALLY dream! Some of the Linux crowd has
> >> claimed that you can get Alpha based motherboards for under $1K. You should be
> >> able to put together a machine for say $2K +/-. Now, the really hard part is
> >> where are you going to get Open Genera?
> >
> > From Symbolics, I would think.
>
> From WHO? There is no Symbolics anymore. They got liquidated at the
> beginning of the year. It's "assets" were bought by two guys who disappeared
> with them.
>
> > Then, where do I get Digital Unix? How much would it cost?
>
> You run Linux instead. Alpha Linux supposedly runs DEC Unix apps.
>
> > One company I have heard of calls their thing "Extended Common Lisp".Also, I
> > wonder what XPORT is: http://www.ascent.com/tools.htm .
>
> Hmm. Doesn't seem to say much. I wonder what it really is?
>
> Mike McDonald
> ·······@mikemac.com
From: David Hanley
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <6moct5$m7$2@eve.enteract.com>
Mike McDonald <·······@mikemac.com> wrote:
> If we're going to dream, let's REALLY dream! Some of the Linux crowd has
> claimed that you can get Alpha based motherboards for under $1K. You should be
> able to put together a machine for say $2K +/-. Now, the really hard part is
> where are you going to get Open Genera?
The key question for me is: If I were to write cool lisp apps for
this system, who would my clients be?
I've thought of writing lisp apps for linux in cases where the
clients want a box on the network to do something specific. This is
probably a market that hasa sales in it. but I'm not sure how many.
dave
From: Harley Davis
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <6mpfas$m54$1@news1-alterdial.uu.net>
>Mike McDonald <·······@mikemac.com> wrote:
>
>> If we're going to dream, let's REALLY dream! Some of the Linux crowd
has
>> claimed that you can get Alpha based motherboards for under $1K. You
should be
>> able to put together a machine for say $2K +/-. Now, the really hard part
is
>> where are you going to get Open Genera?
>
> The key question for me is: If I were to write cool lisp apps for
>this system, who would my clients be?
>
> I've thought of writing lisp apps for linux in cases where the
>clients want a box on the network to do something specific. This is
>probably a market that hasa sales in it. but I'm not sure how many.
There seems to be a real market for "network appliances" - machines that do
noe dedicated function that you can just plug into a network and let loose,
with minimal distant browser-based configuration. You can get file servers,
HTTP servers, why not POP3/IMAP servers, etc. No reason not to write one of
these apps in Lisp running on Linux.
-- Harley
From: Espen Vestre
Subject: Re: Lisp Implementations and Lisp in Life (was: Harlequin vs. Allegro)
Date:
Message-ID: <w6btrk8vnt.fsf@gromit.nextel.no>
······@lavielle.com (Rainer Joswig) writes:
> Compaq will offer cheaper Alpha machines with Digital Unix (they
> have bought Digital). They recently announced that they will
> bring the Alpha to "mass market".
Their recent announcement were interpreted (e.g. by http://www.slashdot.org)
as the death announcemnet of Digital Unix :-(
(However, the Titanic movie special effects story has showed that
Alpha with linux is a very interesting platform!)
--
Espen Vestre
* Will Hartung wrote:
> Franz's Linux release mitigates this somewhat, but if someone wanted
> to create a small application for Windoze, and then even give it away
> to others, that would not be practical with Franz's Windows based
> product.
I don't think this is correct -- if you give away the resulting
application I think the runtime license is free. This is the case for
our Solaris license anyway, but it might be some strange academic-only
license or something.
--tim
* Zeno wrote:
> Setting distribution fees (royalties, run-time fees, etc.) aside for
> the moment, is it possible to create a small application with Lisp
> which could be distributed over the web? Let's say, you wanted to
> create a small utility which would search the files on Win95/NT
> computers and find duplicates, then provide a list to the user
> allowing them to delete duplicate files as they see fit. Can you do
> this with Lisp? Can you do this with Allegro Common Lisp? Are small
> utilities like these necessarily large programs because of the size of
> Lisp itself?
Yes, typically they are unfortunately. The real problem is that Lisps
don't typically share their runtime support with the OS & other apps
on the system, so it has to be bundled with the program making image
sizes very large. This wouldn't be true if the runtime support was
already on the machine (you could distribute just the code). That's
the case with, for instance, Java & C/C++ stuff where you have oodles
of libraries on the machine already.
--tim
Tim Bradshaw <···@aiai.ed.ac.uk> writes:
> Yes, typically they are unfortunately. The real problem is that Lisps
> don't typically share their runtime support with the OS & other apps
> on the system, so it has to be bundled with the program making image
> sizes very large. This wouldn't be true if the runtime support was
> already on the machine (you could distribute just the code). That's
> the case with, for instance, Java & C/C++ stuff where you have oodles
> of libraries on the machine already.
MCL uses shared libraries. You can install them in the extensions folder
in the system. Disk sizes: compiler=448K, library=2.9MB, kernel=168k.
The base development image starts with 1.5MB disk size.
Zeno <····@deltanet.com> wrote:
> Setting distribution fees (royalties, run-time fees, etc.) aside for
> the moment, is it possible to create a small application with Lisp
> which could be distributed over the web?
Well, the problem there is the size of the runtime system.
Lisp comes with a lit of library support, and because you client probably
doesn't already have that, they will need to get than along with your
executable.
Of course, this happens with C programs too; just that these
.dll files are often placed there by the manufacturer. I've had
to download huge VBRUN files for windoze apps though...
Let's say, you wanted to
> create a small utility which would search the files on Win95/NT
> computers and find duplicates, then provide a list to the user
> allowing them to delete duplicate files as they see fit. Can you do
> this with Lisp? Can you do this with Allegro Common Lisp?
You could write the program with either of these. Small?
I dunno. Some newer systems allow you to split off the library in a
.dll, so that after the first program, they could receive small
executables.
dave
····@deltanet.com (Zeno) writes:
> the moment, is it possible to create a small application with Lisp
> which could be distributed over the web? Let's say, you wanted to
> create a small utility which would search the files on Win95/NT
> computers and find duplicates, then provide a list to the user
> allowing them to delete duplicate files as they see fit. Can you do
> this with Lisp? Can you do this with Allegro Common Lisp? Are small
> utilities like these necessarily large programs because of the size of
> Lisp itself?
I hate repeating myself, but:
1) For very small, _self-contained_ utilities, Common Lisp may simply
not be the right language. On Unix you'd rather either use the
built-in utilites, or perl/scsh. On Windows, this could probably be
hacked-up in Delphi/VB/VC++ in a couple of minutes. Or you could
even use <HYPE>Java</HYPE> and be really portable (OTOH Java also
needs much support-code, but this sometimes happens to be already
installed on some systems).
Another posibility are several Scheme implementations, which are
able to produce small stand-alone images...
Things look very different if you want the utility to work in CL,
as a utility for developers, since then the user already has a CL
environment, and you only distribute the (compiled) files.
2) If you really wanted to distribute smallish applications, then
either ECL[1] or CLiCC[4], which compile to ANSI-C, and IIRC only
link to those functions really needed, might be better choices.
Another option might be clisp.
So this all boils down to using the right tools for the problem: Use
CL where it works best: In medium[3] to large applications that work on
complex data and/or implement complex functionality/logic. CL is
_not_ the language of choice for small, stand-alone applications,
which could even be implemented in perl or VB and still be
maintainable. You wouldn't use C++ with the STL, CORBA, OpenGL and
embedded Tcl for your utility either (other than for demoing purposes,
that is).
BTW: Your questions here are giving me the idea, that you might indeed
be better served with Harlequin's Dylan: Dylan is another member of the
Lisp family of languages, in that it shares a number of it's features,
but Dylan emphasizes more the kind of things you seem to be interested
in. Take a look at Harlequin's site, and the various pages dedicated
to Dylan on the WWW (I imagine Yahoo has a section on Dylan, which
might be a good starting point). See also the newsgroup
comp.lang.dylan, although there is little traffic at the moment.
Regs, Pierre.
Footnotes:
[1] commercial
[2] free (though discontinued, and probably Unix-only)
[3] Here I define medium to be at least around 3k lines of Lisp code,
which would roughly translate to a minimum of 6-10k lines of C++ code.
I know that this is not in line with SE classification of projects,
which label much larger projects medium...
--
Pierre Mai <····@cs.tu-berlin.de> http://home.pages.de/~trillian/
"Such is life." -- Fiona in "Four Weddings and a Funeral" (UK/1994)
* Zeno the Anonymous Poster
| the moment, is it possible to create a small application with Lisp which
| could be distributed over the web?
yes, of course. the question is: what does the recipient need to use it?
for some languages, the user needs to have installed *huge* libraries of
run-time support systems, a *massive* operating system, and have
contributed a lot to Bill Gates's now 50 billion dollar fortune. for
other languages, the user needs to have installed libraries of run-time
support systems, an operating system, and not have contributed to any
top-20 fortune list. that's the difference between VB and Common Lisp,
for instance, and all you can really argue is "but I already have all the
stuff for VB installed". the question is then why you do that instead of
the much more sensible thing, and perhaps why you should stop installing
stuff now, just _before_ you get yourself a Common Lisp environment, too.
| Let's say, you wanted to create a small utility which would search the
| files on Win95/NT computers and find duplicates, then provide a list to
| the user allowing them to delete duplicate files as they see fit. Can
| you do this with Lisp? Can you do this with Allegro Common Lisp? Are
| small utilities like these necessarily large programs because of the size
| of Lisp itself?
yes, of course you can. however, how do you run this utility? with the
VB approach, you basically run the huge run-time system from the command
line or GUI environment and feed it some input that happens to be your
"small utility". with the Common Lisp approach, you basically run the
Common Lisp environment, then invoke the function from the listener.
(sufficiently advanced Lisp listeners are indistinguishable from GUI's.)
Emacs users everywhere run some surprisingly large programs in Emacs and
don't seem to have any craving for an "executable" version of, say, Gnus.
the same applies to other Lisp software. of course, you _can_ run a
separate Emacs that you dumped with Gnus and argue against the size of
the damn thing, but why bother? running Gnus inside your already running
Emacs is _easier_. an that's the issue with Common Lisp, too: it's just
a whole lot easier to have a Common Lisp environment running and then do
all the cool stuff in it than to run whatever other environment your
computer came with and fire up huge run-time systems every time you need
the services of a small utility.
* Pierre Mai
| I hate repeating myself, but:
|
| 1) For very small, _self-contained_ utilities, Common Lisp may simply
| not be the right language.
I disagree profoundly. first, "self-contained" is no longer a valid
concept when used about general purpose computers like PC's or
workstations -- the interdepencies between individual programs and the
operating system environment has been so blurred as to make the choice
between "data for the run-time system" and "executable program" all but
meaningless. second, I run self-contained utilities in Common Lisp all
the time, if I may twist your words a little: I call them "functions".
| On Unix you'd rather either use the built-in utilites, or perl/scsh.
really? if you consider perl and scsh "utilities" to be _small_, just
because they fit in very small source files, I'm afraid that you really
don't see the whole picture. Common Lisp programs are easily shorter
than Perl and Scheme programs, so I would imagine that if you could run
Common Lisp programs with the hash-bang convention, they'd be even
smaller than "very small", since you completely ignore the *enormous*
costs of the Perl and Scheme Shell execution environment.
| Another posibility are several Scheme implementations, which are
| able to produce small stand-alone images...
let me tease you a bit: do you mean images from which you can boot your
computer, like form a floppy disk? that is the truly "stand-alone"
program. dedicated software like Internet routers running on PC's meet
this definition of "stand-alone", and they do exist -- they aren't even
very hard to build -- all it takes is linking in the basic facilities
from a library and asking the linker to write a slightly different file
format that the boot loader recognizes. matter of fact, the whole
operating system meets this definition of "stand-alone". (however, they
aren't _small_, anymore, and frequently need more than one CD-ROM. :)
incidentally, no Scheme implementations I know of can write such images.
| Things look very different if you want the utility to work in CL,
| as a utility for developers, since then the user already has a CL
| environment, and you only distribute the (compiled) files.
but the user _always_ "already has" the prerequisite environment!
this actually reminds me of a user who sued an Internet provider over
here because their marketing line for a "complete" package of software
and hardware for Internet users, "all you need to use the Internet", was
_entirely_ false -- you actually needed a whole _computer_ (and it had to
pledge allegiance to Bill Gates to boot, but that didn't seem to worry
anybody) and only _then_ did you have all you needed.
if you don't have a Common Lisp environment, the question should _not_
be: "how can we make our software available to users without computers?",
but "how do we make them understand that they need a computer _with_ a
Common Lisp environment?"
the problem, then, is only selling them the first Common Lisp application
that needs a Common Lisp environment. once he's got it, it's just
rolling them in!
| So this all boils down to using the right tools for the problem: Use CL
| where it works best: In medium[3] to large applications that work on
| complex data and/or implement complex functionality/logic. CL is _not_
| the language of choice for small, stand-alone applications, which could
| even be implemented in perl or VB and still be maintainable.
wrong! this is _so_ wrong! *cringe* (uncringe, breathe.)
nothing ever _starts_ large except in some extremely specialized areas.
all over the place, you're expected to buy the large starting point and
then make some small application on top of it, then get more funding and
boss approval and peer recognition and all that, and _then_ you let the
thousand cancers grow. Perl is *huge*, VB is even bigger. on top of
these monsters, you can write a small program that does something useful.
Common Lisp is not different in any possible regard, except it makes for
smaller, more elegant, and more playable toys. however, when Perl or VB
code gets bigger than is good for them (a screenful in my opinion, but
people seem to have very large screens these days), two things happen at
once: (1) they realize they should have used a better language, and (2)
they can't use a better language, for several reasons: (a) they don't
know the better language because they have never started using it for
small toys, (b) they don't have the better language available because
they were encouraged to use Perl and VB for "small" applications and
never got any experience with it, (c) they don't have the chutzpah to ask
their boss for the better language that they don't know anything about
because they didn't dare realize that their application would work on
complex data and/or implement complex functionality/logic until just
after they ran out of the chutzpah needed to say Perl or VB (and they)
could handle it, (d) it means scrapping working code, and finally, (e)
because the change to the better language is no longer incremental, like
all the other changes to code they have worked on or written have been.
what happens to a Common Lisp system that grows? with a marginal amount
of nurturing by people who care about elegance, it stays elegant, and it
acquires functionality the same way that the Common Lisp language did: by
careful consideration of the costs of changing one's ways as well as good
design in general functionality where observed necessary. I think a
Common Lisp programmer who is able to think in the terms of Common Lisp
the Standard (i.e, ANSI X3.226) and who is not afraid to write functions
and macros and interfaces that need specification on the same level of
precision as the standard facilities will necessarily write elegant code,
and not end up with a hodge-podge of special-cased crud that fails to
achieve abstraction by it's sheer lop-sided overweightness, which is what
happens to the cancerously growing masses of code in languages that had
all their abstraction done by the language designer and then you just get
to use whatever they left you (C, C++, Perl, VB, etc, etc).
my favorite examples of just this kind of development on top of Common
Lisp are the MOP, the Gray Stream proposal, and logical pathnames (with
which I've spent the last few days struggling...). one could view CLOS
as just such an extension to the first Common Lisp language. their
commonalities are: being well-integrated, solving very hard problems
elegantly through abstraction in the right places, and exposing no
essential differences between "application", "extension", and "language".
properties like this is why I like Common Lisp over languages that make a
tremendous effort to separate the three categories, and I want to write
my code the same way, potentially leaving something of lasting value, not
just some piece of code that "works". what makes this both possible and
impossible, however, is that adding to the Common Lisp heritage is not
for random enthusiasts in the pre-burn-out phase, but for those who are
willing to grok the language (if "grok" is still recognized as a word by
our younger audience -- Merriam-Webster's Collegiate dropped it between
the Ninth and the Tenth Edition -- boo hiss). this takes a lot of time
and effort and is a sometimes humbling experience, but there is also no
shortage of people a lot smarter and more experienced than yourself who
have gone before you. (at least in my experience.) making small steps
in this way in Common Lisp has given me a lot more pleasure than most of
the stuff I have ever been proud of in C or under Unix.
in contrast, just about any newbie can make suggestions for (real and
important) improvements to Perl or C++ and actually get them into the
language! knowing how cool it is to see "my feature" in a large system
like Emacs, it's no wonder that Perl enthuiasts feel the way they do
about their tools, but I contend that they would feel a _lot_ better if
they had not had to invent the mindless kludges that made some trivial
thing marginally easier, but could have used a well-functioning system
from the start and could write some small piece of code that did
something neatly and cleanly that would otherwise be _very_ ugly.
my suggestion is simply: start with small problems, not with large ones.
(and don't start with problems that you think are ideal for Perl or VB --
you'll find that it was the language mindset that defined "ideal", not
the problem itself or indeed _any_ part of its nature.) the goal is to
gain experience, just like you gained experience with Perl or VB: by
doing little things that looked sufficiently fun. don't think for a
minute that "what's cool in Common Lisp is different from what's cool in
Perl or VB", but ask "what's cool in _this language_?" (which is actually
what you should think for any language, or tool, or system, or whatever).
if you start off thinking "how will I solve this Perl problem in Common
Lisp?" you will only find that the Perl mindset is your limitation, and
you will probably blame Lisp, not the least because Larry Wall does.
for your entertainment, a small piece of code that almost completely
hides the annoying habit of at least one operating system to delimit
lines in text files with CRuft instead of just newline characters, and
for the moment ignoring the fact that this problem should have been
solved even more elegantly by the file transfer/sharing software...
;;; administrivia
(defpackage "STREAM-EXAMPLE"
#+allegro
(:use "COMMON-LISP"
"STREAM" ;the Gray stream proposal
"CLOS") ;the Meta-Object Protocol
#-allegro
(:use whatever is appropriate))
(in-package :stream-example)
;;; a typical simple mixin filter class and methods.
(defclass macintosh () ())
(defmethod stream-write-char ((stream macintosh) (character (eql #\newline)))
(call-next-method stream #\return))
(defmethod stream-read-char ((stream macintosh))
(let ((character (call-next-method)))
(if (char= character #\return)
#\newline
character)))
;;; interface function to dynamic stream class creation
(defun push-mixin (mixin stream)
"Push the MIXIN (named by a symbol) on STREAM (a stream object).
Actually, dynamically create, if necessary, a subclass of MIXIN and the
current class of STREAM, and change the class of STREAM to this class."
(let* ((name (concatenate 'string
(symbol-name mixin) "+"
(symbol-name (type-of stream))))
(symbol (intern name #.*package*))
(class (or (find-class symbol nil)
(ensure-class symbol
:direct-superclasses (list mixin (class-of stream))))))
(change-class stream class)))
;; typical usage
(with-open-file (stream "some-file-from-a-mac" :direction :input)
(push-mixin 'macintosh stream)
(read-line stream))
I think this is pretty cool, not the least because I can string together
all sorts of filter functions and mapping tables with mixin classes like
this, bunch up a number of thus created streams in a broadcast stream and
write to a whole bunch of files at once in places in the networked file
system I have relegated to the logical pathname stuff to select for me.
an 8,000-line piece of C code got replaced by 400 lines of Common Lisp in
this fashion. (yes, it's also faster. :)
#:Erik
PS: the other articles that need replying to will be answered shortly.
--
http://www.naggum.no/spam.html is about my spam protection scheme and how
to guarantee that you reach me. in brief: if you reply to a news article
of mine, be sure to include an In-Reply-To or References header with the
message-ID of that message in it. otherwise, you need to read that page.
Zeno> [...] I do not, however, understand the
Zeno> hatred of Microsoft/Gates which echoes through your posts
Zeno> consistently. To me, when I purchase something, it is to me
Zeno> as a vote for that product or company, much the same as my
Zeno> one vote contributes to the election of politicians and
Zeno> laws. [...]
The answer to your first sentence is contained in the second. The
general impression is that MS goes a bit beyond being a software house
and they use as much trickery as the market allows to grow. I am not
talking about arguably illegal trickery, but about gratuitous
incompatibilities, constantly changing file formats, stuff line
ms-tnef mime attachments, etc. etc. These things look stupid at first
to people who were not exposed to MS before, and then you get this
eerie feeling that these people are NOT stupid but really know what
they are doing and you don't like what they want to accomplish. So
your "vote" is important. Ordinarily I wouldn't give a damn what
other people use, but now I'm beginning to because what others are
"choosing" to use in multitudes is scaring me some.
Zeno> For instance, in the above paragraph you say that for both
Zeno> something like VB and for Lisp, one needs to have an
Zeno> operating system and libraries which are huge, and that the
Zeno> difference between VB and CL is that if you use CL, you will
Zeno> be contributing to the less successful company and not given
Zeno> money to someone who already has a lot. I assume this is
Zeno> because it is better to help the underdog. But what I am
Zeno> more concerned with is my own bottom line. Selfish?
Zeno> Perhaps, but truthful.
If you are convinced you can do what you need to do with Visual Basic
and CL wouldn't get you any efficiency gains (time wise) and you don't
have to deliver on anything but MS platforms, then why bother with CL
at all? Franz isn't there to compete with Visual Basic or Microsoft,
so I don't understand why you think you should compare them? As far
as their market go I don't think they are the underdog. BTW, if
Microsoft wanted to go into the Lisp business, they'd buy a vendor and
then try to kill the others by seeing to it that their V-Lisp is
somehow more suitable to use on Windows (including the runtime on the
Windows CD just might do that). Hasn't happened yet.
Zeno> [...] You say that *all* that can be argued is that "I already
Zeno> have all the stuff for VB installed". But this is a huge
Zeno> argument, because not only do I already have it, but every
Zeno> business that I go to has it, and they do not want to buy
Zeno> another if the one they have serves their purpose. [...]
Of course, but none of these arguments are technical arguments showing
the superiority of Visual Basic. It might well make business sense to
use it for your application. What people are reacting to is not that
it might make sense for you to use it, but the implication that Lisp
is huge and thus is somehow flawed. At least that's my take on this
thread. The only OS+GUI bundle you need to support comes with
libraries that can [only?] be used by the language the same vendor
sells. That's a good business decision by MS to set things up that
that's the case, it doesn't have anything to do with what common lisp
might be capable of.
[...]
Zeno> I have been told that CL programs can be an icon on the
Zeno> screen, and the user can just click on it to start a program
Zeno> without having to start the correct environment first. If
Zeno> this is true, then to the user, running a Lisp program is
Zeno> the same as a VB program, but not any easier.
How is it supposed to be any easier? Direct commands by the brain?
Zeno> [...] The users I deal with are
Zeno> not even used to starting the word processor to edit a
Zeno> document, they just double-click on the document from
Zeno> whatever program they happen to be in, and the word
Zeno> processor starts.
I've paid my rent and fed myself more than once by dealing with
such users. These are the same folks who call you up because
someone somewhare in their organization has installed Office-mumble
and they can no longer "open" the excel spreadsheets they receive
from that guy. This invariably is "our" problem, because MS is
way too big to call and bitch at ... the guy who's upgraded?...
well he's UPgrading -- can't argue against progress... So yes I think
I know what kind of users you are dealing with, right?
[...]
Zeno> ... They would not understand why I needed them to
Zeno> install an expensive environment on top of Windows in order
Zeno> to run my programs, while my competition does not require
Zeno> this. [...]
Clearly, if people who use Visual Basic are your competition and Franz
will not give you good pricing then they are out of the picture unless
you are convinced common lisp will help you do things that would be
hard to do with VB. My understanding though is that their runtime
royalty is a fraction of _your_ selling price. If you are delivering on
CD, runtime size won't matter, if ACL will make you more productive the
several thousand they charge initially might not matter also. One
would assume that if you are looking for alternatives maybe VB isn't
doing what you want anyway? It depends on the nature of the work and
the numbers involved. ACL is would be a bad investment for writing a
simple directory tree recurser, but maybe they want some fancy
"intranet" thing that could be built on top of cl-http? Or maybe you
heard about companies like viaweb using CL (in their case it clearly
seems to have paid off) and want to do something similar? We don't
know. Maybe you could tell us more? (Or did you already, my
apologies if that's so)
BM
* Zeno the Anonymous Poster
| ... I do not want technology to halt because file formats should remain
| consistent forever. I want innovation to continue, and the best that can
| be done is to give the new program the ability to save in the old format,
| can you see a better way?
yes, I can, and so can everybody else who has ever defined data formats
and communication protocols. again, this is not about Microsoft, but
about how you have let your entire intellectual capacity be reduced to
whatever Microsoft lets you imagine. why would technology halt because
file formats remained consistent? this is a given in _your_ world, but
it is a falsehood everywhere else. it is mind-boggling that you believe
this crap. you _must_ break loose from and look beyond the propaganda.
first, gratuitous incompatibilities with the past is _not_ evidence of
innovation, but proof of rampant stupidity and technical inferiority. if
it is perpetrated on purpose, which it is by Microsoft, it is also evil.
when the concept of innovation is expressed through incompatibilities,
which it is by Microsoft, we're talking about conceptual pollution much
more advanced than George Owell's Newspeak. when customers believe this
crap, we're talking about successful marketing and business smarts and
billions of dollars going into the wrong hands. it's all tied together
like this, and it's somewhat disconcerting that a Microsoft victim is so
dead set on defending the company and never listen. Microsoft is really
good at the propaganda game.
second, designing a forward-compatible data format or communications
protocol is trivial (except for people who think it's macho to be
complex). those who do not base their entire business on the hope and
expectations that all their users will upgrade, have no problem applying
the decade-old knowledge and experience that tells them that it is
fantastically stupid _not_ to use forward-compatible techniques when they
are so well known throughout the entire industry, _except_ for Microsoft,
and apparently their believers.
for the archetypical example of forward-compatible formats, consider the
header-value pairs used in Internet mail. the semantics of certain
headers were defined in RFC 822 and a pattern to header names was defined
never to obtain standard semantics (X-headers). that's the extent of the
core definition. this format was _explicitly_ forward-compatible, in
that new RFCs were supposed to add headers according to application
needs, as they indeed have.
if you need an elaborate example of what a data format would be like if
people were concerned about standardizing syntax and not semantics, take
a look at the Standard Generalized Markup Language, the foundation of the
hugely successful and at least somewhat forward-compatible "HTML". SGML
is basically a meta-language for other languages, but while it is clouded
in a _lot_ of problems because of its inexperienced designers at the time
it got its core definition hammered out (early 1970's), it has some
rather unique and interesting properties with respect to forward- and
backward-compatibility. it would take too long to explain them here, but
consider the HyperText Markup Language. HTML made the same mistake by
accident that Microsoft does on purpose: they defined the behavior of
future additions to be intrusive on the past. in particular, an unknown
element in HTML _always_ causes its contents to be treated as if the
element's boundary syntax ("tags" to the public) were absent. this is
_really_ stupid. the smart choice would be to have standardized on a
property, perhaps syntactic, which defined whether an element should
display or should be skipped. it doesn't take a genius to think up these
things, but it does require people who haven't been brain-washed to think
in terms that _exclude_ such ways. so, surprisingly, Microsoft's RTF
does exactly that. one is left to wonder how that can be when they argue
so strongly that data formats _must_ change for them to "innovate".
don't you think it's time to get off your Microsoft horse and stop saying
things like "the best that can be done" when you clearly lack the theory
and the practical experience to even be _able_ to say those words? not
only does it _offend_ some people to see monumentally stupid ways to do
things being paraded as "the best that can be done", it indicates that
you are not prepared to listen, you are not coming to a new world with a
desire to learn, but you are (in effect) coming to a new world to defend
your old world and your old ways. if you hear arguments intended to make
you realize that your old world is mostly bogus, it is because you make a
point out of telling people it's great when they know that it isn't.
what started as small stabs at somebody's even then apparent marriage to
Visual Basic, actually went to prove that that person would be defending
his Microsoft ways, not because they are better, but because he wouldn't
_let_ anything better than them _exist_, and would even be upset that
anybody could charge a lot more money for a kind of product that he
thought he already knew everything about and which would compete with
"the best that could be done", because surely, _nothing_ can ever exceed
Microsoft's work in technical excellence. right? (wrong!)
as long as people say Microsoft's mass-marketed crapware is technically
superior or even quality products, they _will_ be beat over the head with
huge sticks by those who know better, and they are becoming more and more
numerous because it is impossible to keep a whole nation subdued with
propaganda _all_ the time. operating systems that actually manage to
keep their data formats compatible and consistent for decades have begun
to destroy the belief systems of die-hard Microsoftians everywhere.
BTW, Lisp's syntax and data formats have remained consistent for several
decades. nobody needed to define a new and incompatible _syntax_ just
because they issued a new definition of the _language_. this is what
happens when good people think. gratuitous incompatibilities is what
happens when stupid people don't think or bad people think too much.
#:Erik
--
http://www.naggum.no/spam.html is about my spam protection scheme and how
to guarantee that you reach me. in brief: if you reply to a news article
of mine, be sure to include an In-Reply-To or References header with the
message-ID of that message in it. otherwise, you need to read that page.
[I have to edit this some sorry if I lose some relevant content]
[my response deleted]
Zeno> You snipped the end.
My apologies. My point didn't concern what I snipped IMHO (but IMHO
of course!), anyhow...
[...]
Zeno> Is
Zeno> there a reason why no one says that Lisp programs can be
Zeno> developed more efficiently than with VB, or that Lisp
Zeno> programs are more portable than C programs? [...]
Can't speak for anyone else. My only exposure to VB is through
VB for apps. that I needed to use for some instrumentation stuff
that needed to plug into Excel and some minor guidance I gave one of
our student techs. I didn't like it, but then I never liked basic
anyway. I didn't particularly care for the environmant either -- but
that's mainly cultural -- I'm used to X, emacs, etc. I start getting
red the first time windows grabs my keyboard focus... So I cannot tell
you what the contrast is with Basic. The vague _similarity_ could be
having a listener at hand while developing, and the ability to
call individual functions to test them etc. Your best bet is to get
Norvig's, and Graham's books and try to do what they are doing there
in Basic and then look at their solutions. If the things like
elegance, succint expression of ideas, etc. jump from the page,
you're all set!
(You mention prolog and AI below, that's why I suggest Norvig's book.
Paradigms of AI programming in Common Lisp is the name I believe).
[...]
Zeno> The last part of your paragraph brings up a question that I
Zeno> have been wondering about. If Microsoft had a version of
Zeno> Common Lisp, it would necessarily provide hooks into the OS,
Zeno> and make use of their proprietary ActiveX/Com technology.
[...]
Depends on how they do it. People (your buddy Erik included) tend to
get harsh and tight about non-ANSI-ness sometimes. If they build all the
stuff in and claim that they are part of Common Lisp, then people might
make noises.
[...]
Zeno> I do not understand what you mean when you say that the
Zeno> libraries provided for VB/VC++ can only be used with
Zeno> Microsoft languages. I use a lot of ActiveX and DLL
Zeno> components, and they're almost all written by firms other
Zeno> than MS. These components can be called from any language
Zeno> which supports them, and from what I understand, the newest
Zeno> version of Allegro Common Lisp can use them. [...]
Ah, might be my mistake. I assumed vbrun.dll etc were VB specific.
[on my whining about people to whom I send bills sometimes...]
Zeno> It's far worse than that at times, which I'm sure you know.
Zeno> As far as installing Office-mumble, all the "upgraded" user
Zeno> has to do is save the file in the previous version. They
Zeno> often don't want to because they may lose something in the
Zeno> translation, such as their dancing graph or font which
Zeno> flashes like a billboard.
No they don't do that because they don't want to go into the additional
truble of selecting the same thing from a menu over and over again and be
greeted with a warning message that tells them they would be missing
"advanced" features when nobody really knows what those are.
Zeno> But I do not want technology to
Zeno> halt because file formats should remain consistent forever.
Zeno> I want innovation to continue, and the best that can be done
Zeno> is to give the new program the ability to save in the old
Zeno> format, can you see a better way?
This ain't innovation, this is an under-handed way to get people
to upgrade. Fer cryin' out loud, how many of your clients _need_
anything beyond Office 4.3 or 95? What are these innovations anyway?
And why do our clients need to spend hundreds of dollars a pop to
get them? Is it the patronizing little digital paperclip?
But I digress...
Yes there can be a better way, if one's willing to look for it.
HTML changes all the time, for example, and practicaly nobody suffers
because their browser cannot understand some new tag. Anyhow,
MS can do these things because the market lets them. Erik is right,
innovation is no longer a useful word in this context.
[...]
Zeno> I am putting together a website which uses a large database.
Zeno> This site must access the data fast, display forms and
Zeno> graphical data quickly, and learn from experience.
What kind of connectivity do the clients have? How many?
Zeno> It must
Zeno> be flexible enough that as the web interface change over
Zeno> time, and as the amount of data grows and the data server is
Zeno> changed, the main-program/rules-engine should not become
Zeno> obsolete or overloaded and have to be rewritten.
I doubt either problem will happen _because of_ Lisp. It is
inevitable that things will get re-written as you learn more about the
problem you are solving. Lisp is very suitable for projects like this
because (and as Graham illustrates in his macro book) you have the
ability to write in Lisp a language that is suitable for your problem.
You then solve your problem in that language. As for efficiency when
you do know more about your program ask around for experiences. Folks
here are very helpful for stuff like that. Sometimes inner loops get
optimized in a group effort!
Zeno> I was
Zeno> looking to see if I should stick with the languages I
Zeno> currently use with some intelligence and adaptability added
Zeno> by calling Prolog, or if I should write the whole thing in
Zeno> Lisp. I am still not sure about Lisp for this since I saw
Zeno> one post which said web-objects (?) were priced at $15,000,
Zeno> but may have come down since then. I have the expense of
Zeno> research, sales, programming, hardware, and communications
Zeno> (and that's trimming it down).
Not knowing the figures, I cannot help you with the money part, but I
can tell you it doesn't take $6k to see if you can do what you want
in Common Lisp. Here's what I can see:
(1) I don't see why you're tied to Microsoft at all for this. They
hardly have the best of anything you need for this project. You'll
need a database engine of sorts, and a web server. Neither SQL server
nor IIS are that great. You can use them probably but you don't have to.
Since your UI is what the browser will display, you don't need to have
access to windows API or anything like that. All this is good, because
you have a choice and don't need to wait for things to appear on NT.
(2) I understand the AI-ish part of the project is yet to be developed.
You don't need an $6k system to start trying out ideas. Franz does have
a Free CL for Windows, which might be OK for small programs (the editor
in the IDE chokes on big files). Or you might bite the bullet and see
if you can develop under Unix. A decent headless Linux box with
very respectable power would cost you less than $1k; grab ACL for Linux and
their fi interface, run an X-server on your NT machine and you're in
business. ACL for Linux is identical (as far as the Common Lisp part goes)
to ACL for other Unices and Windows. If you don't get discouraged at
the beginning, you might find the development environment and emacs
very useful also.
The above scheme costs nothing other than time (hardware can
move very quickly when clients hear 15%-off as you may know).
[...]
Zeno> The website would also allow users to download smallish
Zeno> utilities to run on their own computers which would help
Zeno> them use the website. Since a different version of each
Zeno> program would be necessary for each type of operating
Zeno> system, I was looking to see if Lisp would provide me with
Zeno> the necessary portability, and if libraries were then
Zeno> available for each OS so that the program could run in
Zeno> different environments with (almost) the same code. The
Zeno> utilities will necessarily be sophisticated, as they would
Zeno> be very graphically oriented (just the nature of the beast).
It looks like you'll need Java or something that compiles to JVM code
for the client end.
Zeno> At this point, I have no idea if Lisp is suited to the task,
Zeno> or whether the costs of using it for this would be
Zeno> astronomical. [...]
It wouldn't be. You can explore the ways of writing the guts of this
thing using Lisp for free. The UI stuff will have to come later if I
understand you correctly. Actually, you can have the entire thing
running with free software if you use CMUCL, a free Unix and apache
(ACL for Linux is not free if you're making money off it).
If you go that route you'll also find that you get away with not
installing keyboards and mice and video cards in servers (I suppose
that is an innovation), and that you can admin this thing remotely
(another innovation) and that you don't need to reload the OS to
change certain simple things (big innovation, really revolutionary) and
that it doesn't crash or leak memory as often (oh boy, that's the next
generation). Sorry couldn't resist.
cheers,
BM
Zeno wrote:
>
> What I am trying to do is to figure out if I would get efficiency
> gains if I become proficient in Lisp. The only answers I seem to get
> are such as yours, which say if I don't think it will benefit me, then
> why bother with it? Is there a reason why no one says that Lisp
> programs can be developed more efficiently than with VB, or that Lisp
> programs are more portable than C programs?
I actually know an expert Prolog programmer who likes VB for what it is. He
likes the ease of UI development and the availability of components for that
purpose. Of course he still does all the heavy-duty algorithmic stuff in
Prolog; he wouldn't think of trying to do that in VB. I haven't used VB myself,
but I expect that a comparison between Lisp and VB would go similarly: Lisp
would be far preferable for heavy-duty computation, while VB would have
advantages over most Lisps for UI development. (There are some Lisps with very
nice UI toolkits, notably Digitool's Macintosh Common Lisp, but I don't know
what's available in this regard for the Windows environment.)
As for portability, there's good news and bad news. Common Lisp programs are
much more portable between CL implementations than C programs (provided, of
course, they don't take advantage of vendor-specific goodies like UI toolkits).
But CL implementations do not exist for as many different machines. Still, it
doesn't sound like portability is that high a priority for you. It sounds like
you could be happy doing everything on NT.
> I want innovation to continue, and the best that can be done
> is to give the new program the ability to save in the old format, can
> you see a better way?
As dismayed as I am by the manner in which Erik Naggum chooses to express
himself, I do think he has a good point about file formats designed for forward
compatibility. Microsoft could certainly put more effort into designing these,
if they cared to.
> [W]hen
> someone else posted that Lisp may not be suitable for small programs,
> another flurry started which said that Lisp was good for such programs
> because all programs become larger than originally expected
There is some truth to that. Most programs do grow.
> I am putting together a website which uses a large database. This
> site must access the data fast, display forms and graphical data
> quickly, and learn from experience. It must be flexible enough that
> as the web interface change over time, and as the amount of data grows
> and the data server is changed, the main-program/rules-engine should
> not become obsolete or overloaded and have to be rewritten. I was
> looking to see if I should stick with the languages I currently use
> with some intelligence and adaptability added by calling Prolog, or if
> I should write the whole thing in Lisp.
I think either Prolog or Lisp could serve you well. Lisp has the advantage (as
I believe someone has mentioned) that there is already an HTTP server written in
it which you could probably use (I don't know what the licensing considerations
are). Either Prolog or Lisp will take you some time to become proficient in,
however.
> The website would also allow users to download smallish utilities to
> run on their own computers which would help them use the website.
Someone else suggested Java for this purpose, and I agree -- that's exactly what
it's designed for. The caveat is that Java implementations are not mature and
robust yet. Still, given what you're trying to do, my guess is that it's worth
putting up with a bit of flakiness in order to be able to run inside people's
Web browsers (very convenient).
In any case, I don't recommend Lisp for this application.
> If there are no computers in the office when I
> get there, they are in the Windows environment when I leave. This is
> not going to change. I wondered if Lisp proficiency would help me to
> be more productive and serve them better. (The first answer to this,
> I know, will be that I can serve them better by making sure they're
> not in the Windows environment--but this is not true.)
Well, we Macintosh users would beg to differ :-) But I know, the dearth of
applications for the Mac these days is discouraging.
I would say that Lisp proficiency will help you to serve your customers better
insofar as you may be able to use it for fairly sophisticated tasks such as your
web server. As you note, most of the other things you do are probably not
appropriate applications for Lisp, under the circumstances.
-- Scott
* * * * *
To use the email address, remove all occurrences of the letter "q".
Zeno wrote:
>
> The first day I came to this site, I had just finished reading in the
> Prolog group how a small version of Prolog was quite easy to implement
> in Lisp. I thought, then, that Lisp was more powerful, but I still
> expected a full version of it to plug into VB as Prolog will. I was
> mistaken!
What Prolog are you using that plugs into VB? I'm not very familiar with the
Prolog market, though I have bought a copy of SICStus Prolog for
experimentation.
> I am learning Lisp, but the intelligent features I want add into a
> program I am currently writing for a client will be in Prolog. Much
> of their system is already written in VB, I don't really have to learn
> Prolog as much as I just need to get back in the swing of it, and it's
> features and ability to modify its own code are just a call away from
> VB.
If you already know Prolog, you certainly will have an easier time learning Lisp
than most people would have. On the other hand, if you already know Prolog,
it's a fair question exactly what is to be gained by learning Lisp. Well, the
Web server might be a good reason.
-- Scott
* * * * *
To use the email address, remove all occurrences of the letter "q".
In article <·················@zeta-sqoft.com>, "Scott L. Burson"
<·····@zeta-sqoft.com> wrote:
[snip]
>As for portability, there's good news and bad news. Common Lisp programs are
>much more portable between CL implementations than C programs (provided, of
>course, they don't take advantage of vendor-specific goodies like UI
toolkits).
Yes. My own experience supports this assertion.
>But CL implementations do not exist for as many different machines.
I'm not sure that I believe this. If you count Windows9x, WindowsNT,
Unix, Linux, MacOS and embedded systems, there is at least one good
commercial implementation of Common Lisp for each.
C and C++ compilers tend to be a bit more prolific per platform. I
believe that there are four main reasons for this:
(1) C/C++ is easier to develop and deliver,
(2) they have a larger incumbent market,
(3) they have "standards" (the quotes signal derision) which leave
a lot of room for interpretation, making it easier for a vendor
to carve out a market niche by adapting the language semantics
and runtime library to meet a particular software development
need, and
(4) they match better the APIs exposed by common operating systems
However, I think that C/C++ and Common Lisp are commerically available on
all the same platforms (at least until you start including small embedded
microprocessors.)
[snip]
--
David B. Lamkins <http://www.teleport.com/~dlamkins/>
David B. Lamkins wrote:
>
> In article <·················@zeta-sqoft.com>, "Scott L. Burson"
> <·····@zeta-sqoft.com> wrote:
> >But CL implementations do not exist for as many different machines.
>
> I'm not sure that I believe this. If you count Windows9x, WindowsNT,
> Unix, Linux, MacOS and embedded systems, there is at least one good
> commercial implementation of Common Lisp for each.
Well, but "Unix" isn't a single platform, as there are multiple versions of Unix
running on a wide variety of hardware. Even Linux, which you might think of as
a single thing, runs on various CPUs. (And then there's NT on the Alpha.) The
less popular combinations are generally not served by Common Lisp, but *all* of
them have C.
Of course, depending on what kind of application one is trying to deliver, this
may or may not be an issue. The major platforms are pretty well covered these
days, and that may be enough.
-- Scott
* * * * *
To use the email address, remove all occurrences of the letter "q".
In article <·················@zeta-sqoft.com>, "Scott L. Burson"
<·····@zeta-sqoft.com> wrote:
>David B. Lamkins wrote:
>>
>> In article <·················@zeta-sqoft.com>, "Scott L. Burson"
>> <·····@zeta-sqoft.com> wrote:
>> >But CL implementations do not exist for as many different machines.
>>
>> I'm not sure that I believe this. If you count Windows9x, WindowsNT,
>> Unix, Linux, MacOS and embedded systems, there is at least one good
>> commercial implementation of Common Lisp for each.
>
>Well, but "Unix" isn't a single platform, as there are multiple versions
of Unix
>running on a wide variety of hardware. Even Linux, which you might think of as
>a single thing, runs on various CPUs. (And then there's NT on the Alpha.) The
>less popular combinations are generally not served by Common Lisp, but *all* of
>them have C.
>
Yeah, it's hard to break the Macintosh mindset. Remember that programs
written for 68K Mac OS 0.x (i.e. 1983 vintage prerelease) can still run,
unmodified and un-recompiled, on the latest hardware and software. And
the processor today doesn't even run the same instruction set!
>Of course, depending on what kind of application one is trying to deliver, this
>may or may not be an issue. The major platforms are pretty well covered these
>days, and that may be enough.
Yes. I haven't seen the Franz price list lately, but the last time I saw
it there was a fairly long list of hardware/Unix OS combinations. I don't
know what Harlequin offers.
How far can you expect Common Lisp vendors to push? Do they really _need_
to have an environment available everywhere there's a C/C++ compiler in
order to be successful? The Common Lisp market in general is a niche,
relative to the C/C++ market (especially given the bundled and/or free
compilers available under Unix); it would probably mean death to a CL
vendor to attempt to cover all of the minor platform variations.
Hypothetically, a single customer with deep pockets could commission a
Common Lisp port to one particular fringe platform. But I'd have a hard
time imagining a business case that could justify doing that instead of
using an already-available port.
BTW, I'm not aware of all the details, but I believe that the initial MCL
PPC port was enabled by a consortium of MCL customers who were willing to
fund its development.
--
David B. Lamkins <http://www.teleport.com/~dlamkins/>
David B. Lamkins wrote:
>
> Yeah, it's hard to break the Macintosh mindset. Remember that programs
> written for 68K Mac OS 0.x (i.e. 1983 vintage prerelease) can still run,
> unmodified and un-recompiled, on the latest hardware and software. And
> the processor today doesn't even run the same instruction set!
Uh, I guess. I haven't always had luck running ancient Mac apps under System 7
(haven't tried 8 yet).
> How far can you expect Common Lisp vendors to push? Do they really _need_
> to have an environment available everywhere there's a C/C++ compiler in
> order to be successful?
No, probably not, but some potential customers may nonetheless find that an
environment they care about is not supported.
-- Scott
* * * * *
To use the email address, remove all occurrences of the letter "q".
········@teleport.com (David B. Lamkins) writes:
> Hypothetically, a single customer with deep pockets could commission a
> Common Lisp port to one particular fringe platform. But I'd have a hard
> time imagining a business case that could justify doing that instead of
> using an already-available port.
It's not too hard to port KCL/AKCL/GCL even to fairly random machines.
It's basically a large, though somewhat unusual, C program. (And the
compiler produces C.) The biggest problems are usually dynamically
loading compiled code and saving an executable image.
It should be possible to have a Common Lisp that is even easier
to port. Perhaps someone has already made one.
Another way to make things easier would be to have a Common Lisp
that worked in a different way. You would have to link compiled
files and library routines to produce an executable, just as you
do with C. (Interpreted code could still be dynamically loadable,
since that's trivial.)
This would not be everyone's ideal Lisp, but it would be very
useful nonetheless.
-- jd
In article <··············@gairsay.aiai.ed.ac.uk>, Jeff Dalton
<····@gairsay.aiai.ed.ac.uk> wrote:
>········@teleport.com (David B. Lamkins) writes:
>
>> Hypothetically, a single customer with deep pockets could commission a
>> Common Lisp port to one particular fringe platform. But I'd have a hard
>> time imagining a business case that could justify doing that instead of
>> using an already-available port.
>
>It's not too hard to port KCL/AKCL/GCL even to fairly random machines.
>It's basically a large, though somewhat unusual, C program. (And the
>compiler produces C.) The biggest problems are usually dynamically
>loading compiled code and saving an executable image.
>
I was speaking of ports of commercial CL environments, in the context of
making them available on "unsupported" Unix platforms.
GCL and its ancestors are way behind the curve w.r.t. conformance to
standards, GC technology, performance, and runtime reliability.
>It should be possible to have a Common Lisp that is even easier
>to port. Perhaps someone has already made one.
>
If this has been done, it's not in the public domain, AFAIK. If I'm
wrong, I'd love to hear about it. (Please don't tell me about CMUCL until
it breaks free of its Unix dependencies.)
There's a lot of catching up to be done for a PD Lisp system to even come
close to systems offered by Franz and Digitool.
>Another way to make things easier would be to have a Common Lisp
>that worked in a different way. You would have to link compiled
>files and library routines to produce an executable, just as you
>do with C. (Interpreted code could still be dynamically loadable,
>since that's trivial.)
>
>This would not be everyone's ideal Lisp, but it would be very
>useful nonetheless.
I think there's a PD Lisp (ECL?) that does just that. Again, not full
Common Lisp, but if you just want the Lispish syntax and semantics, it
just might be good enough. Personally, I wouldn't use something like that
unless there was a corresponding _interactive_ development environment to
go along with it. (You _don't_ want to have to run your source-level
debugger against the C code generated by a Lisp cross-compiler.)
--
David B. Lamkins <http://www.teleport.com/~dlamkins/>
David B. Lamkins wrote:
>
> In article <··············@gairsay.aiai.ed.ac.uk>, Jeff Dalton
> <····@gairsay.aiai.ed.ac.uk> wrote:
>
> >········@teleport.com (David B. Lamkins) writes:
> >
> >> Hypothetically, a single customer with deep pockets could commission a
> >> Common Lisp port to one particular fringe platform. But I'd have a hard
> >> time imagining a business case that could justify doing that instead of
> >> using an already-available port.
> >
> >It's not too hard to port KCL/AKCL/GCL even to fairly random machines.
> >It's basically a large, though somewhat unusual, C program. (And the
> >compiler produces C.) The biggest problems are usually dynamically
> >loading compiled code and saving an executable image.
> >
I am trying to port MAXIMA to run under Win95/WinNT. The only (GNU) Lisp
that seems to do it it AKCL. I am trying to port AKCL to produce an
executable with Visual C++ 5.0. Unfortunately, the (LISP C source) code
is hopelessy
entangled with out-of-date DOS-extender technology or Unix #defines that
makes me
want to puke (e.g. - you can't run it under NT, only if you boot up
under DOS :(
(Not to mention the last "port" to GO32 still uses K&R style coding.)
Well, if I ever DO get the thing ported so MAXIMA will compile and run
under NT, I will let everyone know! It just seems to me that no GNU
effort
has been made since 32-bit PCs have become available in the last 4
years.
Or maybe I am mistaken, if so please tell me where to find it!
>
> I was speaking of ports of commercial CL environments, in the context of
> making them available on "unsupported" Unix platforms.
>
> GCL and its ancestors are way behind the curve w.r.t. conformance to
> standards, GC technology, performance, and runtime reliability.
The last development was in 1994 as far as I can tell. I guess I may
be the next link in the chain...
>
> >It should be possible to have a Common Lisp that is even easier
> >to port. Perhaps someone has already made one.
> >
>
> If this has been done, it's not in the public domain, AFAIK. If I'm
> wrong, I'd love to hear about it. (Please don't tell me about CMUCL until
> it breaks free of its Unix dependencies.)
If I can find a PD Lisp that will compile MAXIMA, I will be convinced of
its glory. I haven't found anything that can do it (and Go32 DOS
extenders
don't count).
>
> There's a lot of catching up to be done for a PD Lisp system to even come
> close to systems offered by Franz and Digitool.
>
> >Another way to make things easier would be to have a Common Lisp
> >that worked in a different way. You would have to link compiled
> >files and library routines to produce an executable, just as you
> >do with C. (Interpreted code could still be dynamically loadable,
> >since that's trivial.)
> >
> >This would not be everyone's ideal Lisp, but it would be very
> >useful nonetheless.
>
> I think there's a PD Lisp (ECL?) that does just that. Again, not full
> Common Lisp, but if you just want the Lispish syntax and semantics, it
> just might be good enough. Personally, I wouldn't use something like that
> unless there was a corresponding _interactive_ development environment to
> go along with it. (You _don't_ want to have to run your source-level
> debugger against the C code generated by a Lisp cross-compiler.)
>
> --
> David B. Lamkins <http://www.teleport.com/~dlamkins/>
FF
Firefly <·······@foo.com> writes:
< I am trying to port MAXIMA to run under Win95/WinNT. The only (GNU)
< Lisp that seems to do it it AKCL. I am trying to port AKCL to
< produce an executable with Visual C++ 5.0. Unfortunately, the (LISP
< C source) code is hopelessy entangled with out-of-date DOS-extender
< technology or Unix #defines that makes me want to puke (e.g. - you
< can't run it under NT, only if you boot up under DOS :( (Not to
< mention the last "port" to GO32 still uses K&R style coding.)
Have you tried CLISP? I'm pretty sure that it's actively being
developed; this may minimize your efforts. If I remeber correctly, it
even has support for NT (or at least win32).
I've tried to compile Punimax with Win32 and, DOS version of Clisp.
So far no success.
Punimax is Maxima ported to Clisp.
--
········@one.net
Steve Gonedes wrote in message <··········@bgtnsc03.worldnet.att.net>...
>
>Firefly <·······@foo.com> writes:
>
>
>< I am trying to port MAXIMA to run under Win95/WinNT. The only (GNU)
>< Lisp that seems to do it it AKCL. I am trying to port AKCL to
>< produce an executable with Visual C++ 5.0. Unfortunately, the (LISP
>< C source) code is hopelessy entangled with out-of-date DOS-extender
>< technology or Unix #defines that makes me want to puke (e.g. - you
>< can't run it under NT, only if you boot up under DOS :( (Not to
>< mention the last "port" to GO32 still uses K&R style coding.)
>
>
>Have you tried CLISP? I'm pretty sure that it's actively being
>developed; this may minimize your efforts. If I remeber correctly, it
>even has support for NT (or at least win32).
>
>
>
>
>
Firefly <·······@foo.com> writes:
> David B. Lamkins wrote:
> >
> > In article <··············@gairsay.aiai.ed.ac.uk>, Jeff Dalton
> > <····@gairsay.aiai.ed.ac.uk> wrote:
> >
> > >········@teleport.com (David B. Lamkins) writes:
> > >
> > >> Hypothetically, a single customer with deep pockets could commission a
> > >> Common Lisp port to one particular fringe platform. But I'd have a hard
> > >> time imagining a business case that could justify doing that instead of
> > >> using an already-available port.
> > >
> > >It's not too hard to port KCL/AKCL/GCL even to fairly random machines.
> > >It's basically a large, though somewhat unusual, C program. (And the
> > >compiler produces C.) The biggest problems are usually dynamically
> > >loading compiled code and saving an executable image.
> > >
>
> I am trying to port MAXIMA to run under Win95/WinNT. The only (GNU) Lisp
> that seems to do it it AKCL. I am trying to port AKCL to produce an
> executable with Visual C++ 5.0. Unfortunately, the (LISP C source) code
> is hopelessy
> entangled with out-of-date DOS-extender technology or Unix #defines that
> makes me
> want to puke (e.g. - you can't run it under NT, only if you boot up
> under DOS :(
> (Not to mention the last "port" to GO32 still uses K&R style coding.)
[some snips]
> Well, if I ever DO get the thing ported so MAXIMA will compile and run
> under NT, I will let everyone know! It just seems to me that no GNU
> effort
> has been made since 32-bit PCs have become available in the last 4
> years.
> Or maybe I am mistaken, if so please tell me where to find it!
None for dos/windows. The last version of gcl was released last august
(2.2.2). I'm not sure that anything has been happening since then, and
I saw a post here a little while ago suggesting that the mailing list
is pretty much dead. OTOH, I found maxima+gcl to work well under
linux.
[more snipperdoodles]
> If I can find a PD Lisp that will compile MAXIMA, I will be convinced of
> its glory. I haven't found anything that can do it (and Go32 DOS
> extenders
> don't count).
There's also a port of maxima to clisp out there somewhere. I don't
know how well clisp will run under dos/window, tho.
[and, of course, still more snips]
--
Greg Harvey, http://www.biosys.net/gregh
In any case, no matter what you're doing rhythmically, it's important
to play in sync. That means in time. The exception, of course, is if
your desired result is to sound drunk (or, if you are drunk).
In article <·················@news.deltanet.com>, ····@deltanet.com (Zeno)
wrote:
>If you write a CL application which runs on Linux, will it run with
>Linux on any hardware platform, or would you need different code or a
>different compilation for each? If a CL application runs on Linux,
>will it also run on Unix?
>
No. Here's my take on the situation; perhaps a Unix expert will offer
corrections: Unix implementations are flavored by both processor
instruction set architecture (ISA) and OS version. Any pairing is, in
general, incompatible with all others, with the possible exception of OS
bugfix releases on the same hardware.
This means that any Unix program _must_ be recompiled for a particular ISA
and OS version. This is the essential technical reason that free Unix
software is distributed in source code form.
>- Zeno
>-----------
>On Sun, 28 Jun 1998 21:36:06 -0700, "Scott L. Burson"
><·····@zeta-sqoft.com> wrote:
>
>>David B. Lamkins wrote:
>>>
>>> In article <·················@zeta-sqoft.com>, "Scott L. Burson"
>>> <·····@zeta-sqoft.com> wrote:
>>> >But CL implementations do not exist for as many different machines.
>>>
>>> I'm not sure that I believe this. If you count Windows9x, WindowsNT,
>>> Unix, Linux, MacOS and embedded systems, there is at least one good
>>> commercial implementation of Common Lisp for each.
>>
>>Well, but "Unix" isn't a single platform, as there are multiple versions
of Unix
>>running on a wide variety of hardware. Even Linux, which you might
think of as
>>a single thing, runs on various CPUs. (And then there's NT on the
Alpha.) The
>>less popular combinations are generally not served by Common Lisp, but
*all* of
>>them have C.
>>
>>Of course, depending on what kind of application one is trying to
deliver, this
>>may or may not be an issue. The major platforms are pretty well covered these
>>days, and that may be enough.
>>
>>-- Scott
>>
>> * * * * *
>>
>>To use the email address, remove all occurrences of the letter "q".
--
David B. Lamkins <http://www.teleport.com/~dlamkins/>
David B. Lamkins wrote:
>
> In article <·················@news.deltanet.com>, ····@deltanet.com (Zeno)
> wrote:
>
> >If you write a CL application which runs on Linux, will it run with
> >Linux on any hardware platform, or would you need different code or a
> >different compilation for each? If a CL application runs on Linux,
> >will it also run on Unix?
> >
>
> No. Here's my take on the situation; perhaps a Unix expert will offer
> corrections: Unix implementations are flavored by both processor
> instruction set architecture (ISA) and OS version. Any pairing is, in
> general, incompatible with all others, with the possible exception of OS
> bugfix releases on the same hardware.
>
> This means that any Unix program _must_ be recompiled for a particular ISA
> and OS version. This is the essential technical reason that free Unix
> software is distributed in source code form.
This is basically correct, although I gather there have been attempts at greater
binary compatibility among the x86 Unices -- since I don't have an x86, I
haven't paid close attention to this -- and some Unices have specific
compatibility support for running most if not quite all executables from another
version (for instance, Sun Solaris on the SPARC will run most SunOS binaries).
So back to Zeno's question: to run on a specific hardware/OS combination, you
need a Common Lisp that runs on that combination. Exactly how hard it is to
move to a different platform depends on the situation. If, for instance, you
develop for ACL (Allegro Common Lisp) on x86 under Linux, and you want to move
to SPARC under Solaris, you can just buy the ACL for Solaris from Franz and
recompile; you probably won't have to change your code at all. On the other
hand, if you wanted to port to a platform that required changing Lisp vendors,
you might have to modify your code somewhat, depending on exactly what it is
doing. Code that is primarily computational, and uses only features defined in
the Common Lisp specification, almost always ports with little or no
modification. If you're doing OS-related stuff (like sockets or something) that
relies on features outside the standard, you're likely to have a bit more work
to do. And of course if you're using a vendor-specific GUI or other such
package, you'll have quite a bit of work to do.
However I must warn you about something. Franz in particular has had a couple
of Lisp implementations from other sources that they have at various times
purchased the rights to (along with at least one they developed in-house). They
have had what I consider to be the somewhat misleading practice of calling them
all Allegro, even though they were separate implementations that didn't share
source code. Coral Common Lisp, for instance, was developed by Coral, bought by
Apple, then sold to Franz; Franz subsequently sold it to Digitool, but during
the time they carried it, they called it "Macintosh Allegro Common Lisp". And
it is my understanding that their Windows products, also called "Allegro", were,
at least originally, of yet another lineage; I'm not sure about the situation
today.
In contrast, to my knowledge, all of their Lisps for any version of Unix share
the vast bulk of their source code, excepting only the compiler backend and
various OS-interface routines that (unfortunately) are not 100% portable between
different Unices.
While it is true that code that sticks to the Common Lisp specification is
highly portable, it is also the case that code that uses a lot of
implementation-specific extensions is not so portable. Secondly, there are
differences in garbage collectors, compilers, and other environmental aspects,
that sometimes become relevant to very complex applications. And thirdly, the
overall quality of different implementations can certainly vary. I think that
Franz has called all of their Lisps "Allegro" quite intentionally, to obscure
their distinct lineages and to suggest that one can move an application between
any of them by just recompiling, period. This certainly was not the case a few
years ago, when some colleagues of mine were looking into porting a Unix ACL
application to Windows. It may be true, or more nearly true, now, but I don't
trust Franz in this area, and if I wanted to develop a Windows application I
would use their (or someone's) Windows product and not try to cross-develop from
Unix. It should be clear that I don't think highly of this practice on their
part.
-- Scott
* * * * *
To use the email address, remove all occurrences of the letter "q".
"Scott L. Burson" <·····@zeta-sqoft.com> writes:
> However I must warn you about something. Franz in particular has had a couple
> of Lisp implementations from other sources that they have at various times
> purchased the rights to (along with at least one they developed in-house). They
> have had what I consider to be the somewhat misleading practice of calling them
> all Allegro, even though they were separate implementations that didn't share
> source code. Coral Common Lisp, for instance, was developed by Coral, bought by
> Apple, then sold to Franz; Franz subsequently sold it to Digitool, but during
> the time they carried it, they called it "Macintosh Allegro Common Lisp". And
> it is my understanding that their Windows products, also called "Allegro", were,
> at least originally, of yet another lineage; I'm not sure about the situation
> today.
I don't think this is correct. Apple sold MCL to Digitool. Apple owned
MCL and after the introduction of the PowerPC it was obvious that
(again) Apple was not able to support developers appropriately -
by porting MCL to the PowerPC Macintosh. Apple wanted to discontinue
MCL. Some brave Apple customers and the guys from Digitool
finally took over and did the port (I wish something like this would have happened
for the Newton technology, too.).
I think neither MCL
nor one of its earlier versions were owned by Franz. But I can't remember
for sure. True, it was once called Macintosh Allegro Common Lisp. Does
anybody remember why it was called that way?
Rainer Joswig wrote:
>
> "Scott L. Burson" <·····@zeta-sqoft.com> writes:
>
> > Coral Common Lisp, for instance, was developed by Coral, bought by
> > Apple, then sold to Franz; Franz subsequently sold it to Digitool, but during
> > the time they carried it, they called it "Macintosh Allegro Common Lisp".
>
> I don't think this is correct. Apple sold MCL to Digitool.
>
> I think neither MCL
> nor one of its earlier versions were owned by Franz. But I can't remember
> for sure. True, it was once called Macintosh Allegro Common Lisp. Does
> anybody remember why it was called that way?
There may have been some twists in the story of which I am unaware, but I am
fairly certain that when it was called "Macintosh Allegro Common Lisp", it was
in Franz's possession. (And what other reason would there be for calling it
that?) Perhaps it reverted to Apple for some reason before being sold to
Digitool. Anyone know for sure?
-- Scott
* * * * *
To use the email address, remove all occurrences of the letter "q".
"Scott L. Burson" <·····@zeta-sqoft.com> writes:
> There may have been some twists in the story of which I am unaware, but I am
> fairly certain that when it was called "Macintosh Allegro Common Lisp", it was
> in Franz's possession. (And what other reason would there be for calling it
> that?)
For the same reason people are calling their software "Power"-,
"Open"- or "Visual"-something.
> Perhaps it reverted to Apple for some reason before being sold to
> Digitool. Anyone know for sure?
If Franz ever had influence over MCL (I can't remember that),
then it was minor. Apple had much more influence, since
they were a long time owner of MCL.
From Apple's Dylan manual (1992), Foreword by Larry Tesler:
"...
We felt we could not design a good language without concurrently
implementing it. But we did not have the skills in house to implement the
kind of language we envisioned. Around the time we were approached by
Massachusetts-based Coral Software, Inc., which was seeking to be aquired.
Coral had created a popular Common Lisp implementation distinguished by
its small memory footprint, very usable speed, interesting object system,
and thorough integration into the Macintosh. A few months later, we aquired
the assets of the company, hired most of the staff, and created "ATG East"
in Cambridge.
We asked the new group to accept two challenges: (1) continue to develop
their Common Lisp implementation, adding CLOS, ephemeral garbage
collection, and other features; and (2) [...]
The first led to the very popular product know as Macintosh Common Lisp (MCL) 2.0.
..."
O.k. I looked, I have the original Macintosh Allegro Common Lisp 1.3 from 1989.
It is an Apple product.
Scott L. Burson wrote:
>
> There may have been some twists in the story of which I am unaware, but I am
> fairly certain that when it was called "Macintosh Allegro Common Lisp", it was
> in Franz's possession. (And what other reason would there be for calling it
> that?) Perhaps it reverted to Apple for some reason before being sold to
> Digitool. Anyone know for sure?
Duane Rettig of Franz has set me straight. Reposted with permission, and with
the disclaimer that this is not an official statement of Franz Inc. but just one
developer's recollection:
> From: ·····@franz.com (Duane Rettig)
> Date: Mon, 29 Jun 98 12:07:17 -0700
> No, Franz never owned Coral Common Lisp. We had had some business dealings
> with them, in which they agreed to call their product "Allegro", with the
> intention of eventually cross-pollinating the implementations. This was long
> before the CL standard, and both were fairly close CLtL1 implementations at
> the time. But the deal fell through with no technical relationship, and Apple
> subsequently bought Coral and eventually dropped the Allegro name.
-- Scott
* * * * *
To use the email address, remove all occurrences of the letter "q".
In article <············@lise.lavielle.com>, Rainer Joswig
<······@lavielle.com> wrote:
>I think neither MCL
>nor one of its earlier versions were owned by Franz. But I can't remember
>for sure. True, it was once called Macintosh Allegro Common Lisp. Does
>anybody remember why it was called that way?
Yes. It was a co-marketing arrangement between Coral Software and Franz.
Paul Snively
Former Unofficial MCL Support Tech @ Apple Computer
* Scott L. Burson
| And it is my understanding that their Windows products, also called
| "Allegro", were, at least originally, of yet another lineage; I'm not
| sure about the situation today.
the latest Windows version, ACL 3.0.2, was a completely different product
from the Unix version, ACL 4.3.1, much to my dismay. ACL 5.0 is the same
product on both, with Windows-specific stuff coming from the ACL 3.0.2
heritage, like the IDE. the very real and noticeable shortcomings of the
Windows offering that I ran forcefully into when one of my first clients
insisted on NT (which I didn't know how crappy was at the time) cost me a
tremendous amount of work. I have recommended strongly against ACL for
Windows since then, and I have always been precise that I'm talking about
the Unix version when praising Allegro Common Lisp, but the consolidated
version should be as good as the Unix version is. I'm very happy that
this particular confusion is now history.
| It may be true, or more nearly true, now, but I don't trust Franz in this
| area, and if I wanted to develop a Windows application I would use their
| (or someone's) Windows product and not try to cross-develop from Unix.
yup, been there, tried that. cross-development wasn't even close to
doable between ACL 4.3 (an ANSI CL) and ACL 3.0.2 (barely CLtL1 with some
CLOS), so I have also yet to fully regain the necessary confidence, but
watching the bug fixes and other stuff about ACL 5.0 strongly reinforces
their claims to common sources. maybe we can finally get to the point
where I wanted to be almost two years ago when this client wanted NT:
that one can run Allegro CL and Emacs under Neanderthal Technology and
not have to deal with that ill-designed "operating system" at all.
to be honest, part of the reason I like and use Common Lisp is that I
don't have deal with all the demented Unix heritage, either -- it's a lot
of work, but possible, to write clean and neat abstractions on top of the
many Unices and not have to deal with their low-level crud and mistakes.
speaking of mistakes, I _want_ file versioning. :if-exists :clobber is
_not_ my favorite option, and I want (close ... :abort t) to leave the
old file intact. that Unix does not offer a way to do this so it's quite
hard to get it right has been a pet peeve of mine for years, long before
I shifted focus to Common Lisp. safe file I/O generally appears to be up
to the application. e.g., I don't want to make a file available to the
world until it has been fully written and closed, but that means I have
to use temporary filenames and do a rename on close manually under Unix.
this is annoying. what's more annoying is that I haven't found a way to
hook into the operation of OPEN and CLOSE to automate this behavior.
#:Erik
--
http://www.naggum.no/spam.html is about my spam protection scheme and how
to guarantee that you reach me. in brief: if you reply to a news article
of mine, be sure to include an In-Reply-To or References header with the
message-ID of that message in it. otherwise, you need to read that page.
In article <·················@zeta-sqoft.com>, "Scott L. Burson"
<·····@zeta-sqoft.com> wrote:
[snip]
>Coral Common Lisp, for instance, was developed by Coral, bought by
>Apple, then sold to Franz; Franz subsequently sold it to Digitool, but during
>the time they carried it, they called it "Macintosh Allegro Common Lisp".
> And it is my understanding that their Windows products, also called
>"Allegro", were, at least originally, of yet another lineage; I'm not
>sure about the situation today.
See <http://www.teleport.com/~dlamkins/mcl-review.html#companies> and
<http://www.teleport.com/~dlamkins/mcl-aclnt-comparison.html#heritage>.
[snip]
--
David B. Lamkins <http://www.teleport.com/~dlamkins/>
* Zeno wrote:
> If you write a CL application which runs on Linux, will it run with
> Linux on any hardware platform, or would you need different code or a
> different compilation for each? If a CL application runs on Linux,
> will it also run on Unix?
No. It will need to be recompiled for each platform it needs to run
on, the same way a C/C++ program would be. You rely therefore on a
compiler for each platform. Assuming the compiler (`lisp system') is
compatible across the different platforms, recompilation should be
trivial.
I expect that something like Eclipse provides such a compatible
compiler, as it compiles into C and all these platforms are likely to
have reasonably compatible versions of gcc. (Gcl may also be an
option, or a commercial Lisp if they support all the platforms you
need).
--tim
Tim Bradshaw <···@aiai.ed.ac.uk> writes:
> * Zeno wrote:
> > If you write a CL application which runs on Linux, will it run with
> > Linux on any hardware platform, or would you need different code or a
> > different compilation for each? If a CL application runs on Linux,
> > will it also run on Unix?
>
> No. It will need to be recompiled for each platform it needs to run
> on, the same way a C/C++ program would be.
An interesting note: last I checked, binaries compiled under CMUCL for
Linux will run under CMUCL on FreeBSD. I wouldn't be surprised if it
worked the other way also, though I don't remember trying it.
--
Fred Gilham gilham @ csl . sri . com
King Christ, this world is all aleak, / And life preservers there are none,
And waves that only He may walk / Who dared to call Himself a man.
-- e. e. cummings, from Jehovah Buried, Satan Dead
Fred Gilham <······@snapdragon.csl.sri.com> writes:
> An interesting note: last I checked, binaries compiled under CMUCL for
> Linux will run under CMUCL on FreeBSD. I wouldn't be surprised if it
> worked the other way also, though I don't remember trying it.
>
FreeBSD has an emulation mode for Linux binaries. So this is not CMUCL
specific. Linux doesn't provide the same for FreeBSD binaries however.
--
Lieven Marchand <···@bewoner.dma.be>
------------------------------------------------------------------------------
Few people have a talent for constructive laziness. -- Lazarus Long
Lieven Marchand <···@bewoner.dma.be> writes:
> Fred Gilham <······@snapdragon.csl.sri.com> writes:
>
> > An interesting note: last I checked, binaries compiled under CMUCL for
> > Linux will run under CMUCL on FreeBSD. I wouldn't be surprised if it
> > worked the other way also, though I don't remember trying it.
> >
>
> FreeBSD has an emulation mode for Linux binaries. So this is not CMUCL
> specific. Linux doesn't provide the same for FreeBSD binaries however.
I don't think that's the reason because CMUCL binaries have nothing to
do with the usual executables. CMUCL has it's own binary format. I
think this is what Fred Gilham is talking about.
And yes Linux CMUCL will load and run binaries created on FreeBSD.
That is how I was able to get a long-float version for Linux: someone
compiled up the code on FreeBSD and I grabbed it and loaded it up and
ta da! a Linux version with long-float support.
Ray
In article <·················@news.deltanet.com>,
Zeno <····@deltanet.com> wrote:
>If you write a CL application which runs on Linux, will it run with
>Linux on any hardware platform, or would you need different code or a
>different compilation for each? If a CL application runs on Linux,
>will it also run on Unix?
>
No. As a matter of fact, this is why I'm not getting Linux yet.
I just got a nice new Mac, and can spare a 1.3Gb drive for MkLinux,
which is the Linux that runs on Power Macintoshes. There's stuff
I want to do.
Unfortunately, I know of no good lisp system for MkLinux. I don't
want to have to use GCL or CLisp. Linux on Intel has CMU-CL and
ACL available. If I can't get a good Lisp for a system, why
install it?
(Yeah, I know that this question would lead to endless amusement
in most of the comp* groups.)
Anyway, if somebody does know of a good Lisp for MkLinux, please
let me know.
--
David H. Thornley | These opinions are mine. I
·····@thornley.net | do give them freely to those
http://www.thornley.net/~thornley/david/ | who run too slowly. O-
David Thornley wrote:
>
> Anyway, if somebody does know of a good Lisp for MkLinux, please
> let me know.
Yeah, that would be nice. A port of CMUCL would probably not be too hard, as
such things go. Code generators already exist for SPARC and Alpha, so PPC
should be fairly straightforward to add.
-- Scott
* * * * *
To use the email address, remove all occurrences of the letter "q".
Zeno wrote:
>
> If you write a CL application which runs on Linux, will it run with
> Linux on any hardware platform, or would you need different code or a
depends in which mode you run it: interpreted: yes; bytecode compiled:
very likely, native mode compiled: obviously not
> different compilation for each? If a CL application runs on Linux,
> will it also run on Unix?
Most likely: the core of Unix systems (including Linux) is very
compatible. You'll run into troubles when you use proprietary
extensions. You will have to recompile, though.
--
Hartmann Schaffer
Guelph, Ontario, Canada
········@netcom.ca (hs)
> You snipped the end. The point of the paragraph was that I am much
> more likely to listen to the benefits of choosing one over the other,
> as opposed to listening to what the other one does wrong. I wondered
> why, whenever I asked a question about Lisp, the result was always a
> bashing of Microsoft. The practices of big business and the wealth of
> CEOs is not my concern while trying to figure out if Lisp is "right
> for me", and I do not wish to take a stand against any corporate
> giants. I just want to write quick, intelligent, robust programs.
Amen, brother. I use free software/Linux etc. because it happens to work
better for the tasks I want to perform, personal politics not
withstanding.
> What I am trying to do is to figure out if I would get efficiency
> gains if I become proficient in Lisp. The only answers I seem to get
> are such as yours, which say if I don't think it will benefit me, then
> why bother with it? Is there a reason why no one says that Lisp
> programs can be developed more efficiently than with VB, or that Lisp
> programs are more portable than C programs? I do not know, because
> almost all the answers I receive are negative and only infer
> superiority like the first sentence of your paragraph above--and
> always included are references to the amount of money the president of
> Microsoft earns, or that MS business practices are suspect.
>
> The last part of your paragraph brings up a question that I have been
> wondering about. If Microsoft had a version of Common Lisp, it would
> necessarily provide hooks into the OS, and make use of their
> proprietary ActiveX/Com technology. Would this make everyone cut
> loose with a barrage of MS hatred whenever someone posted a question
> about V-Lisp in this group? Does this happen when someone posts a
> question about using ODBC or Com technology with Franz Lisp? Or does
> everyone but me know better than to ask such questions here?
I have seen quite reasoned discussion upon making use of COM/ActiveX
with LISP
in this newsgroup. The majority of people here seem to be pragmatic
rather than
reactionary.
I originally began using Common LISP because it was substantially
different from
the imperative languages like Pascal and C I had been using before. I
was a student
at the time (still am, part-time) and thought learning LISP and Prolog
would help me
think about programming differently. And they have done.
I continued to use Common LISP because the John Koza wrote his genetic
programming
framework in it, and I was interested in this form of AI. While I got
more familiar with
Common LISP, I found several things:
1) Common LISP is a big language/environment that does a lot more than
many languages. It
is also very extensible. This can be a double-edged sword.
2) The ability to either interpret or compile code, and mix the two, can
provide you with new
and fast ways of rapidly developing and prototyping code. If a system is
not working quite right,
you can interrupt it at run-time, redefine the function, and continue,
with very little down-time.
None of this waitiing for your entire project to recompile ala C++. This
is just one example.
This is not isolated to LISP. For example, I am told some Java
environments can allow you to modify
the code at run-time.
3) Good Common LISP programming is hard, but not in the way C++ is hard.
In C++, it's hard to write something that runs at all, but if it does,
it will tend to be very efficient (provided good algorithms and
appropriate data structures are used. Unfortunately, the C/C++ community
seems to have a large faction of people who see the array as the one
true data structure). In Common LISP, you can more easily write
something which works, but to get it running efficiently requires
understanding how your given compiler optimises, declaring types, which
functions should be inlined, writing tail-recursive code and so forth.
4) LISP is based on S-expressions. S-expressions all have the same basic
form. Programs and data
in LISP have the same form (S-expressions). This, combined with the
ability of LISP to evaluate itself, allows you to write programs which
manipulate programs, or programs which produce programs.
The gap between programs and data in LISP is somewhat blurry sometimes,
which can be a very good thing.
This method of programming (programs which produce other programs by
returning closures etc.) was totally unknown to me before I started
getting into LISP. Also, LISP is able to have a safe macro
system, largely because of the constant form of the language. C++
#defines are notoriously unsafe
(see Effective C++).
> To choose the most expensive
> programming environment does not seem beneficial to me, but then
> again, I am still less than halfway through "Lisp", by Winston/Horn,
> and am still a total Lisp neophyte, with no understanding of what it
> can do for me.
This is a brilliant text which, should you become more conversant with
LISP,
you will probably find yourself re-reading. Some would argue that the
added productivity
of a LISP environment can justify the high cost. I personally don't find
this argument sufficiently
compelling.
> Damn, I write long messages.
Yes, you do. Perhaps if you want a small system you could use one of the
many good Scheme implementations that are popping up lately. Because
there is less in Scheme than in Common LISP, the runtime image size
might be more to your liking. But it depends on the application.
Mark
····@deltanet.com (Zeno) writes:
> Your post was extremely informative. I was completely surprised when
> you compared C++ and Lisp in your point #3, about getting programs to
> run and optimizing them. I originally left C++ because I wanted to be
> able to get things going as quickly as possible, and then optimize the
> code later. Now, I can get things running quickly, but have lost the
> ability to optimize the code in many ways. Being insulated from the
> low-level functions means far fewer glitches in initial development,
> but later "being insulated from" can turn into "am denied access to."
>
Actually there's just a different way of getting access to them.
You can bit-fiddle with the best of them in Lisp, but you aren't
bit-fiddling by default.
A good example of this is destructive list operations. These
operations allow you to change lists without copying or creating new
lists. They let you do grody things to lists that can cause
unexpected results. But their use is not encouraged and their
existence is not trumpeted like their non-destructive counterparts.
The declaration system also takes some getting used to. Proper use of
declarations requires you to get into the lower-level aspects of the
system and understand the way lisp objects are represented. Doing so
may result in large speedups because you can get the system to
actually switch from less-efficient to more-efficient representations.
C and C++ not only make dealing with low-level concerns easy; they
encourage and almost require it. Lisp makes it possible but makes it
largely unnecessary for simply getting working programs. You don't
have to know about rplaca or (declare (fixnum foo)) to get your lisp
program to work. You may need to know about them to get your program
to run quickly. But the point is that lisp makes clear distinctions
between high- and low-level concerns, and this lets you see the forest
for the trees. When you need to look at the trees you can; you just
need to take out a different instrument. The admittedly difficult
thing in lisp is to find out what those instruments are and how they
work.
--
Fred Gilham gilham @ csl . sri . com
King Christ, this world is all aleak, / And life preservers there are none,
And waves that only He may walk / Who dared to call Himself a man.
-- e. e. cummings, from Jehovah Buried, Satan Dead
Mark Greenaway <········@scholar.nepean.uws.edu.au> writes:
> 3) Good Common LISP programming is hard, but not in the way C++ is hard.
> [...] In Common LISP, you can more easily write
> something which works, but to get it running efficiently requires
> understanding how your given compiler optimises, declaring types, which
> functions should be inlined, writing tail-recursive code and so forth.
I don't think it's difficult to get reasonably good performance from
(most) Common Lisp(s) in a largely implementation-independent way.
The main problem is that it's not "obvious" _a priori_ (we might
say) what things tend to be efficient. C (and even C++) is more
transparent in this respect, though this is more a matter of
implementation tradition than of the language itself. C (and
even C++) also, to a large extent, simply leave out things that
implementations might not make efficient in a fairly straightforward
way. (For example, there are no nested procedure definitions
and no closures.)
-- jd
* Zeno wrote:
> For instance, in the above paragraph you say that for both something
> like VB and for Lisp, one needs to have an operating system and
> libraries which are huge, and that the difference between VB and CL is
> that if you use CL, you will be contributing to the less successful
> company and not given money to someone who already has a lot. I
> assume this is because it is better to help the underdog. But what I
> am more concerned with is my own bottom line. Selfish? Perhaps, but
> truthful.
No, it's because monopolies are *bad* for the free market. I have my
problems with MS SW *as* *software* -- it's generally unreliable &
hard to manage compared to other OS-type products (I'm a system
manager), but I have much worse problems with seeing a monopolist
stifle competition, and MS hold a monopoly on the PC OS market (they
would deny this of course).
Damn, *more* topic drift. I should post articles about Lisp...
--tim
* Zeno the Anonymous Poster
| Agreed. Monopolies are bad. But I have no interest in breaking up any
| monopolies, and I wondered why answers to questions about Lisp in this
| group have more space dedicated to the evil empire of Microsoft than
| about the benefits of using Lisp.
in brief, because you have let Microsoft define the value to you of any
development system that you could ever want to purchase (and thus the
price you would be willing to pay for "a development system", regardless
of what value it would or could offer you), and because you parade this
microsoftization of your value system before others all the time by
incessantly comparing with Visual Basic.
#:Erik
--
http://www.naggum.no/spam.html is about my spam protection scheme and how
to guarantee that you reach me. in brief: if you reply to a news article
of mine, be sure to include an In-Reply-To or References header with the
message-ID of that message in it. otherwise, you need to read that page.
In article <················@naggum.no>, ······@naggum.no says...
> * Zeno the Anonymous Poster
> | Agreed. Monopolies are bad. But I have no interest in breaking up any
> | monopolies, and I wondered why answers to questions about Lisp in this
> | group have more space dedicated to the evil empire of Microsoft than
> | about the benefits of using Lisp.
>
> in brief, because you have let Microsoft define the value to you of any
> development system that you could ever want to purchase (and thus the
> price you would be willing to pay for "a development system", regardless
> of what value it would or could offer you), and because you parade this
> microsoftization of your value system before others all the time by
> incessantly comparing with Visual Basic.
I'm working of ridding the world of Microsoft. I'll start by
assassinating Bill Gates, and then try to obtain a tactical nuke so I can
take out Seatle.
That just leaves all the people who helped make MS what they are today,
i.e. the people who bought their software. My personal believe - please
correct me if I'm wrong - is that the demands of the free market have
made MS into the competitive monster that it is. Just like every other
corporation. So, more tactical nukes will be needed.
My efforts to provoke conflict between Indian and Pakistan have failed,
thanks to another corporate outfit called the United Nations. So, first
Seatle, then New York...
The risk is that the market will attempt to put replace MS with another
company (Apple? Sun? SGI?) and turn it into yet another monster. I cite
as evidence all the previous tycoons who've testified before a senate
committee this century. The only difference is that this time it's the
computer industry, instead of oil, steel, rail, etc.
I'm unwilling to let anyone define what I value. Alas, money talks. Only
guns (and nukes) speak louder...And that's still money talking. This is
_not_ about Lisp. It's way bigger than that! The problem is so bad that
most people will think you're mad if you question the desire to make
money. My alternately plan is to just wait until a year and half, and
then ask people how much they think software can hurt them. I don't want
a body count (but we'll probably get one anyway), but number with a lot
of zeros will help wake up the bean counters.
Perhaps I won't actually need any nukes...Money talks.
Followups adjusted.
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
From: Erik Naggum
Subject: Re: free market bashing (was Re: Harlequin vs. Allegro)
Date:
Message-ID: <3107946235085816@naggum.no>
* Martin Rodgers
| The problem is so bad that most people will think you're mad if you
| question the desire to make money.
if you can't understand, then just _memorize_ that that's not the issue
at all, you perennial dimwit.
#:Erik
--
http://www.naggum.no/spam.html is about my spam protection scheme and how
to guarantee that you reach me. in brief: if you reply to a news article
of mine, be sure to include an In-Reply-To or References header with the
message-ID of that message in it. otherwise, you need to read that page.
In article <················@naggum.no>, ······@naggum.no says...
> * Martin Rodgers
> | The problem is so bad that most people will think you're mad if you
> | question the desire to make money.
>
> if you can't understand, then just _memorize_ that that's not the issue
> at all, you perennial dimwit.
Have you anything better to offer than bile?
--
Please note: my email address is munged; You can never browse enough
Not coming to you from Glastonbury
From: Erik Naggum
Subject: Re: free market bashing (was Re: Harlequin vs. Allegro)
Date:
Message-ID: <3107964015015320@naggum.no>
* Martin Rodgers
| Have you anything better to offer than bile?
what you do think you deserve? you change the subject to "free market
bashing" as if anyone but you had a problem with the free market and you
direct followups to alt.fan.bill-gates and comp.os.ms-windows.advocacy,
as if anyone but you would like to post to such moronic newsgroups.
why do you whine every time you get what you deserve but never learn to
do something different? you don't _have_ to post stupid drivel, do you?
#:Erik
--
http://www.naggum.no/spam.html is about my spam protection scheme and how
to guarantee that you reach me. in brief: if you reply to a news article
of mine, be sure to include an In-Reply-To or References header with the
message-ID of that message in it. otherwise, you need to read that page.
In article <················@naggum.no>, ······@naggum.no says...
> * Martin Rodgers
> | Have you anything better to offer than bile?
>
> what you do think you deserve? you change the subject to "free market
> bashing" as if anyone but you had a problem with the free market and you
> direct followups to alt.fan.bill-gates and comp.os.ms-windows.advocacy,
> as if anyone but you would like to post to such moronic newsgroups.
Nice ranting, Erik. It belongs in an advocacy newsgroup, with all the
other politics. Is comp.lang.lisp about MS or a programming language?
> why do you whine every time you get what you deserve but never learn to
> do something different? you don't _have_ to post stupid drivel, do you?
You don't have to post such bile, do you? I'm serious. I'm not whining.
As I explained, I expect the problem to go away in a few years, or at
least radically change. Money talks, and the market listens. All you have
to do is show how using Lisp leads to making more money. If the gain is
big enough, the market will follow. If you rant and call all businesses
using MS software stupid, don't be surprised if such people write you off
as a loony. It's all in the message. Make it sweet, and they'll listen.
Make it sour, and they'll ignore you. Simple human nature.
The problem is that most people see the world in terms of money and a 5
year business plan. Why should we expect them to appreciate the long term
effects of poor software? The irony is that, in a few years, such
software could easily result in the bean counters getting their butts
kicked very hard.
So, my plan is simple. Wait until the butt kick starts, say "I told you
so", and offer some alternatives to the current "popular" tools.
Apparently, Mrs Thatcher said, "You can't buck the market". While I'd
disagree with most of the things she's said, I don't deny that some
people may still believe it. All I'm saying is that we might use the
market. Let it choose better software.
Is this whining or a plan to constructively change the world?
--
Please note: my email address is munged; You can never browse enough
Not coming to you from Glastonbury
From: Erik Naggum
Subject: Re: free market bashing (was Re: Harlequin vs. Allegro)
Date:
Message-ID: <3108022262619929@naggum.no>
* Martin Rodgers
| Nice ranting, Erik.
get the hell out of my face, Martin Rodgers! go find somebody else to
pester with your depressingly moronic whining and disgusting obsequity.
if you want my wholehearted approval for something you do, try suicide.
#:Erik
--
http://www.naggum.no/spam.html is about my spam protection scheme and how
to guarantee that you reach me. in brief: if you reply to a news article
of mine, be sure to include an In-Reply-To or References header with the
message-ID of that message in it. otherwise, you need to read that page.
In article <················@naggum.no>, ······@naggum.no says...
> get the hell out of my face, Martin Rodgers! go find somebody else to
> pester with your depressingly moronic whining and disgusting obsequity.
> if you want my wholehearted approval for something you do, try suicide.
Ah, more bile. This is your "go away your stupid" response. Ignore what I
say, pretend that we have nothing in common, and we don't both use Lisp?
I don't ask for your approval; I merely refuse your bile. I offer a
positive way to attack MS, while you only offer your hate. I point out
that bashing MS isn't enough, that we must kill the ideas that created
and support MS, and you flame me. Am I too extreme for you, or not
extreme enough?
I say beware of not merely Gates, but also Murdock. Not even Murdock can
question the power of the market, of the dollar. If crap software costs
companies enough money, they'll realise that they've been paying money to
the wrong people. It's a simple plan: wait a year and a half and then
say, "Told you so".
Or are you naive enough to think that discussing the failings of MS on
UseNet will change as much as a global software disaster?
The revolution begins on Jan 1, 2000. See you there.
--
Please note: my email address is munged; You can never browse enough
Not coming to you from Glastonbury
From: Erik Naggum
Subject: Re: free market bashing (was Re: Harlequin vs. Allegro)
Date:
Message-ID: <3108056295654868@naggum.no>
* Martin Rodgers
| I don't ask for your approval; I merely refuse your bile.
so refuse it in a way that _works_! geez, how stupid _are_ you?
one good way to refuse my "bile" is to stop being such an annoying pest
that "bile" is the only thing you deserve. you can stop posting more of
the insane drivel you constantly impute to me; just stop telling me what
I think and stick to telling us what _you_ think, if you do -- you appear
to have the mental capacity of a dodo, but only because it's extinct, and
I'd vastly prefer if you matched it in that particular capacity, as well.
_just_ _go_ _away_, OK?
damn, there's no alt.fan.martin-rodgers newsgroup.
#:Erik
--
http://www.naggum.no/spam.html is about my spam protection scheme and how
to guarantee that you reach me. in brief: if you reply to a news article
of mine, be sure to include an In-Reply-To or References header with the
message-ID of that message in it. otherwise, you need to read that page.
In article <················@naggum.no>, ······@naggum.no says...
> * Martin Rodgers
> | I don't ask for your approval; I merely refuse your bile.
>
> so refuse it in a way that _works_! geez, how stupid _are_ you?
Like using a killfile? Sure, but most of your posts are worth reading. I
could perhaps write a rule that kills anything by you that contains
"Microsoft" or "Bill Gates".
> one good way to refuse my "bile" is to stop being such an annoying pest
> that "bile" is the only thing you deserve. you can stop posting more of
> the insane drivel you constantly impute to me; just stop telling me what
> I think and stick to telling us what _you_ think, if you do -- you appear
> to have the mental capacity of a dodo, but only because it's extinct, and
> I'd vastly prefer if you matched it in that particular capacity, as well.
> _just_ _go_ _away_, OK?
Do you mean stop being so reasonable? You may disagree with what I say,
which is unfortunately not always good news, but that hardly makes me
stupid. I will not go away simply because you can't cope with it.
In case you've missed the point, I'm not a fan of Microsoft. I don't
choose to use their software. Nor do many other people. My message is
merely that. I know it's bad news that so many people use MS software.
So, it looks like you prefer to kill the messenger. I prefer to look at
realistic ways to change the situation.
We agree about the problem, we just disagree about the solution. Most
people can do that without name calling.
If you think that this is whining, then I doubt that you've read a word
I've written. I'm not attacking you. The worst that I might be guilty of
is wondering how many people will embrace Lisp when MS and all their
works are removed the surface of the planet. If you can't imagine that MS
might not be guilty for all the obstacles in Lisp's path, then you've
probably forgotten which company gave us C++ and Java, nor wondered why
anyone might stop using Lisp and switch to another language and looked
for a plausible reason. I don't for a moment suggest that C++ is superior
to Lisp, but I can still admit that some intelligent people _might_.
Nor am I alone. ISTR Kent Pitman making an observation on this subject,
not too many months ago. Was that also whining?
You're the one complaining about MS in a newsgroup for Lisp discussions.
> damn, there's no alt.fan.martin-rodgers newsgroup.
So create one. Stop whining.
Followups set to a more appropriate newsgroup. Killfile updated.
--
Please note: my email address is munged; You can never browse enough
Not coming to you from Glastonbury
In article <·························@news.demon.co.uk>,
···@wildcard.this.email.address.intentionally.left.crap.demon.co.uk (Martin Rodgers) writes:
> I'm working of ridding the world of Microsoft. I'll start by
> assassinating Bill Gates, and then try to obtain a tactical nuke so I can
> take out Seatle.
> The risk is that the market will attempt to put replace MS with another
> company (Apple? Sun? SGI?) and turn it into yet another monster.
You can take SGI off that list. They've already gone over to the dark side.
Mike McDonald
·······@mikemac.com
In article <····················@news.teleport.com>, ·······@mikemac.com
says...
> You can take SGI off that list. They've already gone over to the dark side.
If I may make an alternative cultural reference, they've not yet taken
the Ring from MS. For now, it's still Gates who is saying, "All shall
love me and dispair."
However, I agree about SGI's current position, which is essentially that
of Saruman. I now wince whenever I see one of their "old" machines on TV.
An era has ended, albeit that of just one company. It just happens to be
a vendor of Unix machines with a _very_ high profile.
If Symbolics were in a similar position, I might mourn more. Oops, I
guess they are. Still, until MS release a 64bit version of NT, there's no
chance of Symbolics following SGI down the same path.
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
From: Rob Warnock
Subject: Re: free market bashing (was Re: Harlequin vs. Allegro)
Date:
Message-ID: <6na4mf$7k24t@fido.engr.sgi.com>
Martin Rodgers <···@...munged_address...> sed:
+---------------
| Still, until MS release a 64bit version of NT, there's no
| chance of Symbolics following SGI down the same path.
+---------------
Hey, until MS releases a 64-bit version of NT that runs anywhere *close*
to as well as Irix on 128-processor single-system-image CC/NUMA machines,
there's no chance in hell that SGI'll stop supporting/developing Irix...
at least, on the high-end server systems. [So don't hold your breath.]
Though as has been announced in the press, the single-user desktop system
is a different story.
-Rob
-----
Rob Warnock, 7L-551 ····@sgi.com http://reality.sgi.com/rpw3/
Silicon Graphics, Inc. Phone: 650-933-1673
2011 N. Shoreline Blvd. FAX: 650-933-4392
Mountain View, CA 94043 PP-ASEL-IA
In article <············@fido.engr.sgi.com>, ····@rigden.engr.sgi.com
says...
> Hey, until MS releases a 64-bit version of NT that runs anywhere *close*
> to as well as Irix on 128-processor single-system-image CC/NUMA machines,
> there's no chance in hell that SGI'll stop supporting/developing Irix...
> at least, on the high-end server systems. [So don't hold your breath.]
Exactly. While it would be a chance, it would be a bloody remote one!
> Though as has been announced in the press, the single-user desktop system
> is a different story.
So it is.
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
From: Randy Sims
Subject: Re: free market bashing (was Re: Harlequin vs. Allegro)
Date:
Message-ID: <1998Jun30.062647.10551@srs.gov>
····@rigden.engr.sgi.com (Rob Warnock) writes:
> Martin Rodgers <···@...munged_address...> sed:
> +---------------
> | Still, until MS release a 64bit version of NT, there's no
> | chance of Symbolics following SGI down the same path.
> +---------------
>
> Hey, until MS releases a 64-bit version of NT that runs anywhere *close*
> to as well as Irix on 128-processor single-system-image CC/NUMA machines,
> there's no chance in hell that SGI'll stop supporting/developing Irix...
> at least, on the high-end server systems. [So don't hold your breath.]
>
As an aside, what commercial lisp runs natively mutli-threaded on that
128-processor system?
Randy.
>
> Though as has been announced in the press, the single-user desktop system
> is a different story.
>
>
> -Rob
>
> -----
> Rob Warnock, 7L-551 ····@sgi.com http://reality.sgi.com/rpw3/
> Silicon Graphics, Inc. Phone: 650-933-1673
> 2011 N. Shoreline Blvd. FAX: 650-933-4392
> Mountain View, CA 94043 PP-ASEL-IA
--
Randal N. Sims (Randy) | Tel: (803)725-1387
Westinghouse Savannah River Co. | Fax: (803)725-8829
SRS, 773-A, A1128, Rm. 2 | Email: ···········@srs.gov
Aiken, SC 29808-0001 USA | URL: http://www.srs.gov (generic)
DISCLAIMER: The opinions expressed are mine and do not necessarily
represent Westinghouse Savannah River Co. or the
United States Department of Energy.
From: Rob Warnock
Subject: Re: free market bashing (was Re: Harlequin vs. Allegro)
Date:
Message-ID: <6nc5k5$7qeoh@fido.engr.sgi.com>
Randy Sims <···········@srs.gov> wrote:
+---------------
| ····@rigden.engr.sgi.com (Rob Warnock) writes:
|
| > [NT] as well as Irix on 128-processor single-system-image CC/NUMA machines,
| > there's no chance in hell that SGI'll stop supporting/developing Irix...
|
| As an aside, what commercial lisp runs natively mutli-threaded on that
| 128-processor system?
+---------------
Actually, I'd like to know that myself. (*sigh*)
-Rob
-----
Rob Warnock, 7L-551 ····@sgi.com http://reality.sgi.com/rpw3/
Silicon Graphics, Inc. Phone: 650-933-1673
2011 N. Shoreline Blvd. FAX: 650-933-4392
Mountain View, CA 94043 PP-ASEL-IA
···@wildcard.this.email.address.intentionally.left.crap.demon.co.uk (Martin Rodgers) writes:
> If I may make an alternative cultural reference, they've not yet taken
> the Ring from MS. For now, it's still Gates who is saying, "All shall
> love me and dispair."
Comparing Bill Gates and Galadriel should be a capital offense ;-)
--
Lieven Marchand <···@bewoner.dma.be>
------------------------------------------------------------------------------
Few people have a talent for constructive laziness. -- Lazarus Long
In article <············@xenon.inbe.net>, ···@bewoner.dma.be says...
> Comparing Bill Gates and Galadriel should be a capital offense ;-)
True. That's why I was comparing Gates to Sauron, and Galadriel to who
ever you think the "good guys" are. ;)
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
Martin Rodgers wrote:
[someone else said:]
>> Comparing Bill Gates and Galadriel should be a capital offense ;-)
>
> True. That's why I was comparing Gates to Sauron, and Galadriel to who
> ever you think the "good guys" are. ;)
I'm not sure I quite believe we're discussing this in c.l.l,
but: it was Galadriel who said `all shall love me and despair'[1]
(although of course she didn't stay in that frame of mind for
long[2]), which is presumably why you were accused of comparing
her with Big Bad Bill when you suggesting he was saying it.
Um. I haven't read that book for about ten years. I wonder
whether any of what I'm saying is right. And I wonder whether
it has any possible connection with Lisp.
[1] When she was offered the Ring, yes?
[2] `I shall diminish, and go into the West, and remain Galadriel'.
--
Gareth McCaughan Dept. of Pure Mathematics & Mathematical Statistics,
·····@dpmms.cam.ac.uk Cambridge University, England.
In article <··············@g.pet.cam.ac.uk>, ·····@dpmms.cam.ac.uk
says...
> Martin Rodgers wrote:
>
> [someone else said:]
> >> Comparing Bill Gates and Galadriel should be a capital offense ;-)
> >
> > True. That's why I was comparing Gates to Sauron, and Galadriel to who
> > ever you think the "good guys" are. ;)
Those last two lines were definitely me. ;)
> I'm not sure I quite believe we're discussing this in c.l.l,
> but: it was Galadriel who said `all shall love me and despair'[1]
> (although of course she didn't stay in that frame of mind for
> long[2]), which is presumably why you were accused of comparing
> her with Big Bad Bill when you suggesting he was saying it.
Yes, this is exactly what I was saying. While most others, like Gandalf,
merely hint at what they'd do with the Ring, Galadriel makes it explicit.
Boromir - and his father - think that they might master and control the
Ring, when instead it would be _their_ master.
Now, who was it who said that, "We can't afford to be less ruthless than
the opposition"? I think it was a comment on the Cold War. Tolkien seemed
to be saying that if we go down that path, we lose. We should destroy the
Ring, rather then use it ourselves.
I interpret that to do mean, in the context of the computer industry,
that we need to change the demands of the market. Destroy the myth that
"good enough" is profitable. Put a figure on the _true_ cost.
My loathing of the current state of affairs began in the early 80s, not
with MS - who were at that point small, insignificant, and still run by
talented people - but Intel. While Intel weren't big in the Bus Wars
strip, it does document some of the flaws in the industry. Important
technical decisions are not always influenced by technical issues, and
this strip gives examples. It's almost a techie version of Dilbert.
Perhaps I'm flattering Gates by comparing him to Sauron. Gates could be
just one of the Ring Wraiths. Lord of the Nazg�l? Hmm.
> Um. I haven't read that book for about ten years. I wonder
> whether any of what I'm saying is right. And I wonder whether
> it has any possible connection with Lisp.
Only in the sense that the problems of the computer industry are bigger
than any one company. This is much like saying that the problems of the
world are bigger than any one country. It's easy for some people to just
blame, say, the US. (Some people blame the British Empire, which IMHO is
thankfully dead. Alas, new empires arise.) The real problems are hard to
conceive of, due to their scale.
So, where is our Mount Doom, into which we can cast the Ring?
> [1] When she was offered the Ring, yes?
Yes, offered to her by Frodo. Just as he offered it to Gandalf. Those
were who wise, refused it. Those who desired it, perished.
By contrast, Scott Adams is a lot more cynical. Look at what Dogbert is
currently doing...
> [2] `I shall diminish, and go into the West, and remain Galadriel'.
That sounds a lot like IBM. ;) Still big in the mainframe world, tho.
Anyway...
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On 30 Jun 1998 21:46:27 +0200, Lieven Marchand <···@bewoner.dma.be>
wrote:
>Comparing Bill Gates and Galadriel should be a capital offense ;-)
Well, Galadriel is no saint. She was among the Noldor who came back to
the Middle Earth; and though she didn't participate, she certainly was
a witness of the killing of the Teleri, when the boats were robbed;
there's blood on her hands.
Now, if you were speaking of her brother Finrod, that's a nice guy;
only half-decent elf in all of Tolkien's works :)
/L/e/k/t/u
-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 5.5.3i; see <http://www.pgpi.com>
iQA/AwUBNZyFm/4C0a0jUw5YEQJuswCfV04ni0TD5SHWBHCWkiMAI9D7BK0AoPTT
5NdHp5q+dciEkRKpxA164Ovh
=lu8J
-----END PGP SIGNATURE-----
···@wildcard.this.email.address.intentionally.left.crap.demon.co.uk (Martin Rodgers) writes:
> That just leaves all the people who helped make MS what they are today,
> i.e. the people who bought their software. My personal believe - please
> correct me if I'm wrong - is that the demands of the free market have
> made MS into the competitive monster that it is. Just like every other
> corporation. So, more tactical nukes will be needed.
Words are much stronger weapons than nukes or money will ever be,
because words can change people, weapons only kill them. The next
history book will bear this out in all it's glory and all it's
tragedy. They might be slower weapons, but stronger none the less.
If you disagree with the above, then I think that Microsoft or the
programming language of your choice is your smallest problem, since
then the future of mankind looks very bleak indeed.
Regs, Pierre.
--
Pierre Mai <····@cs.tu-berlin.de> http://home.pages.de/~trillian/
"Such is life." -- Fiona in "Four Weddings and a Funeral" (UK/1994)
In article <··············@torus.cs.tu-berlin.de>, ····@cs.tu-berlin.de
says...
> Words are much stronger weapons than nukes or money will ever be,
> because words can change people, weapons only kill them. The next
> history book will bear this out in all it's glory and all it's
> tragedy. They might be slower weapons, but stronger none the less.
Agreed. Perhaps you missed the irony. If you read to the end of that
post, you might see that I make exactly the same point as you.
> If you disagree with the above, then I think that Microsoft or the
> programming language of your choice is your smallest problem, since
> then the future of mankind looks very bleak indeed.
I not only agree with you, but I suggest that flames and hatemail are no
more constructive than the threat of nukes. I strongely recommend that
you re-read my post, keeping in mind that it is a comment on perception.
I don't wish anyone to equate criticisms of MS as radical politics. Read
all of mt post and you should see that. Perhaps I should've made the use
of irony more obvious. At times, any discussion of MS is as provocative
as discussion of the "Troubles", as we quaintly call them here in the UK.
It's important to realise that there is no "right" side in this kind of
politics, and it's deeply harmful to any communication to suggest
otherwise. This is why I'm suspicious of any hatemail; I've heard too
much of it from the Real World. And it continues...
As I've said before, I don't believe in magic bullets. There will be no
simple solution to this problem. No "peace process" will heal the wounds
overnight. The damage is way too profound - on all sides.
And finally, I don't believe that hatemail of any kind belongs in
comp.lang.lisp. It only gets in the way of the real purpose: Lisp.
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
* Zeno the Anonymous Poster
| I assume this is because it is better to help the underdog.
I think this line pretty much sums up your whole line of inquiry, here.
so, why did I hammer on your Microsoft affiliation? I doubt that I could
have gotten you to express the above sentiment in so many words unless I
had hammered real hard on it. I have been doing custom software since
1980 (starting my own business in 1987), and I have _never_ worked under
the Microsoft regime, but I have had ample opportunity to watch it from
the outside and see projects to which I was invited crumble or fail, and
I have watched people blame _everybody_ but the people really to blame:
Microsoft, for their sheer lack of quality products and responsibility
towards their customers. their customers have _always_ believed that
they have been to blame if something failed to work right, and the lies
that Bill Gates personally have served his every partner have likewise
been attributed to his "smartness" and others have just been stupid to
believe him. I saw this pattern _very_ early on, and _that_ is why I
want nothing of their ilk or fandom, and I have acquired an eerie skill
in predicting the few categories Microsoft users can be in. my distinct
impression is that managers who insist on Microsoft products don't want
to _succeed_, they just want to avoid failing in a world they don't
understand, and as long as Microsoft stays afloat, they think they will
stay afloat, too. funny as it may seem, fin-de-si�cle syndromes make
managers act weird, too. the Year 2000 problem is little more than the
age-old myth that the end of world coincides with new centuries, and my
guess there are still so many managers who haven't reacted in time is
that they don't actually believe there will be any life on planet earth
in new millennium, anyway, so why waste all the money? think about it.
however, I must admit that you surprised me a bit today, and that's quite
a feat for a hardened cynic like myself. I didn't know it was _possible_
for anybody to be so arrogant and so stupid at the same time as to
express the above sentiment towards a "competing" product or solution,
and so utterly lacking in understanding of basic economics. sorry to be
harsh, but you need to snap out of your dream and _listen_, not just to
the stuff you are prepared to listen to, but to the underlying arguments.
when you pay for something, there is a tacit assumption that you somehow
value whatever you pay for higher than the money you part with, and an
explicit assumption that whoever parts with the goods values your money
higher than the goods he parts with. the explicit assumption is well
known and nobody argues its validity. the tacit assumption is what makes
the price he charges possible in his market, and the science of marketing
and market management is founded in psychology so murky that you would
prefer it to be unknown to you as a customer, but you _need_ to know.
these tacit assumption need to be made explicit in order to change the
direction and focus of a market, and that's where Microsoft's marketing
is at its very best: they address your belief system, not their own
products. this is especially important for people who don't realize that
these assumptions are there to begin with, and that includes you, "Zeno".
a few people, yours truly among them, argue in various fora that kids
should be exposed to psychology from kindergarten up to withstand the
mind-wiping techniques of twenty-first century marketing, and although
this sounds paranoid to people who know nothing about marketing, those
who do and who know how the mass media work have been trying to alert the
sleeping masses for many decades. I'm trying to alert you now. it will
fail with a 98% probability (a statistical fact), because your belief
system is so constructed as to block any suggestion that those beliefs
are manufactured by others and not your own. the consensus among your
peers, upon which you base your judgments, is not your own, it has been
deliberately manufactured by those who benefit from it, from politicans
to businesses via religious leaders and news anchors. however, it is
_not_ a conspiracy, it is _not_ an evil plot to subdue the masses (as if
they weren't to begin with!), and it is _not_ a take-over plot by aliens
visiting earth, so listen up. this is the _natural_ development of mass
media at work in huge societies, and it could not have become otherwise;
the only way _not_ to get where we are would have been to destroy the
onset of mass communication, which would never have been tolerated by the
same masses that are controlled by it today, in exactly the same way that
tobacco and the automobile would have been prohibited immediately had
they been proposed today, together with their attendant costs and loss of
human lives, but cannot be removed from society today.
the tacit assumption at work in your world is that a development system
has a _fixed_ upper limit to its value to developers and that leads to
the sentiment that you are willing to pay anything _below_ that upper
limit, but that you would not feel you would get your money's worth if
you paid more -- you would, in your own judgment, be better off keeping
your money and being without that development environment. put even more
strongly, albeit less certain, a development system has a fixed _value_
in your judgment: you know what you can get out of the development
environment you use today, and you cannot imagine, nor will you listen to
testimonials, that it is possible to be an order of magnitude more
productive in another, nor will you consider slightly different modes of
operation that would make your development environment look like the
cheap plastic toy it is to many other people. the result of your tacit
assumptions is that it would not be possible to sell you anything more
_valuable_ than your current development environment. thus you see the
world as fundamentally limited, and to you, it _would_ be helping some
underdog who overcharges for his products, not because they do, but
because _you_ are unable to see the extra value they charge for, and
which _others_ accept as worth more than the money they part with.
the key question when it comes to your Microsoft affiliation, if not
marriage, is that you have "let" Microsoft implant in you the upper value
of the goods that its _competitors_ could sell you, but not the price of
their own products. the _problem_ is that you have not _let_ this happen
to you in any conscious way, of course -- it has been shaped by the
extremely talented and equally manipulative people at Microsoft who have
successfully set the entire agenda for the PC industry trade rags and
(almost) all of their journalists. in particular, the history of the
quite fantastic marketing of the vaporware "Windows" product is worth
studying for decades to come. how _could_ a fraudulent little fart in
Seattle manage to con the whole software world into believing he would
release a product and then not do it, over and over and over? why did
_anyone_ believe him?
to the people who are aware of the assumptions that are tacit among the
sleeping masses, it is no wonder at all that Microsoft succeeds: Bill
Gates is very good at playing the right lullabies at the right time so
people who wake up and smell the coffee (a blistering asphalt by now),
they are calmed down and go back to sleep for another marketing cycle.
those who are unable to understand or appreciate the dire consequences of
tacit assumptions that go completely unchallenged, appear to the sleeping
masses as if they are on cocaine or paranoid or think they are sleepless
over Seattle. Microsoft is so good at turning off the alarm clock that
it has become impossible to be concerned about their business ethics
without being associated with something the tacit assumptions elsewhere
say is "impossible" or "impractical" or "insane", like high quality
software, communism, or anti-innovation.
| One of my concerns was with the price of Lisp development compared to VB.
precisely, and you are not at all concerned with its _value_ to you,
because the tacit assumption in your world is that you _know_ the value
of any and all development environments, you _know_ all it takes to use
one productively, you _know_ how much you can squeeze out of it, and you
know _exactly_ what you want out of it, too. those tacit assumptions are
so tacit as to be engraved on your MS Vertebra 8.0. nothing could change
your ways, your means of doing business, or your customer base. that you
and millions of other developers think this way is how Microsoft benefits
tremendously from your solidifying your tacit assumptions by being an
arrogant tourist in Common Lisp-land. when Microsoft says "jump", you
rush to compute the optimal height and go for it. when Microsoft falls,
you fall. then, when Microsoft doesn't get up, you will _not_ blame Bill
Gates for it like you should but never did in the past, either, you will
_not_ blame yourself for being so unfuckingbelievably gullible as to buy
his crap and let your own mind turn to mush with their tacit assumptions.
but you _will_ retain all the tacit assumptions that brought you down
intact and you _will_ blame the Department of Justice for "meddling" with
"internal affairs" and you _will_ continue to think that everybody who
doesn't like Microsoft's business practices must also _hate_ them and at
least be _somewhat_ irrational, if not completely gaga. however, _I_
have seen people go nuts over the fact that I refuse to work under the
Microsoft regime and effectively boycott them (although it never gets to
that in practice), as if the fact that I _dare_ to confront them puts
their own belief system in jeopardy and they need to defend themselves
from the living proof that you _don't_ die instantly if you don't believe
in Bill Gates' rhetoric, which it seems is what they actually fear.
_such_ is the set of tacit assumptions that defines the upper limit to
the price of a development system in your world, and _this_ is why you
are not going to become a Franz Inc customer, nor a Common Lisp user, nor
_ever_ understand why some people think Franz Inc's pricing policy for
the Microsoft world is _dangerously low_. you see, to some people,
charging 40% less for a lot more software to people who have already
bought into the tacit assumptions emanating from Seattle is caving in to
the price spiral and the "nah, it cannot _possibly_ be worth that much
money"-attitude that you have so well exposed to us. does the Lisp world
_need_ people who only consider the price and not value-for-money? no.
does the Lisp world _need_ to become ephemeralized the way the Microsoft
world is, where nothing ever lasts till next year? no. a product that
bears a publication date of May 1996 is _not_ out of date in the Lisp
world -- but it would be ancient history to Microsoftians because their
world doesn't really exist except as figments of marketing, and you can't
repeat a marketing lie just any number of times, it _has_ to be upgraded,
or you will stop believing it, you will see your tacit assumptions and
you will question Microsoft, lose your belief system, and die instantly.
finally, the reason I'm valuable to my clients and the reason I can do
all the fun stuff I do is that I'm not afraid to challenge _any_ tacit
assumption or to seek answers to the questions people don't want me or
anybody else to ask, neither of themselves nor of their peers. only that
way can I figure out what people _really_ think and want (and do), and
most of the time, it comes down to much less expensive systems and
solutions than they were brought up by the big consulting firms and
Microsoft to believe it would cost. the tacit assumption I sometimes
find it hard to overcome is "nah, it cannot _possibly_ be accomplished
with that little money or that few people", but this can sometimes be
fixed just by charging more money for it...
#:Erik
--
http://www.naggum.no/spam.html is about my spam protection scheme and how
to guarantee that you reach me. in brief: if you reply to a news article
of mine, be sure to include an In-Reply-To or References header with the
message-ID of that message in it. otherwise, you need to read that page.
In article <················@naggum.no>, Erik Naggum <······@naggum.no> wrote:
[snip!]
> * Zeno the Anonymous Poster
> | One of my concerns was with the price of Lisp development compared to VB.
>
> precisely, and you are not at all concerned with its _value_ to you,
[snip!]
A modern version of Perlis' quip that Lisp programmers know the
value of everything but the cost of nothing. :)
Sashank (fellow Common Lisp programmer)
* Zeno the Anonymous Poster
| You say that you hammered on my Microsoft affiliation so that I would
| express the sentiments I did.
no, that's not what I said, and that's not what I did. I'm beginning to
realize it's a fucking waste of time to write anything to you at all, but
please, in the interest of understanding what I _have_ said, realize the
difference between observing an unexpected result that could not have
happened without a specific antecedent which had its own purpose and
causes, and an action intended to cause the exact same result. just
because you see the result doesn't mean it had to come from the set of
causes your failing brain imagines. this is _extremely_ elementary.
please also refrain from rewriting what other people say and making
stupid summaries of what you fail so utterly to understand. it may be
entertainment to you, just as it is _expected_ of people who have to
defend their very core beliefs, but can't and won't, always find cause to
laugh or find themselves "entertained", but that doesn't mean it wasn't
an attempt to reach a working brain at least somewhat open for new views
(which failed, with the 98% expected probability, which you went on to
prove, for reasons I do not understand -- usually people pull themselves
together and try real hard to be in the select crowd that "gets" an idea
in the face of such statistics -- perhaps the ideas were just to much for
your brain to handle all at once and you _had_ to assume it wasn't real).
| I would have to think extremely hard, and have the help of some mushrooms
| to understand that managers are buying Microsoft products because there
| will be no life on planet earth in a short few years.
excuse me? whoever _wrote_ this insane bullshit but _yourself_? does
your brain work, not by thinking, integrating, and analyzing ideas, but
by juxtaposing random sentences and declaring yourself "entertained" by
results like the above? does your brain run Visual Basic?
| Yes, you certainly do have the most open, inquisitive mind I have ever
| encountered. Not set in your ways, you're not--oh, no.
thank you for the final proof of the upper limit to your capabilities.
it saves me a lot of time and effort to know that I would have been
dealing with someone whose ability to accept information and ideas
contrary to his tacit assumptions is _exactly_ zero.
#:Erik
--
http://www.naggum.no/spam.html is about my spam protection scheme and how
to guarantee that you reach me. in brief: if you reply to a news article
of mine, be sure to include an In-Reply-To or References header with the
message-ID of that message in it. otherwise, you need to read that page.
In article <·················@news.deltanet.com>,
Zeno <····@deltanet.com> wrote:
>On 25 Jun 1998 15:03:32 +0000, Erik Naggum <······@naggum.no> wrote:
>> it has become impossible to be concerned about their business ethics
>> without being associated with something the tacit assumptions elsewhere
>> say is "impossible" or "impractical" or "insane", like high quality
>> software, communism, or anti-innovation.
>
>Your beliefs that communism and anti-innovation are worthwhile
>endeavors does lend more insight to your hatred of one man for making
>so much more than the others, and to your rejection of new technology.
Erik is concerned about Microsoft's business ethics. Zeno has just
associated Erik with high quality software, communism and anti-innovation.
This hasn't proved Erik's assertion, but it's certainly consistent with it.
-dan
In article <··········@fishy.ox.compsoc.net>, ···@fishy.ox.compsoc.net
says...
> Erik is concerned about Microsoft's business ethics. Zeno has just
> associated Erik with high quality software, communism and anti-innovation.
>
> This hasn't proved Erik's assertion, but it's certainly consistent with it.
There are also people who call Richard Stallman a communist. I guess some
ideas are hard for some people to understand.
I've heard exactly the same kind of misunderstandings made on a radio
phone-in show, simply because the presenter has some creative ideas. To
many, they're also provocative ideas.
Now, business is a very different world to talk radio, which here in the
UK is considerably more gentle. Why then are people so hostile to new
ideas? Perhaps any suggestion of change is seen as provocative? Could
that explain why people "redmist", and hear a totally different idea?
For example, mistaking the freedom to use GPL'd software as communism?
Like Erik, I too question Microsoft's business ethics. I also question
the policies of my country's government, as do a lot of other people.
How do we "vote" for better business ethics without sounding like
socialists? (A very real problem here in the UK.)
I think we're preaching to the converted - apart from Zeno, that is.
--
Please note: my email address is munged; You can never browse enough
Not coming to you from Glastonbury
* Martin Rodgers
| How do we "vote" for better business ethics without sounding like
| socialists?
you buy from somebody else. just how hard can it be to grasp this?
#:Erik
--
http://www.naggum.no/spam.html is about my spam protection scheme and how
to guarantee that you reach me. in brief: if you reply to a news article
of mine, be sure to include an In-Reply-To or References header with the
message-ID of that message in it. otherwise, you need to read that page.
In article <················@naggum.no>, ······@naggum.no says...
> you buy from somebody else. just how hard can it be to grasp this?
Have you noticed what people are buying? I may buy non-MS (as I have),
but what difference does that make? Small, but significant. I've not
bought _any_ MS software, other than what came bundled with one machine.
Nor will I willingly buy any. Ever. Small but significant.
Has the world changed yet? I doubt it. I expect to see a bigger change
in, say, two years from now. A _much_ bigger change. My sig file will be
saying, "Y2K? See, I told you so!" Small but significant.
Steel, oil, rail...And software.
--
Please note: my email address is munged; You can never browse enough
Not coming to you from Glastonbury
* Martin Rodgers
| I may buy non-MS (as I have), but what difference does that make?
the only difference that counts. you can't spend anybody else's money
any differently, only your own. if you don't want to persuade others on
the individual level where this principle applies, you have to play with
big marketing money and work to shift public sentiments (which you have
demonstrated that you don't want to understand how works, so let's just
forget that others do). if you don't want to do what it takes, just get
out of the game entirely -- you can only communicate your defeat and
resignation if you are as inefficacious as you imply in every whimpering,
whining article here, and there's no wonder you feel bad.
| Has the world changed yet?
yes, of course it has, but you insist on measuring "the world" with a
resolution that guarantees that you wouldn't see any differences until
something that hits you in the head measures 7.0 on Richter's Scale.
there's a description of depression that says that being depressed is all
about not seeing the small and insignificant tidbits of happiness around
you that all the significant ones are made of while you're getting more
and more depressed waiting for the significant one that could take you
out of your depression, but it never comes, because everything is too
small and too insignificant to "count", and downward it goes from there.
the same principle applies to any form of being and feeling efficacious,
and especially to sales forces who must regard every _single_ sale as
more significant than all the rejections before it, because they only
make one sale at a time, and it's all they can do to keep afloat to let
each sale vindicate their efforts.
if you can't handle this psychological requirement of "evangelizing" or
selling good ideas or products or yourself or whatever, don't even _try_
to pretend that your sense of defeat or inefficacy has anything to do
with anything but yourself. it's you, and _only_ you, that is the
problem if you can't make such a sale --- just wise up to the fact that
you're just not cut out for it and leave the scene to those who _can_
handle it. being a good salesman is all about first selling yourself the
idea that you _can_ actually handle this, and then letting every sale
reinforce that belief. but lose that crucial faith, and you're history,
no matter how many sales you make or don't make. it's _all_ psychology.
the world changes with every significant action you make. if you feel
the world doesn't change when you act, either your actions are truly
insignificant, or you're just depressed. now you can go hang yourself or
you can go do something significant. my suggestion if you want to keep
whining and bemoaning the state of the world is to put yourself out of
your misery as soon as possible (please let me try to sell you the rope),
but you _could_ also try getting out of your depression and one good way
is to start noticing what software people are actually buying. (do you
recognize that question?) every single computer in the world that sports
a Common Lisp environment is important to its programmer and its vendor
(or author); why isn't it to you? every single sale of a Linux CD-ROM is
important to the customer and the Free Software world; why isn't it to
you? every single transfer of the Association of Lisp Users web page to
somebody browser is important to the Lisp community; why isn't it to you?
#:Erik
--
http://www.naggum.no/spam.html is about my spam protection scheme and how
to guarantee that you reach me. in brief: if you reply to a news article
of mine, be sure to include an In-Reply-To or References header with the
message-ID of that message in it. otherwise, you need to read that page.
In article <················@naggum.no>, ······@naggum.no says...
> * Martin Rodgers
> | I may buy non-MS (as I have), but what difference does that make?
>
> the only difference that counts. you can't spend anybody else's money
> any differently, only your own. if you don't want to persuade others on
> the individual level where this principle applies, you have to play with
> big marketing money and work to shift public sentiments (which you have
> demonstrated that you don't want to understand how works, so let's just
> forget that others do). if you don't want to do what it takes, just get
> out of the game entirely -- you can only communicate your defeat and
> resignation if you are as inefficacious as you imply in every whimpering,
> whining article here, and there's no wonder you feel bad.
>
> | Has the world changed yet?
>
> yes, of course it has, but you insist on measuring "the world" with a
> resolution that guarantees that you wouldn't see any differences until
> something that hits you in the head measures 7.0 on Richter's Scale.
This belongs in alt.fan.bill-gates, along with all the other Microsoft
hatemail. Perhaps you've not realised that you're not the only one who
dislikes MS? Playing by your rules, I could use this to call you
clueless, but I find that insulting people isn't constructive.
It's a great way to invite the Murdock press to call us loonies. Here in
the UK, phrases like "loony left" still haunt us. It might not be
apparent outside the UK, but the party currently in power used to be
socialists. Now they're merely New Labour. One of the Old Labour MPs used
to be called "Red Ken".
This kind of politics divides and weakens us, which is just what people
like Murdock and Gates would like.
> there's a description of depression that says that being depressed is all
> about not seeing the small and insignificant tidbits of happiness around
> you that all the significant ones are made of while you're getting more
> and more depressed waiting for the significant one that could take you
> out of your depression, but it never comes, because everything is too
> small and too insignificant to "count", and downward it goes from there.
I can speak of clinical depression from direct experience; It's like not
being able to see any beauty. Imagine a world without beauty. I've seen
it. While that was a couple of decades ago, I remember it well.
> the same principle applies to any form of being and feeling efficacious,
> and especially to sales forces who must regard every _single_ sale as
> more significant than all the rejections before it, because they only
> make one sale at a time, and it's all they can do to keep afloat to let
> each sale vindicate their efforts.
Agreed. This is the problem with bean counters. They count the wrong
things. My point, which you've ignored, is that a year and half from now
those bean counters are going to get a re-education. While we can talk,
and we can code, the bean counters won't notice because they're not
counting us. Yep, you _and_ me. We're on the same side, Erik, even if we
disagree about the solutions.
My bet is that class of software bugs is going to cost the business world
a lot more than we can hope to make in our lifetimes. Loads of zeroes.
> if you can't handle this psychological requirement of "evangelizing" or
> selling good ideas or products or yourself or whatever, don't even _try_
> to pretend that your sense of defeat or inefficacy has anything to do
> with anything but yourself. it's you, and _only_ you, that is the
> problem if you can't make such a sale --- just wise up to the fact that
> you're just not cut out for it and leave the scene to those who _can_
> handle it. being a good salesman is all about first selling yourself the
> idea that you _can_ actually handle this, and then letting every sale
> reinforce that belief. but lose that crucial faith, and you're history,
> no matter how many sales you make or don't make. it's _all_ psychology.
People have been giving us this "hard sell" for years, not just for
hardware and software, but for everything. Please forgive me for being
just a little cynical. I guess the last 20 years of UK politics have
something to do with that. Who to trust, who to trust?
I don't believe in magic bullets. I do believe that money talks. Anyway,
you don't have to convince me, I'm on your side. I've been there since
the mid 80s. I've sometimes managed to convince other people, and if
that's what you mean by a "sale", then I can do it. However, no amount of
"hard sell" is as effective as an argument that let's money do the
talking. It's naive to assume that everyone will see things as you do.
> the world changes with every significant action you make. if you feel
> the world doesn't change when you act, either your actions are truly
> insignificant, or you're just depressed. now you can go hang yourself or
> you can go do something significant. my suggestion if you want to keep
> whining and bemoaning the state of the world is to put yourself out of
> your misery as soon as possible (please let me try to sell you the rope),
> but you _could_ also try getting out of your depression and one good way
> is to start noticing what software people are actually buying. (do you
> recognize that question?) every single computer in the world that sports
> a Common Lisp environment is important to its programmer and its vendor
> (or author); why isn't it to you? every single sale of a Linux CD-ROM is
> important to the customer and the Free Software world; why isn't it to
> you? every single transfer of the Association of Lisp Users web page to
> somebody browser is important to the Lisp community; why isn't it to you?
Why do you assume that I'm pro-Microsoft and anti-Linux, pro-C++ and
anti-Lisp? I've spent more money on Common Lisp than C++, dispite being
_paid_ to use C++. I'm on _your_ side. However, I also recognise that
Bill Gates isn't the only problem in the world. Some people feel just as
you do about Rupert Murdock. Not too long ago, one of them was telling me
that he refuses to use cable or satellite TV because he won't give any of
his money to Murdock. I told him that it could go to Gates instead. Gates
isn't alone in creating monopolies. His is computers, Murdock's is TV and
newspapers, Ted Turner also has a TV empire, and others in the past have
created monopolies with oil, steel, etc. The more things change, eh?
The problem is that most people see the world in terms of money and a 5
year business plan. Why should we expect them to appreciate the long term
effects of poor software? The irony is that, in a few years, such
software could easily result in the bean counters getting their butts
kicked very hard.
So, my plan is simple. Wait until the butt kick starts, say "I told you
so", and offer some alternatives to the current "popular" tools.
Apparently, Mrs Thatcher said, "You can't buck the market". While I'd
disagree with most of the things she said, I don't deny that some people
may still believe it. All I'm saying is that we might use the market. Let
it choose better software.
We can win. We're merely disagreeing about tactics. I like using the
tools of the opposition - corporate fears and greed - to work for _us_.
--
Please note: my email address is munged; You can never browse enough
Not coming to you from Glastonbury
···@wildcard.this.email.address.intentionally.left.crap.demon.co.uk (Martin Rodgers) writes:
>
> Apparently, Mrs Thatcher said, "You can't buck the market". While I'd
> disagree with most of the things she said, I don't deny that some people
> may still believe it. All I'm saying is that we might use the market. Let
> it choose better software.
I wonder if this isn't setting up the straw man of the metaphysical
market---the `invisible hand' as a real entity.
The point of the free market is to maximize value. Value is DEFINED
as the sum total of the PERCEIVED value of the things under
consideration. There's really no other way to measure value in an
objective way. If someone is willing to pay a certain amount (pay
supposedly relates to labor in some sense) for something then that
thing has that value in the market.
The reason I'm bringing this up is to point out that there's a strong
element of perception in the market. The free market cannot make
moral prescriptions or determine what a value `ought' to be.
Free-marketeers from Adam Smith to George Guilder have pointed this
out.
For example, take prostitution. Some of us would argue that
prostitution has a negative value to society as a whole. But the free
market is unable to make this judgment. The only way the value of
prostitution can change is if the perception people have of its value
changes.
The whole point of advertising is to affect that perception; i.e.
tobacco companies are able to convince children that a poisonous
substance that will negatively affect one's health and eventually
often kill one is a valuable commodity by advertising. The
combination of parental use and government restriction of its use by
minors also gives it the aura of grown-upness, at a time when the
appearance of such is important to these children. Thus something
that many would argue has a negative value in an absolute sense takes
on a positive value in the perception of these children.
Similarly, some of us argue that Microsoft software has less value
than its competitors. But as long as people think a word processor
that lets word-processing documents erase your hard disk is a better
value than its competitors, the market will reflect that by rewarding
Microsoft.
The market, then, is limited in the value judgments it can make.
Nevertheless the market is the most democratic way to assign value to
a system. As such it rewards the apportionment of resources to things
people want and discourages its apportionment to things people don't
want. By railing against the market we rail against ourselves, or
rather, we rail against people whose judgment we consider inferior to
our own. All arguments against the free market are, at heart, elitist
in the (non-pejorative) sense that they are arguments that some
minority knows better than the majority. The minority may be right;
it is still a minority.
Whenever the government intervenes in the marketplace, it by
definition lessens the value of the system (otherwise people would
have voluntarily done what the government is coercing them to do).
This is not to say that the government shouldn't do that; the same
principle applies as when it puts constraints on other elements of our
nature, such as the element that leads some of us to murder our fellow
citizens. Nevertheless whenever some entity overrides the judgment of
the market, it 1) lessens the value of the system and 2) itself makes
a value judgment that can only be either prophetic or metaphysical
(i.e. it is a prediction about some future evil that will occur or a
statement about the way the universe works in the absolute sense).
So as lisp advocates our job must be to persuade (educate) and produce
valuable things (i.e. things people want). Education changes
perception, and supply can create demand. The world-wide-web is a
clear example of the latter phenomenon; once it existed everyone had
to have a piece of it. Even though the Internet was around for years
before the world-wide-web, for some reason FTP wasn't nearly as
attractive as Netscape.
One can argue over the means; i.e. should we adopt dishonest means
like Microsoft's use of the word `innovate' which basically means
`force users to upgrade and to use Microsoft products', or should we
be constrained by some moral vision that encourages us to accurately
present our case.
O.K. Here's some Microsoft-style propaganda for lisp. (Though I tried
to make sure that all were in some sense true or at least not too
outrageously false.)
o Lisp cuts development time down by at least 100% compared to C.
Lisp's simple, regular syntax makes learning a breeze. <Did you
know that *++*_+++=++*++*__++; compiles under C (with proper
declarations, of course)? Beware of line noise!>
o <some> Lisp programs are as fast as C programs. Lisp allows an
incremental development approach that encourages rapid development
of a working prototype, then selective optimization of `hot-spots'
in the code. <Don't forget the quotes around `hot-spots';
non-technical people need them to know when jargon is being used.>
o Lisp encourages safe programming. Code is safe by default but
safety can be traded for speed. Lisp's innovative :-) optimization
parameters allow the programmer fine-grained control over exactly
how and where the tradeoff will be made.
o Lisp makes memory management problems a thing of the past. It
applies advanced AI techniques to efficiently decide when to obtain
memory and when to recycle unused memory.
o Lisp has the most advanced object system in existence---the ONLY
standardized object system available <at least it was last time I
checked....>
o Lisp has an integrated development environment that is more
advanced than anything else on the market <unfortunately the company
that sells it is bankrupt>.
o Lisp programs are portable to all platforms on the market and
becoming more so every day.
o Lisp is the outcome of more than thirty years of software evolution
and development; its foundations are in mathematical logic and its
implementation is the product of years of experience running the
most demanding applications in existence. <AI programming: the art
of making exponential problems less exponential.>
o Lisp has integral support for the World Wide Web. Lisp-based Web
Servers bring the full power of Lisp to your real-time
HTML-generation applications.
Please feel free to use the above. Copyright aban^H^H^H^Hdisavowed.
--
Fred Gilham gilham @ csl . sri . com
King Christ, this world is all aleak, / And life preservers there are none,
And waves that only He may walk / Who dared to call Himself a man.
-- e. e. cummings, from Jehovah Buried, Satan Dead
In article <··············@snapdragon.csl.sri.com>,
······@snapdragon.csl.sri.com says...
> O.K. Here's some Microsoft-style propaganda for lisp. (Though I tried
> to make sure that all were in some sense true or at least not too
> outrageously false.)
This is, IMHO, one way to defeat MS. Use their own methods of promotion.
If corporations listen to a certain kind of language, then we might also
use that language to promote Lisp. In fact, I expect that many Lisp
programmers already make the same points that you do to support their use
of Lisp. I know I do.
In other words, Know Thy Enemy.
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
Martin Rodgers wrote:
>
> In article <··············@snapdragon.csl.sri.com>,
> ······@snapdragon.csl.sri.com says...
>
> > O.K. Here's some Microsoft-style propaganda for lisp. (Though I tried
> > to make sure that all were in some sense true or at least not too
> > outrageously false.)
>
> This is, IMHO, one way to defeat MS. Use their own methods of promotion.
> If corporations listen to a certain kind of language, then we might also
> use that language to promote Lisp. In fact, I expect that many Lisp
> programmers already make the same points that you do to support their use
> of Lisp. I know I do.
>
IMHO, the problem with Lisp isn't Lisp or Microsoft, but the screwy
pre-ANSI C code in which it is written. It takes a MASSIVE effort
(time and money) to port this stuff so that it works on current
technology
(be it PC, HP or whatever). Companies which are able to invest resources
in
it have to charge huge prices for their ports, there-by reducing
interest with new
programmers who may actually benefit from it.
FF
> In other words, Know Thy Enemy.
> --
> Please note: my email address is munged; You can never browse enough
> "Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
In article <·············@foo.com>, ·······@foo.com says...
> IMHO, the problem with Lisp isn't Lisp or Microsoft, but the screwy
> pre-ANSI C code in which it is written. It takes a MASSIVE effort
> (time and money) to port this stuff so that it works on current
> technology
> (be it PC, HP or whatever).
I'm not sure what you mean. Lisp is already available for all the popular
platforms. If you have a not so popular platform that doesn't yet have an
indudustrial strength Lisp, you could have a problem. The same is true
for a lot of other software, too.
> Companies which are able to invest resources in
> it have to charge huge prices for their ports, there-by reducing
> interest with new
> programmers who may actually benefit from it.
There's an obvious and often used counter argument to this. The true
value of the software is what you do with it. Counting the cost alone
tells us nothing.
This applies to any software, ported or otherwise. If you can't afford
it, you probably don't need it badly enough. This is as true for apps
like Photoshop (which has, as it happens, been ported) as it is for Lisp.
BTW, I believe that ACL/PC was written for Windows, not ported to it,
and that Golden Common Lisp was ported to Windows from DOS. Both were,
last time I checked, considerably more expensive than LispWorks for
Windows, which was ported from Unix.
Counting money is fine, so long as you count _all of it_.
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
Martin Rodgers wrote:
>
> In article <·············@foo.com>, ·······@foo.com says...
>
> > IMHO, the problem with Lisp isn't Lisp or Microsoft, but the screwy
> > pre-ANSI C code in which it is written. It takes a MASSIVE effort
> > (time and money) to port this stuff so that it works on current
> > technology
> > (be it PC, HP or whatever).
>
> I'm not sure what you mean. Lisp is already available for all the popular
> platforms. If you have a not so popular platform that doesn't yet have an
> indudustrial strength Lisp, you could have a problem. The same is true
> for a lot of other software, too.
>
> > Companies which are able to invest resources in
> > it have to charge huge prices for their ports, there-by reducing
> > interest with new
> > programmers who may actually benefit from it.
>
> There's an obvious and often used counter argument to this. The true
> value of the software is what you do with it. Counting the cost alone
> tells us nothing.
>
> This applies to any software, ported or otherwise. If you can't afford
> it, you probably don't need it badly enough. This is as true for apps
> like Photoshop (which has, as it happens, been ported) as it is for Lisp.
>
> BTW, I believe that ACL/PC was written for Windows, not ported to it,
> and that Golden Common Lisp was ported to Windows from DOS. Both were,
> last time I checked, considerably more expensive than LispWorks for
> Windows, which was ported from Unix.
Will it compile MAXIMA?
>
> Counting money is fine, so long as you count _all of it_.
Show me a Lisp for NT (or 95) which will compile and run MAXIMA. And
Go32
under a DOS reboot doesn't count.
Have you ever looked at the code for AKCL or GCL and tried to get it
to compile under, say Cygnus?
I guess I am just complaining that the GNU Lisps out there? This has got
to be
the most non-portable code I have ever seen. There are so many OS
dependent
#ifdefs in that code it is impossible to port it to Windows without a
huge
investment in resources (which probably explains why commercial Lisp
charge $1000 for their compilers).
Plain and simple, I want to run MAXIMA under NT, and even if I DID shell
out
$1000 out of my pocket, how would I know that it would even compile
MAXIMA?
MAXIMA supposedly compiles under GCL and AKCL, but every attempt I have
mae
to run it on an HP9000 series machine has resulted in horrible failures
due
to some bad syntax by this software to run 'ld'.
Yeah, yeah, I know. Free software. You get what you pay for. But it's a
shame because
MAXIMA is a good resource, and it is impossible to integrate it with
current technology
based on my experience over the last few months.
I guess I should let the "Big Buxs" folks do this for me, and then
charge me $15,000
so I can use it on my home PC. But if I DO ever port this thing to run
under NT,
I guarantee it will be distrubuted for free over the visa-vis the GNU
public
license (not MAXIMA, but a port of either KCL, ACKL or GCL which will
actually
compile MAXIMA).
FF
> --
> Please note: my email address is munged; You can never browse enough
> "Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
Firefly <·······@foo.com> writes:
> Will it compile MAXIMA?
With a *little* work, I guess. I can't imagine anything *really*
problematic in the MAXIMA code.
> investment in resources (which probably explains why commercial Lisp
> charge $1000 for their compilers).
Actually, LWW is cheaper. Last time I looked it was at $495.
$700 including "CLIM".
In article <·············@foo.com>, ·······@foo.com says...
> Will it compile MAXIMA?
I've no idea. Sorry.
> > Counting money is fine, so long as you count _all of it_.
>
> Show me a Lisp for NT (or 95) which will compile and run MAXIMA. And
> Go32
> under a DOS reboot doesn't count.
I'll let somebody with an interest in math software try that. If you're
making a more general point, like the inability of certains Lisps to run
significant Lisp apps, I can't comment on that. If you were asking about
CL-HTTP, others here will be better qualified to comment, but it looks
like it will work with ACL/PC and LWW. The real test for a web server is
running a real website.
You see, I have more interest in Internet software than math apps. I hope
someone else here has more experience with MAXIMA and/or the Lisps that
you'd like to use.
> Have you ever looked at the code for AKCL or GCL and tried to get it
> to compile under, say Cygnus?
No, I've never looked at MAXIMA, nor am I ever likely to. I can only tell
you what kind of Lisps are available. If your only interest in Lisp is
for running existing apps like MAXIMA, then this may well constrain your
choice of Lisp. Pick one that runs the software, ignore the rest. The OS
vendor issues are less significant than the _Lisp vendor_ issues.
If you'd like to argument that LispWorks for Unix is better than
LispWorks for Windows, then I can't disput that, as I've never used the
Unix version. I have no interest, financial or otherwise, in such issues.
I've only briefly ran GCL under Linux. I didn't have time to do anything
meaningful with it, so this is a Lisp with which I have practically no
experience. This may someday change, but it won't be soon. I've no idea
what kind of OS dependancies GCL might have, but Cygnus have done some
wonderful work. AFAIK, it may be easy to do, or at least possible.
> I guess I am just complaining that the GNU Lisps out there? This has got
> to be
> the most non-portable code I have ever seen. There are so many OS
> dependent
> #ifdefs in that code it is impossible to port it to Windows without a
> huge
> investment in resources (which probably explains why commercial Lisp
> charge $1000 for their compilers).
"There is no such thing as portable software, only software
that has been ported." -- Jim Gettys
You can even find Lisp vendor dependant #+ and #- expressions in Lisp
source code. I've been aware of this kind of problem for almost as long
as I've been programming. Every implementation of a language can have
this problem because each implementation uses different code. So, we can
argue that each implementation defines a subtly different language to
another implementation.
It's a pain, but an unavoidable one. We can limit the effects to as small
a number of lines as possible. Of course it takes effort. That takes
time, however small it may be.
> Plain and simple, I want to run MAXIMA under NT, and even if I DID shell
> out
> $1000 out of my pocket, how would I know that it would even compile
> MAXIMA?
Try it with an evaluation version of the vendor's Lisp, or explain to the
vendor why you wish to purchase their product and ask them to help you
get the code running in their Lisp.
> MAXIMA supposedly compiles under GCL and AKCL, but every attempt I have
> mae
> to run it on an HP9000 series machine has resulted in horrible failures
> due
> to some bad syntax by this software to run 'ld'.
Perhaps Jim Gettys was right.
> Yeah, yeah, I know. Free software. You get what you pay for. But it's a
> shame because
> MAXIMA is a good resource, and it is impossible to integrate it with
> current technology
> based on my experience over the last few months.
What is MAXIMA worth to you? How much time? Your comments above suggest
that if a Lisp can compile and run this app, then you're willing to spent
$1000. You must also ask yourself how much time you're willing to spend
on this software, and how much that time costs you.
> I guess I should let the "Big Buxs" folks do this for me, and then
> charge me $15,000
> so I can use it on my home PC. But if I DO ever port this thing to run
> under NT,
> I guarantee it will be distrubuted for free over the visa-vis the GNU
> public
> license (not MAXIMA, but a port of either KCL, ACKL or GCL which will
> actually
> compile MAXIMA).
Another question worth asking is: How much time are you willing to freely
donate to this port? Since we all benefit from this, you could ask for
others to help you. You're not alone!
Good luck.
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
Firefly <·······@foo.com> writes:
> Show me a Lisp for NT (or 95) which will compile and run MAXIMA. And
> Go32
> under a DOS reboot doesn't count.
CLISP runs on NT and '95. There was a version of maxima (punimax)
that used to compile out-of-the-box for CLISP.
> Plain and simple, I want to run MAXIMA under NT, and even if I DID shell
> out
> $1000 out of my pocket, how would I know that it would even compile
> MAXIMA?
If you have $1000 to run maxima, why not get the real Macsyma?
Ray
Raymond Toy <···@rtp.ericsson.se> writes:
> Firefly <·······@foo.com> writes:
>
> > Plain and simple, I want to run MAXIMA under NT, and even if I DID shell
> > out
> > $1000 out of my pocket, how would I know that it would even compile
> > MAXIMA?
>
> If you have $1000 to run maxima, why not get the real Macsyma?
>
Yeah! It is incredibly cheaper than Mathematica and/or Maple.
<http://www.macsyma.com> advertises 349 USD for the Pro version.
--
Marco Antoniotti ===========================================
PARADES, Via San Pantaleo 66, I-00186 Rome, ITALY
tel. +39 - (0)6 - 68 80 79 23, fax. +39 - (0)6 - 68 80 79 26
http://www.parades.rm.cnr.it
Good point http://www.macsyma.com/priceuc.html
Has great GUI and Help system. You can if you like, drop into lisp from
Macsyma and back again.
--
········@one.net
Raymond Toy wrote in message <··············@rtp.ericsson.se>...
>Firefly <·······@foo.com> writes:
>
>
>> Show me a Lisp for NT (or 95) which will compile and run MAXIMA. And
>> Go32
>> under a DOS reboot doesn't count.
>
>CLISP runs on NT and '95. There was a version of maxima (punimax)
>that used to compile out-of-the-box for CLISP.
>
>> Plain and simple, I want to run MAXIMA under NT, and even if I DID shell
>> out
>> $1000 out of my pocket, how would I know that it would even compile
>> MAXIMA?
>
>If you have $1000 to run maxima, why not get the real Macsyma?
>
>Ray
Raymond Toy wrote:
>
>
> > Plain and simple, I want to run MAXIMA under NT, and even if I DID shell
> > out
> > $1000 out of my pocket, how would I know that it would even compile
> > MAXIMA?
I said 'If I did', not that I would :)
>
> If you have $1000 to run maxima, why not get the real Macsyma?
>
> Ray
If you charged $100 maybe I would. Let me ask you this, if you charge
$100 for a piece of software and sell 100,000 copies or you charge $1000
for a piece of software and sell 1,000 copies, which makes you more
money?
FF
Firefly <·······@foo.com> writes:
> Raymond Toy wrote:
> >
>
> >
> > > Plain and simple, I want to run MAXIMA under NT, and even if I DID shell
> > > out
> > > $1000 out of my pocket, how would I know that it would even compile
> > > MAXIMA?
>
> I said 'If I did', not that I would :)
But you didn't say you wouldn't, either. :-)
But I did point out CLISP runs on NT and punimax does (did?) run on
CLISP, so you can have maxima on NT.
>
> >
> > If you have $1000 to run maxima, why not get the real Macsyma?
> >
> > Ray
>
> If you charged $100 maybe I would. Let me ask you this, if you charge
> $100 for a piece of software and sell 100,000 copies or you charge $1000
> for a piece of software and sell 1,000 copies, which makes you more
> money?
There's no way to know. That $100 software may cost $99 to make; but
the $1000 may cost $999 to make.
Ray
Fred Gilham wrote:
>
> o Lisp cuts development time down by at least 100% compared to C.
Development time with Lisp is zero or negative??? Could this just possibly be a
slight exaggeration??? :-)
(Seriously though -- what were you trying to say?)
-- Scott
* * * * *
To use the email address, remove all occurrences of the letter "q".
"Scott L. Burson" <·····@zeta-sqoft.com> writes:
> Fred Gilham wrote:
> >
> > o Lisp cuts development time down by at least 100% compared to C.
>
> Development time with Lisp is zero or negative??? Could this just possibly be a
> slight exaggeration??? :-)
>
> (Seriously though -- what were you trying to say?)
>
Well, I DID say this was `Microsoft-style propaganda'. :-)
--
Fred Gilham gilham @ csl . sri . com
King Christ, this world is all aleak, / And life preservers there are none,
And waves that only He may walk / Who dared to call Himself a man.
-- e. e. cummings, from Jehovah Buried, Satan Dead
<<So as lisp advocates our job must be to persuade (educate) and
produce valuable things (i.e. things people want).>>
What do people want that can be programmed in LISP? I wrote a LISP
program that teaches pre-school children how to read, but I can't find
anybody interested in trying it for free much less buying it.
<<o Lisp cuts development time down by at least 100% compared to C.>>
That's impossible. If you cut something down 100%, the result is zero.
Not even the most ardent LISP advocates would honestly say that it
takes zero time to develop a new application in LISP.
The rest of your 'advertising' claims for LISP looked valid.
In article <················@naggum.no>, Erik Naggum wrote:
> the world changes with every significant action you make. if you feel
> the world doesn't change when you act, either your actions are truly
> insignificant, or you're just depressed. now you can go hang yourself or
> you can go do something significant. my suggestion if you want to keep
> whining and bemoaning the state of the world is to put yourself out of
> your misery as soon as possible (please let me try to sell you the rope),
> but you _could_ also try getting out of your depression and one good way
> is to start noticing what software people are actually buying. (do you
> recognize that question?) every single computer in the world that sports
> a Common Lisp environment is important to its programmer and its vendor
> (or author); why isn't it to you? every single sale of a Linux CD-ROM is
> important to the customer and the Free Software world; why isn't it to
> you? every single transfer of the Association of Lisp Users web page to
> somebody browser is important to the Lisp community; why isn't it to you?
I hate to do the sophisticated version of a lame 'me too' but, Erik,
I simply couldn't agree more.
The only thing I can really control is my own desire and my own action.
Every time I choose to use Linux or Lisp or any other superior product, I
not only gain the benefit of using that superior product, but I deprive
the champions of mediocrity of another ally.
It all adds up. Slowly but surely, it all adds up.
Best,
Kendall Grant Clark
--
One day the day will come when the day will not come.
In article <·····················@superfly.ntlug.org>, ······@cmpu.net
says...
> It all adds up. Slowly but surely, it all adds up.
Does anyone really think I'm denying with this? I'm not. The hard sell
isn't necessary here; we're preaching to the converted. Contrast this
thread with the similar thread current in comp.lang.functional.
Consider the words of Kent Pitman, not too long ago:
http://www.wildcard.demon.co.uk/lisp/notawhale.html
Now, if we were the only people offering a "magic bullet", a hard sell
might work on those pointy haired managers. Alas, we're not alone. We
understand what makes Lisp special, but there are people in this world
who only recognise a thing exists when you can put a $ on it.
Hence the phrase, "Money talks". When Jan 1 rolls by, I want to know how
many billions of $ were lost per hour. That's a hard sell that the pointy
haired managers will understand.
BTW, I'm using Linux. I like it. I don't need the hard sell treatment!
In fact, that will only make me suspicious. Loads of people give me the
hard sell treatment, and many of them are people I wouldn't trust to sit
on a toilet seat the right way. I live in a country in which people have
died because of a religious conflict, so please forgive me if I'm just a
_little_ bit cynical - and waring of anyone with a political agenda.
By all means put MS out of business, but please don't tell me that'll
make the world a wonderful place. People will still fight over which end
of a boiled egg you eat first. People will still die because this BS.
If you really want to change the world, follow the money. Otherwise,
you're just waving a placard. Have you seen how the establishment look at
people with placards? Put a $ on your placard and make a real difference.
Still, this is only an opinion. I know people have died because of such
things, and as far as I know a flamewar is non-lethal. So you won't shut
me up by flaming me. I'm not attacking Lisp, nor am I supporting MS! I'm
just delivering an unpleasant message. If it were _my_ msg, then at least
the flames might be somewhat justified. However, it's not.
Look around you. Read Dilbert. Look at the world. Follow the money.
--
Please note: my email address is munged; You can never browse enough
Hit the advocates three times over the head with the Linux Advocacy
mini-HOWTO <URL:http://sunsite.unc.edu/LDP/HOWTO/mini/Advocacy.html>
>In article <·····················@superfly.ntlug.org>, ······@cmpu.net
>says...
>
>> It all adds up. Slowly but surely, it all adds up.
Martin Rodgers, responded:
>Does anyone really think I'm denying with this? I'm not. The hard sell
>isn't necessary here; we're preaching to the converted. Contrast this
>thread with the similar thread current in comp.lang.functional.
Actually, no, I don't think you are denying this; but if you'll read a little
more carefully, you'll notice that I was simply affirming Erik's point, and
I even took the extra precaution of replying to Erik's post, not to anyone
else's.
I don't know or care much what you deny or don't deny. I happen to agree
with Erik about the exercise of personal responsibility even, or, perhaps,
especially, in technology decisions.
Is that really hard to understand?
Best,
Kendall G. Clark
PS -- Perhaps you weren't as careful in replying to my message or in reading
it, but, just for the record, I'm not selling anything, nor am I preaching,
either to the converted or the damned.
I was just agreeing with something I read on Usenet. I know, it's an unusual
thing to do, what can I say?
In article <·····················@superfly.ntlug.org>, ······@cmpu.net
says...
> Is that really hard to understand?
I understand you, and I agree. I'm just pointing out that many others
will find it harder to understand, and some may even disagree. Now,
Erik calls this whining. I refer to a post by Kent Pitman, and ask if
that too is whining.
> PS -- Perhaps you weren't as careful in replying to my message or in reading
> it, but, just for the record, I'm not selling anything, nor am I preaching,
> either to the converted or the damned.
Ditto. I'm not a sales person. I don't choose Lisp to make more money,
nor any other tool. I choose tools based on how they can help me. If
anyone asks me about them, I'll answer. I don't preach.
There's too much advocacy in this world. I've been on the receiving end
of a lot of it. For as long as I've used computers, people have made
assumptions about the hardware and software I use, as if I choose all
these things myself.
I didn't choose the CPU in my first machine, as I didn't even know what a
CPU was. One day a bloke tried to convince me that the CPU in his machine
was superior to the CPU in mine. I suspect that he didn't know which CPU
that was when he chose his machine. We settled the issue by letting the
two machines play chess. It was a draw. ;)
Every time I encounter advocacy it's the same, except that there isn't
such a simple and well mannered way of resolving the conflict. I wish
there was, but it is first necessary to realise that everything that one
side believes makes them right may convince the other side that _they_
are right. When the feelings are strong enough, rational arguments are
thrown away and the mud slinging begins.
> I was just agreeing with something I read on Usenet. I know, it's an unusual
> thing to do, what can I say?
I'm just pointing out an ugly truth; that your milage may vary. I don't
need to take sides to do this. I prefer _not_ to take sides in arguments
expressing hate. By all means point out faults, but if you can't win by
being scrupulously honest, as is the case with many advocates here on
UseNet and the web, then I wish to distance myself.
Perhaps I'm just too honest for my own good. I see the flaws in arguments
on both sides, and when I point them out, I get flamed - by both sides.
I don't claim everything they say is wrong. In fact, I usually agree with
most things they say. However, when somebody says, "This can't be done",
and I know damn well it can, because I'm doing it, or I know others who
are, then why should I not say so? Well, I get flamed for using Lisp by
C++ advocates, and I get flamed for not using Lisp enough by people here.
It's a lose lose situation, but I can live with it.
It's same with operating systems. I use NT _and_ Linux, but some people
are not satisfied with this. Well, I learned long ago that you can never
please everyone, and more recently I see that some people will never
accept what I do. I might only win by doing nothing, and that is no win
at all. What did Richard Feynman used to say? "What do _you_ care what
other people think?"
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
In article <··························@news.demon.co.uk>,
···@wildcard.this.email.address.intentionally.left.crap.demon.co.uk (Martin
Rodgers) wrote:
>What did Richard Feynman used to say? "What do _you_ care what
>other people think?"
I feel compelled to point out that it was actually Arlene Feynman,
Richard's first wife, who said this--to Richard, who at that moment was
plagued by an admittedly uncharacteristic bout of self-doubt with respect
to what someone else thought of him.
Ob. Lisp observation: no one in this group should flame anyone else for
"not using Lisp *enough*." Given the loopy state of what passes for the
"real world," we should be delighted than anyone is using Lisp at all.
>Please note: my email address is munged; You can never browse enough
>"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
Paul
In article <·································@news.mci2000.com>,
·····@mcione.com says...
> I feel compelled to point out that it was actually Arlene Feynman,
> Richard's first wife, who said this--to Richard, who at that moment was
> plagued by an admittedly uncharacteristic bout of self-doubt with respect
> to what someone else thought of him.
I know, I just didn't want to go into that much detail. ;) Thanks.
> Ob. Lisp observation: no one in this group should flame anyone else for
> "not using Lisp *enough*." Given the loopy state of what passes for the
> "real world," we should be delighted than anyone is using Lisp at all.
Thanks.
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
Martin Rodgers wrote:
> ... I've not
> bought _any_ MS software, other than what came bundled with one machine.
> Nor will I willingly buy any. Ever. Small but significant.
You could make a point by buying hardware that does not come bundled
with MS software (difficult in the PC world, but possible).
> ...
Btw, this is leading too far away from the subject of this NG. Any
suggestions where to redirect it to (other than alt.null)?
--
Hartmann Schaffer
Guelph, Ontario, Canada
········@netcom.ca (hs)
In article <·················@netcom.ca>, ········@netcom.ca says...
> You could make a point by buying hardware that does not come bundled
> with MS software (difficult in the PC world, but possible).
This is exactly what I did, just a few months ago.
> Btw, this is leading too far away from the subject of this NG. Any
> suggestions where to redirect it to (other than alt.null)?
I made that point a few days ago. I set followups to more appropriate
newsgroups, which Erik ignored. In fact, he flamed me for doing it.
Ho hum.
--
Please note: my email address is munged; You can never browse enough
Not coming to you from Glastonbury
On 27 Jun 1998 19:23:06 +0000, Erik Naggum <······@naggum.no> posted:
>* Martin Rodgers
>| How do we "vote" for better business ethics without sounding like
>| socialists?
>
> you buy from somebody else. just how hard can it be to grasp this?
It may not be that simple.
US politics has often been accused of providing two parties that are
*both* undeserving of votes. I don't get to vote in US elections; this
is not an unhappy situation as I can't really be supportive of either
of the significant parties.
The same can be true in business when power in the marketplace
concentrates into relatively few hands. I can despise Microsoft; that
does not necessarily leave me with alternatives when the "witless cattle"
that decide to market products decide that M$ is the only platform
they'll support.
--
Those who do not understand Unix are condemned to reinvent it, poorly.
-- Henry Spencer <http://www.hex.net/~cbbrowne/lsf.html>
········@hex.net - "What have you contributed to Linux today?..."
* Erik Naggum
| you buy from somebody else. just how hard can it be to grasp this?
* Christopher B. Browne
| It may not be that simple.
well, yes, it always is that simple. it's called "acting on principle".
| The same can be true in business when power in the marketplace
| concentrates into relatively few hands. I can despise Microsoft; that
| does not necessarily leave me with alternatives when the "witless cattle"
| that decide to market products decide that M$ is the only platform
| they'll support.
have you really _needed_ a piece of software that runs only on Microsoft?
I have never seen _anything_ published solely in the Windows market that
had me go "wow! that software really _would_ make my life easier! how I
wish I had a PC with Windows!". supposing I already had invested in a PC
with Windows, what would I _want_ to buy for it, except various idiot
tools to help me survive the many inherent limitations of the system and
make it more like a real computer, which by the time I had got all the
crappy add-ons, I could get for less money? never has any Microsoft
victim been able to show me anything they do on their PC's that I cannot
do at least as well on my SPARCstation, and usually faster even the first
time -- of tasks that _matter_, that is. entertainment is something else.
if you are into selling cheap goods that nobody needs to millions of
bored people who crave entertainment, choose Microsoft and make lots of
money. if you think mass marketing is a serious liability, like I do
("if it's acceptable to the large fraction of the population that base
their buying decisions on TV commercials, how could it possibly be good
enough for me?"), avoid them entirely, because the Microsoft world has
_nothing_ to offer anyone except ubiquity at this time.
#:Erik
--
http://www.naggum.no/spam.html is about my spam protection scheme and how
to guarantee that you reach me. in brief: if you reply to a news article
of mine, be sure to include an In-Reply-To or References header with the
message-ID of that message in it. otherwise, you need to read that page.
In article <················@naggum.no>, ······@naggum.no says...
> have you really _needed_ a piece of software that runs only on Microsoft?
When a client insists on it, yes. The alternative is find new clients.
Are there enough clients who don't use MS software for every software
house? Again, this is money talking. We go where the money is. It may be
a little hard to understand, but not everyone has the freedom to choose
who to work for. It's very arrogant to suggest otherwise.
That's why people call it "market pressure". Yep, it's pressure. If you
can resist it, then you're either very lucky, or you have a secret that
is worth a _lot of money_. I say that because so few people believe that
you can, in the words of Mrs Thatcher, "buck the market". That should be
a good enough reason to try it; anything that proves her wrong is fine
with me. I support bucking the market, as I suspect you also do.
This is why I've been saying for years: share the secret. Instead of
insulting people for not using Lisp, we should lead them to Lisp. Instead
of telling them them Lisp is, let them discover it for themselves. That
way, their egos will tell them how wonderful _they_ are for making the
discovery. After all, we don't need the credit ourselves.
The hard sell technique is perfect for people who want the credit for
themselves. MS do this. They're masters of the hard sell. The soft sell
gives a bigger win, because the buyer invests more than money. They
invest their ego. In a few years from now, MS will lose big time, while
we will win merely by saying, "I told you so", and reminding everyone
what we've been saying for years: you can win big with Lisp.
--
Please note: my email address is munged; You can never browse enough
Not coming to you from Glastonbury
In article <·················@news.deltanet.com>, ····@deltanet.com (Zeno)
wrote:
[snip]
>I have never seen a text on business programming with Lisp, or, in
>fact any other reference to it in other than AI texts and books about
>early programming languages.
See <http://www.infinet.com/~btobin/lispteam.html> for some pointers to
use of Lisp in business programming.
[snip]
--
David B. Lamkins <http://www.teleport.com/~dlamkins/>
In article <·················@news.deltanet.com>, ····@deltanet.com
says...
> To me, these were logical questions.
Indeed they are, for anyone who is not already familiar with a language.
It's the same with any tool. Curiously, some people feel that what they
use (iterate for all tools on the planet) is somehow special.
Years ago I learned of "egoless programming", and it changes the way I
looked at every tool I use. I look at people around me, and see that many
of them can't make the distinction between a tool and, um, their ego.
> Actually, if you have let people think that the suggestions you have
> implanted are their own ideas so they will act on them, then you
> really can't say, "I told you so," without blowing your cover. The
> real way you would win is by making Lisp a more popular language, but
> I honestly have no idea if the Lisp community wants that.
Not really. The "I told you so" applies to crap software, for example
code that is dependant on a two digit date format. My favourites are code
that doesn't check for buffer overflows and code that doesn't check the
status returned from a function. You can find both of these very often in
C code. State dependancies are also frequent. I learned how to avoid
these pitfalls soon after discovering C, but it looks like programmers
with years of C experience still don't get it.
In my experience, it looks like most C programmers are "cowboys". Perhaps
that's a UK expression that doesn't travel well, I don't know. It usually
refers to people in the building trade who use poor building materials
and don't finish the job properly. In some cases, like with an electrics,
this can potentially be lethal.
So, in a few years from now, I may be saying to someone, "Dear oh dear,
who did this for you?", in that tone of derision that says, "You made a
right cock-up of this one!" I shall have to restrain my glee. ;)
--
Please note: my email address is munged; You can never browse enough
Not coming to you from Glastonbury
(nothing to do with the rest of this debate, just slang clarification)
On Mon, 29 Jun 1998, Martin Rodgers wrote:
[]
> In my experience, it looks like most C programmers are "cowboys". Perhaps
> that's a UK expression that doesn't travel well, I don't know. It usually
> refers to people in the building trade who use poor building materials
> and don't finish the job properly. In some cases, like with an electrics,
> this can potentially be lethal.
Well, sorta. "Cowboy coders" has more to do with portability (ignoring it)
and using compiler specific hacks that exploit a specific compiler's
handling of something the C standard specifies as "undefined". The
portability issue is something like:
/* store and retrieve an 8-bit value from the second byte of a 4-byte
unsigned int that is only 4 bytes to begin with for address alignment,
where the values it was originally specified for never fall outside
an 8-bit range. This saves using another variable for the flag
value stored in the second byte. */
#define UPSHIFT8(v) (v << 8)
#define DOWNSHIFT8(v) (v >> 8)
/* assign 2nd byte value; semicolon in caller */
#define PUT_2ND(a, v) a &= 0xFFFF00FF; a |= UPSHIFT8(v)
/* extract 2nd byte value; semicolon in caller */
#define GET_2ND(a, b) DOWNSHIFT8(a & 0x0000FF00)
This is only correct on little-endian cpus, and assumes 8-bit bytes.
A portable implementation would use BITS_PER_BYTE instead of "8",
and have #ifdef LITTLE_ENDIAN (shift one way) #else (shift the other
way).
So this code probably worked find on the original little-endian platform,
but it amounts to porting overhead on big-endian platforms. It's broken in
the outer context of re-usable code, but may not be broken at all in the
use it was intended for. This is cowboy code.
Another example is using 0 as an address value to substitute for NULL.
A NULL pointer is an abstract data type. I don't know the actual C
standard wording on it, but glibc's info files say that assuming that
a NULL pointer is interchangeable with 0 is not portable. NULL just
resolves to "false" when tested in conditionals, without also implying
that it resolves to a physical value of 0. When assigned to an address
variable, NULL is a value all of its own. Address 0 is the conventional
way that C runtimes represent NULL internally, so code like this usually
works, but it isn't *guaranteed* to work with a C compiler asserted to
be ansi-standard. So this is cowboy code, too.
It's like using expensive paint on the seawall side and adjoining sides of
a house close to saltwater, and using cheap paint the same color on the
back side of it. The cheap paint is not the right paint for the local
environment, but the user of the house might never notice, because the
salt air that peels paint always blows in from the other side of the
house, and the local landscape just happens not to induce back-eddies
behind the house. That would be cowboy construction.
Regards, Clayton Weaver ······@eskimo.com (Seattle)
--
Regards, Clayton Weaver ······@eskimo.com (Seattle)
"Since this is a technical newsgroup and has mostly done a pretty good job
On Tue, 30 Jun 1998 12:46:17 -0700, Clayton Weaver <······@eskimo.com> wrote:
>Well, sorta. "Cowboy coders" has more to do with portability (ignoring it)
>and using compiler specific hacks that exploit a specific compiler's
>handling of something the C standard specifies as "undefined".
I agree with this.
>Another example is using 0 as an address value to substitute for NULL.
>A NULL pointer is an abstract data type. I don't know the actual C
>standard wording on it, but glibc's info files say that assuming that
>a NULL pointer is interchangeable with 0 is not portable.
Bad example. This is incorrect. NULL is always defined as 0.
Always. From the C FAQ:
=======================================================================
5.10: But wouldn't it be better to use NULL (rather than 0), in case
the value of NULL changes, perhaps on a machine with nonzero
internal null pointers?
A: No. (Using NULL may be preferable, but not for this reason.)
Although symbolic constants are often used in place of numbers
because the numbers might change, this is *not* the reason that
NULL is used in place of 0. Once again, the language guarantees
that source-code 0's (in pointer contexts) generate null
pointers. NULL is used only as a stylistic convention. See
questions 5.5 and 9.2.
=======================================================================
I still agree with your main point though. Most programmers
think if their code compiles then it must be perfect. Warnings?
Turn those nasty things off. :)
Neil
········@acs.ucalgary.ca (Neil Schemenauer) writes:
········@acs.ucalgary.ca (Neil Schemenauer) writes:
> Bad example. This is incorrect. NULL is always defined as 0.
> Always. From the C FAQ:
>
> =======================================================================
> 5.10: But wouldn't it be better to use NULL (rather than 0), in case
> the value of NULL changes, perhaps on a machine with nonzero
> internal null pointers?
>
> A: No. (Using NULL may be preferable, but not for this reason.)
> Although symbolic constants are often used in place of numbers
> because the numbers might change, this is *not* the reason that
> NULL is used in place of 0. Once again, the language guarantees
> that source-code 0's (in pointer contexts) generate null
> pointers. NULL is used only as a stylistic convention. See
> questions 5.5 and 9.2.
> =======================================================================
>
What is supposed to happen with this code, then? Assume we're on an
Alpha, with 64-bit pointers and 32-bit ints.
int
do_it(p, n)
int* p;
int n;
{
if (p == 0)
return n;
else
return p[n];
}
main()
{
do_it(0, 5);
}
I think this is broken code. The call to do_it should use NULL and
not 0.
ObLisp: I wouldn't have this problem in lisp
Ray
In article <··············@rtp.ericsson.se>,
Raymond Toy <···@rtp.ericsson.se> wrote:
>main()
>{
> do_it(0, 5);
>}
>
>I think this is broken code. The call to do_it should use NULL and
>not 0.
You're correct that it's broken. But changing it to NULL won't fix it. It
should use (int *)0 or (int *)NULL. This is because NULL is permitted to
expand either to 0 or (void *)0; in an implementation where NULL expands to
0, the code will still be broken.
Note that if there were a prototype for do_it(), the call as written would
be fine with 0 or NULL and with or without the cast, since the conversion
would be implicit.
--
Barry Margolin, ······@bbnplanet.com
GTE Internetworking, Powered by BBN, Cambridge, MA
*** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups.
From: Hartmann Schaffer
Subject: Re: Harlequin vs. Allegro ("cowboy C coders")
Date:
Message-ID: <359A576E.5B8B827D@netcom.ca>
Barry Margolin wrote:
>
> In article <··············@rtp.ericsson.se>,
> Raymond Toy <···@rtp.ericsson.se> wrote:
> >main()
> >{
> > do_it(0, 5);
> >}
> >
> >I think this is broken code. The call to do_it should use NULL and
> >not 0.
>
> You're correct that it's broken. But changing it to NULL won't fix it. It
> should use (int *)0 or (int *)NULL. This is because NULL is permitted to
> expand either to 0 or (void *)0; in an implementation where NULL expands to
> 0, the code will still be broken.
Well, typically you would use an include that has the implementation
specific definition of NULL that would be right. The only case where
there would be problems is on hadware where the pointer representation
would depend on the data type (I thing some Data General machines had
this problem)
--
Hartmann Schaffer
Guelph, Ontario, Canada
········@netcom.ca (hs)
Barry Margolin <······@bbnplanet.com> writes:
> In article <··············@rtp.ericsson.se>,
> Raymond Toy <···@rtp.ericsson.se> wrote:
> >main()
> >{
> > do_it(0, 5);
> >}
> >
> >I think this is broken code. The call to do_it should use NULL and
> >not 0.
>
> You're correct that it's broken. But changing it to NULL won't fix it. It
> should use (int *)0 or (int *)NULL. This is because NULL is permitted to
> expand either to 0 or (void *)0; in an implementation where NULL expands to
> 0, the code will still be broken.
I had assumed that, on machines where 0 and (void *)0 are different,
NULL would be defined correctly to mean a null pointer, which is what
NULL is supposed to be. I haven't done any heavy-duty C hacking in
ages, so my memory is probably fuzzy here.
>
> Note that if there were a prototype for do_it(), the call as written would
> be fine with 0 or NULL and with or without the cast, since the conversion
> would be implicit.
Yes, that's why I left out the prototype.
Ray
In article <··············@rtp.ericsson.se>,
Raymond Toy <···@rtp.ericsson.se> wrote:
>Barry Margolin <······@bbnplanet.com> writes:
>
>> In article <··············@rtp.ericsson.se>,
>> Raymond Toy <···@rtp.ericsson.se> wrote:
>> >main()
>> >{
>> > do_it(0, 5);
>> >}
>> >
>> >I think this is broken code. The call to do_it should use NULL and
>> >not 0.
>>
>> You're correct that it's broken. But changing it to NULL won't fix it. It
>> should use (int *)0 or (int *)NULL. This is because NULL is permitted to
>> expand either to 0 or (void *)0; in an implementation where NULL expands to
>> 0, the code will still be broken.
>
>I had assumed that, on machines where 0 and (void *)0 are different,
>NULL would be defined correctly to mean a null pointer, which is what
>NULL is supposed to be. I haven't done any heavy-duty C hacking in
>ages, so my memory is probably fuzzy here.
But (void *)0 could be different from (int *)(void *)0, so using NULL
rather than 0 doesn't make the program any more correct.
--
Barry Margolin, ······@bbnplanet.com
GTE Internetworking, Powered by BBN, Cambridge, MA
*** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups.
Barry Margolin <······@bbnplanet.com> writes:
> But (void *)0 could be different from (int *)(void *)0, so using NULL
> rather than 0 doesn't make the program any more correct.
Oops, I forgot about machines where pointers to different types are
themselves different. Used to use such a machine (Harris H800).
Ray
From: Hartmann Schaffer
Subject: Re: Harlequin vs. Allegro ("cowboy C coders")
Date:
Message-ID: <359A568A.33F43767@netcom.ca>
Raymond Toy wrote:
> ...
> What is supposed to happen with this code, then? Assume we're on an
> Alpha, with 64-bit pointers and 32-bit ints.
>
> int
> do_it(p, n)
> int* p;
> int n;
> { if (p == 0)
> return n;
> else
> return p[n]; }
>
> main() { do_it(0, 5); }
>
> I think this is broken code. The call to do_it should use NULL and
> not 0.
Yes, the code is broken, but for a different reason: the definition of
do_it should use the ANSI style rather than the K&R style. If you
rewrite it to
int do_it(int *p, int n)
it will work, because the compiler sees that the 0 is a pointer.
Otherwise the compiler does not have the information available that the
0 in the call is a pointer.
My understanding is that the old style function definition style was
left in the standard in order not to invalidate legacy code. In my
opinion, this clearly was a mistake.
--
Hartmann Schaffer
Guelph, Ontario, Canada
········@netcom.ca (hs)
Hartmann Schaffer <········@netcom.ca> writes:
> Yes, the code is broken, but for a different reason: the definition of
> do_it should use the ANSI style rather than the K&R style. If you
> rewrite it to
>
> int do_it(int *p, int n)
>
> it will work, because the compiler sees that the 0 is a pointer.
> Otherwise the compiler does not have the information available that the
> 0 in the call is a pointer.
ANSI C does not require prototypes (does it?), so the code as written
is correct for this reason.
>
> My understanding is that the old style function definition style was
> left in the standard in order not to invalidate legacy code. In my
> opinion, this clearly was a mistake.
Agreed. But I certainly wouldn't want to throw away every existing
piece of C code or have to convert it to ANSI. It was a reasonable
tradeoff.
Ray
Raymond Toy wrote in message <··············@rtp.ericsson.se>...
>What is supposed to happen with this code, then? Assume we're on an
>Alpha, with 64-bit pointers and 32-bit ints.
<snip old code example>
>I think this is broken code. The call to do_it should use NULL and
>not 0.
Yes, this is broken, but because you're not using C correctly. If you
merely redefine do_it as:
int do_it (int *p, int n)
{
if (p == 0)
return n;
else
return p[n];
}
It will work as you expect. The "old K&R" function heading style is
obsolete, and was only retained to keep from breaking old code.
You can expect ANSI-style headers to correctly interpret parameters
passed in (the 0 is guaranteed to be passed as a pointer type).
>ObLisp: I wouldn't have this problem in lisp
I wouldn't have this problem in C, either. :-)
----------------------------------------
····@charybdis.com
"Throwing fire at the sun"
Dan Higdon <····@charybdis.com> wrote:
> Raymond Toy wrote in message <··············@rtp.ericsson.se>...
> >What is supposed to happen with this code, then? Assume we're on an
> >Alpha, with 64-bit pointers and 32-bit ints.
> <snip old code example>
> >I think this is broken code. The call to do_it should use NULL and
> >not 0.
> Yes, this is broken, but because you're not using C correctly.
If he has to work with old code or an old compiler, this is
correct. Not everyone is using new code on visual C++, you know. :)
<snip>
> It will work as you expect. The "old K&R" function heading style is
> obsolete, and was only retained to keep from breaking old code.
Yes, I suppose if you never have to work with any old code, you're
fine. :)
dave
From: Hartmann Schaffer
Subject: Re: Harlequin vs. Allegro ("cowboy C coders")
Date:
Message-ID: <359A5479.6245416A@netcom.ca>
Neil Schemenauer wrote:
> ...
> I still agree with your main point though. Most programmers
> think if their code compiles then it must be perfect. Warnings?
> Turn those nasty things off. :)
Depends on the C compiler. I have seen compilers that put out warnings
for clearly defined stuff (e.g. assigning between (void *) and other
pointer types). Putting out warnings for something that is clearly
legal, well defined, and portable is a nuisance at best.
--
Hartmann Schaffer
Guelph, Ontario, Canada
········@netcom.ca (hs)
In article <·······································@eskimo.com>,
······@eskimo.com says...
> Well, sorta. "Cowboy coders" has more to do with portability (ignoring it)
> and using compiler specific hacks that exploit a specific compiler's
> handling of something the C standard specifies as "undefined". The
> portability issue is something like:
I know. ;) My homepage has a page just for links to pages describing
various for C problems, and a few solutions.
http://www.wildcard.demon.co.uk/c/problems.html
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
From: Hartmann Schaffer
Subject: Re: Harlequin vs. Allegro ("cowboy C coders")
Date:
Message-ID: <359A5380.16AA6AF@netcom.ca>
Clayton Weaver wrote:
> ...
> Another example is using 0 as an address value to substitute for NULL.
> A NULL pointer is an abstract data type. I don't know the actual C
> standard wording on it, but glibc's info files say that assuming that
> a NULL pointer is interchangeable with 0 is not portable. NULL just
> resolves to "false" when tested in conditionals, without also implying
> that it resolves to a physical value of 0. When assigned to an address
> variable, NULL is a value all of its own. Address 0 is the conventional
> way that C runtimes represent NULL internally, so code like this usually
> works, but it isn't *guaranteed* to work with a C compiler asserted to
> be ansi-standard. So this is cowboy code, too.
I don't have it handy right now, but (at least an older version of) the
C standard stated explicitely that assigning 0 (zero) to a pointer or
comparing a pointer to 0 (zero) was equivalent to using the NULL
pointer. I am pretty sure that pretty much every C implementation
inplements NULL in one of the header files as (void *)0. The
portability problems come onle when somebody passes 0 as an argument to
an unprototyped function (which should be taken out of the stanedard
anyway, regardless of legacy code, IMO)
> ...
--
Hartmann Schaffer
Guelph, Ontario, Canada
········@netcom.ca (hs)
From: Casper H.S. Dik - Network Security Engineer
Subject: Re: Harlequin vs. Allegro ("cowboy C coders")
Date:
Message-ID: <casper.899310436@nl-usenet.sun.com>
[[ PLEASE DON'T SEND ME EMAIL COPIES OF POSTINGS ]]
Hartmann Schaffer <········@netcom.ca> writes:
>I don't have it handy right now, but (at least an older version of) the
>C standard stated explicitely that assigning 0 (zero) to a pointer or
>comparing a pointer to 0 (zero) was equivalent to using the NULL
>pointer. I am pretty sure that pretty much every C implementation
>inplements NULL in one of the header files as (void *)0. The
>portability problems come onle when somebody passes 0 as an argument to
>an unprototyped function (which should be taken out of the stanedard
>anyway, regardless of legacy code, IMO)
I think the C standard is pretty clear; the NULL pointer constant is "0".
Code using "void *x = 0" or comparing a pointer to "0" is correct code.
Casper
--
Expressed in this posting are my opinions. They are in no way related
to opinions held by my employer, Sun Microsystems.
Statements on Sun products included here are not gospel and may
be fiction rather than truth.
Casper H.S. Dik - Network Security Engineer wrote:
> I think the C standard is pretty clear; the NULL pointer constant is "0".
>
> Code using "void *x = 0" or comparing a pointer to "0" is correct code.
Um, I must have missed something, but how did we get off on a thread about
nuances of C programming in the Lisp group? Personally, I find this even
more distracting than the (very entertaining) Microsoft-bashing threads.
The only good thing is that when I read this stuff about ``C'' it just
increases my gratefulness that I am able to use LIsp and have by some kind
of grace been saved from this drudgery of a language called C.
--
David J. Cooper Jr. Genworks International
········@genworks.com http://www.genworks.com
"...Embracing an Open Systems approach to Knowledge-based Engineering..."
In article <·······································@eskimo.com>,
Clayton Weaver <······@eskimo.com> wrote:
>(nothing to do with the rest of this debate, just slang clarification)
>
>On Mon, 29 Jun 1998, Martin Rodgers wrote:
>[]
>> In my experience, it looks like most C programmers are "cowboys". Perhaps
>> that's a UK expression that doesn't travel well, I don't know. It usually
>> refers to people in the building trade who use poor building materials
>> and don't finish the job properly. In some cases, like with an electrics,
>> this can potentially be lethal.
>
Over here in the States, a cowboy coder is one who charges ahead, writes
the software, and generally gets the job done - well, most of the job,
ignoring the bugs, and don't try to maintain or port the code. Any
similarity between this and any major software company practices is
darn well intended.
>Well, sorta. "Cowboy coders" has more to do with portability (ignoring it)
>and using compiler specific hacks that exploit a specific compiler's
>handling of something the C standard specifies as "undefined". The
>portability issue is something like:
>
To give a better example of machine-dependent behavior, consider
the following code:
printf("Size of an int: %d\n", sizeof(int));
which, according to two of the best C textbooks around early this decade,
printed out the size of an int in bytes.
In fact, %d prints a value as a signed int, and sizeof(int) returns
something of type size_t, which will be some sort of unsigned integral
type.
In most cases (definitely in this one), signed vs. unsigned doesn't
matter, since they're guaranteed the same bit patterns for the range
of nonnegative signed numbers, and there won't be over 32K bytes in
an int, as a general rule. The thing that matters is what sort of
unsigned type (usually long).
On many 32-bit systems, int and long are the same size, and so
approximately half of the possible values of an unsigned long print
just fine as ints. On Intel processors, and ones with similar
endianness, even with 32-bit ints, the long considered as an int
may print OK, since the least significant two bytes of the long
are at the start, and so can be treated as ints.
I didn't realize this code was bad until a student of mine tried
it on his Macintosh with two-byte ints set (I always used four-byte,
myself).
I'm going through all of this in painful detail to make a point:
this is what a good C programmer has to know. Some darn good
textbook writers slipped up here. Will the rest of us do as well
overall as they did?
>Another example is using 0 as an address value to substitute for NULL.
>A NULL pointer is an abstract data type. I don't know the actual C
>
0 works just fine, provided it's used in a pointer context. It's
guaranteed, provided that what's used is a constant integer expression
not involving functions (e.g., 0, '\0', or 3 - 3 are OK; log(1.0)
isn't; no variable can be acceptable).
The usual trap here is assuming that all-bits-zero is a null pointer,
which it is on most but not all machines. In particular, a cowboy
would use calloc() and assume that the memory returned would function
as null pointers.
Note that there is no one definition of the null pointer that will
serve everywhere, since there are systems in which pointers are
of different sizes (and, until recently, Intel-based systems in
which function and data pointers could be of different sizes were
almost ubiquitous). The difference between NULL and 0 is almost
purely stylistic ("almost" because NULL might be defined as some
sort of pointer, usable in some but not all contexts).
Again, I've included the long painful detail as exposition.
This is stuff that the non-cowboy programmer is going to have
to get right on a regular basis.
>It's like using expensive paint on the seawall side and adjoining sides of
>a house close to saltwater, and using cheap paint the same color on the
>back side of it. The cheap paint is not the right paint for the local
>environment, but the user of the house might never notice, because the
>salt air that peels paint always blows in from the other side of the
>house, and the local landscape just happens not to induce back-eddies
>behind the house. That would be cowboy construction.
>
Nah, cowboy construction is when you don't wash the house beforehand,
so you're painting over peeling paint and dirt, and you use a
spraybrush, missing a few spots and getting paint on some of the windows.
It's usually connected with a desire to optimize in some mostly-
useless fashion, such as getting fast C code that does almost
the right thing.
Now, if we consider Norvig and Graham, how many little details
have they slipped up in, making code unportable? I haven't
caught any. I think this shows why (a) cowboys are associated
with C and similar languages, and (b) why Lisp is such a nice
environment to work in.
--
David H. Thornley | These opinions are mine. I
·····@thornley.net | do give them freely to those
http://www.thornley.net/~thornley/david/ | who run too slowly. O-
In article <·····················@ptah.visi.com>, ········@visi.com
says...
> I'm going through all of this in painful detail to make a point:
> this is what a good C programmer has to know. Some darn good
> textbook writers slipped up here. Will the rest of us do as well
> overall as they did?
This is why I so often find myself refering people to the following book:
The Elements of Programming Style
by Brian W. Kernighan
2nd Edition
Paperback, 168 pages
Published by Computing McGraw-Hill
ISBN: 0070342075
None of the examples in my copy use C. They're either Fortran or PL/I.
While the authors of Lisp books are good at avoiding these programming
pitfalls, I feel this book still has something to offer Lisp users.
It may be stating what to us is obvious. I can't tell, as I read this
book before learning Lisp and FP.
Anyway, I still appreciate tips like "Make it right before you make it
faster", "Don't stop with your first draft", and "Say what you mean,
simply and directly". Best of all is, "Parenthesize to avoid ambiguity".
Some ideas transcend language.
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
···@wildcard.this.email.address.intentionally.left.crap.demon.co.uk (Martin Rodgers) writes:
> In article <················@naggum.no>, ······@naggum.no says...
>
> > have you really _needed_ a piece of software that runs only on Microsoft?
>
> When a client insists on it, yes. The alternative is find new clients.
I beg to disagree. When a client "insists" on it, I still have
several courses of action open to me, to "convince" that client. You
might wonder how many clients quickly fold-over once you simply state
that you cannot and will not take the responsibility of using MS
(or for that matter similiar) "products" on that project.
And if it indeed comes to this, that the potential client would rather
use MS than do business with you, then this is one of the basic
business lessons, that entrepreneurs have to learn: There is such a
thing as "dangerous" clients/contracts, which you are really better
off _not_ doing business with.
One thing that helps here, is a change of mindset: You don't want
something from the client, the _client_ wants something from you: A
_solution_ to a problem, not some piece of software.
People don't go the doctor and "specify" an open-heart operation, with
the condition that the doctor use this anesthetic and that
voodoo dance. You go the doctor and give him the details of your
problem. Then _the doctor_ will give you an overview of his
diagnosis, and what he intends to do about it. Now if you have
questions, or indeed disagree with the doctor, you are free to discuss
this with him, and/or talk with other doctors. _But_ in the end, it
is the doctor that gets to decide on the kind of treatment, and so it
has to be, since he is the one that has to bear the responsibility for
the final outcome.
No responsible doctor will argue "ok, this patient wants me to operate
on his heart without the use of foreign blood. Now if I refuse, he'll
go to another doctor, so I'll agree to do it, and hope for the best...",
and neither should a responsible software architect!
> Are there enough clients who don't use MS software for every software
> house? Again, this is money talking. We go where the money is. It may be
Are there enough patients who want the kinds of treatment I offer?
This is the wrong question! The question is: Can _I_ cure the
problems those patients suffer, and how? Then what remains is to
convince the patient, that it is in his best interests, that he not
tie my right hand behind my back, while ordering me to do open-heart
surgery. And if _I_ fail to do that (and this is primarily _my_
failure, because I'm the specialist, and he is the patient), than I'll
have to let go of that contract, because I simply cannot carry the
responsibility of complying with his wishes.
> a little hard to understand, but not everyone has the freedom to choose
> who to work for. It's very arrogant to suggest otherwise.
Most of us live in countries with free-market economies, and basic
human rights, which _guarantee_ that everyone has the freedom to choose
who to work for. And we even have some mechanisms to try to cussion
the effects of temporary market-mismatches. Otherwise free-market
economies would simply not work, and we would all be either out of
business, or work for _very_ little money. There are indeed currently
some problems in some countries and professions on the job-market, but
the business we are in, is most certainly not in trouble, to the
contrary...
And especially those of us, who are freelancers have this freedom to a
much, much higher and finer degree! We simply have to _exercise_
those rights, or at least be prepared to do so. Of course this can
have temporary negative personal consequences, but this is the risk
we must be prepared to take, to reap the benefits of this self-same
freedom. Otherwise you will really end up as a slave of the forces of
the market.
> This is why I've been saying for years: share the secret. Instead of
> insulting people for not using Lisp, we should lead them to Lisp. Instead
> of telling them them Lisp is, let them discover it for themselves. That
> way, their egos will tell them how wonderful _they_ are for making the
> discovery. After all, we don't need the credit ourselves.
This whole discussion is really not about Lisp at all[1]. A number of
those now being "forced" to use C++ would probably have been "forced"
to use Lisp in the AI-boom in the 80s. The same problems would have
applied, the same "market pressures", etc.
Maybe this is the hint, that I should now shut up, and return to the
topic of this newsgroup. OTOH I'd probably miss Erik's acute summaries
of this problem, which IMHO are highly compressed, deep observations
about this and other issues in our field of work. And although I'd
agree that Erik's style sometimes might be more agressive than it need
be, he still is able to express things more succinctly, and more
pressing than I could even try to do.
> invest their ego. In a few years from now, MS will lose big time, while
> we will win merely by saying, "I told you so", and reminding everyone
> what we've been saying for years: you can win big with Lisp.
I think this is not going to work: Nobody likes the bearer of bad
news, and simply being able to say "I told you so" will not get you
many contracts. And even worse: When MS falls, it is extremely
likely to be replaced by yet another monopoly within a few years. It
is not MS that we've got to change, or get rid of, it is the patients
belief-systems and their resistance to mass-marketing we've got to
lastingly influence, as Erik puts it so nicely.
Regs, Pierre.
BTW: My use of doctors, voodoo dances, etc. is not to be construed as
a judgement of the maturity of modern medicine, or the supremacy
and/or infallability of those wearing white coats.
Footnotes:
[1] And it was my fault really (which Erik spotted so promptly, and
rebutted acutely) not to clue up to this fact before answering Zeno's
questions. Another case of not tackling the problem behind the
problem, really.
--
Pierre Mai <····@cs.tu-berlin.de> http://home.pages.de/~trillian/
"Such is life." -- Fiona in "Four Weddings and a Funeral" (UK/1994)
In article <··············@torus.cs.tu-berlin.de>, ····@cs.tu-berlin.de
says...
> I beg to disagree. When a client "insists" on it, I still have
> several courses of action open to me, to "convince" that client. You
> might wonder how many clients quickly fold-over once you simply state
> that you cannot and will not take the responsibility of using MS
> (or for that matter similiar) "products" on that project.
Either you don't have 300 pound gorillas for clients, or you're more
successful than most people. Perhaps your clients have fewer pointy
haired managers? It looks a lot like most people would rather take the
money and suffer. I know loads of people who don't have any great
fondness for MS software, so why do they suffer? Pragmatism?
> And if it indeed comes to this, that the potential client would rather
> use MS than do business with you, then this is one of the basic
> business lessons, that entrepreneurs have to learn: There is such a
> thing as "dangerous" clients/contracts, which you are really better
> off _not_ doing business with.
You might wonder how many people quickly fold-over when they need to eat.
> One thing that helps here, is a change of mindset: You don't want
> something from the client, the _client_ wants something from you: A
> _solution_ to a problem, not some piece of software.
Have you ever read Dilbert? Ask yourself where the power is.
IT people don't run corporations. Corporations shit on IT people.
> Most of us live in countries with free-market economies, and basic
> human rights, which _guarantee_ that everyone has the freedom to choose
> who to work for. And we even have some mechanisms to try to cussion
> the effects of temporary market-mismatches. Otherwise free-market
> economies would simply not work, and we would all be either out of
> business, or work for _very_ little money. There are indeed currently
> some problems in some countries and professions on the job-market, but
> the business we are in, is most certainly not in trouble, to the
> contrary...
Ah, no. We can apply for a job at a particular company, but that doesn't
mean they'll employ us. Nor does it mean they'll listen when we say,
"This is wrong, let's change it." There's a heirarchy. The people at the
bottom don't get to run things. IT is at the bottom.
It's a dumb situation, but hardly suprising. How many people running
corporations are computer literate? Obviously some of them are, so that
suggests another question. Why do they choose MS software?
Pehaps because the problem is larger than MS. Why did people used to
choose IBM? Why do some people _still_ choose IBM? Because they can?
> This whole discussion is really not about Lisp at all[1]. A number of
> those now being "forced" to use C++ would probably have been "forced"
> to use Lisp in the AI-boom in the 80s. The same problems would have
> applied, the same "market pressures", etc.
Agreed. The problem is larger than MS, C++, Lisp, you-name-it.
> Maybe this is the hint, that I should now shut up, and return to the
> topic of this newsgroup. OTOH I'd probably miss Erik's acute summaries
> of this problem, which IMHO are highly compressed, deep observations
> about this and other issues in our field of work. And although I'd
> agree that Erik's style sometimes might be more agressive than it need
> be, he still is able to express things more succinctly, and more
> pressing than I could even try to do.
When he makes an effort, which he does most of the time, then he's
certainly worth reading. Unfortunately, I think this is one of those
subjects where he redmists a little.
I recommend reading some Scott Adams. He can write about stupidity in a
language that even those pointy haired managers can understand. They
laugh at the jokes, apparently not realising that the jokes are about
themselves. This is a wonderful way to communicate a message. Instead of
feeling insulted and defensive, these managers listen to what Adams is
saying. The soft sell works.
> I think this is not going to work: Nobody likes the bearer of bad
> news, and simply being able to say "I told you so" will not get you
> many contracts. And even worse: When MS falls, it is extremely
> likely to be replaced by yet another monopoly within a few years. It
> is not MS that we've got to change, or get rid of, it is the patients
> belief-systems and their resistance to mass-marketing we've got to
> lastingly influence, as Erik puts it so nicely.
This is what I've been saying. The soft sell works by winning friends,
while the hard sell only makes enemies. Friends will listen to what you
have to say, and be open minded. Enemies will attack you no matter what
you say.
Did I not say that this is bigger than MS? That if we were to remove MS
and all their works from the surface of the planet, another company would
just move in and take their place?
If not for the redmist, I suspect that Erik might realise that we agree
on far more things than we disagree. Alas, he appears not to read my
posts, but simply flames what he _imagines_ I write. In the past, I've
had to carefully stress the points on which we agree before he'd stop
attacking me. He seems unable to debate even the simplest idea with me.
As I've said several times in this thread, we agree on the problem, but
not the tactics. However, the differences are very small. Are these minor
differences really worth the flames? I suspect not.
I welcome _constructive_ criticism.
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
Martin Rodgers wrote:
> I recommend reading some Scott Adams.
Better yet, read Geoffrey Moore's _Crossing The Chasm_, which explains very
clearly why companies like IBM and Microsoft come to dominate their markets.
The phenomenon he describes is not going to go away, and railing at it does
nothing but raise one's blood pressure.
-- Scott
* * * * *
To use the email address, remove all occurrences of the letter "q".
In article <·················@zeta-sqoft.com>, ·····@zeta-sqoft.com
says...
> Better yet, read Geoffrey Moore's _Crossing The Chasm_, which explains very
> clearly why companies like IBM and Microsoft come to dominate their markets.
> The phenomenon he describes is not going to go away, and railing at it does
> nothing but raise one's blood pressure.
This is a good point. There are a number of books like this, so, thanks
for the recommendation.
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
* Erik Naggum wrote:
> have you really _needed_ a piece of software that runs only on Microsoft?
Yes, often. I've *really needed* something that would read some
proprietary MS format, like Word, and actually deal with it well
rather than leaving me thrashing around for a week trying to reverse
engineer the formatting.
Of course I don't use proprietary document formats, but I can't stop
people sending me stuff in them. Proprietary formats are a virus,
MIME is the disease vector.
--tim
In article <···············@todday.aiai.ed.ac.uk>, ···@aiai.ed.ac.uk
says...
> Yes, often. I've *really needed* something that would read some
> proprietary MS format, like Word, and actually deal with it well
> rather than leaving me thrashing around for a week trying to reverse
> engineer the formatting.
This is one of the pressures I refered to in another post. When a client
is several orders of magnitude larger than your employer, you may find it
hard to pursuade them to change their software. Instead, they'll expect
you to change _your_ software.
> Of course I don't use proprietary document formats, but I can't stop
> people sending me stuff in them. Proprietary formats are a virus,
> MIME is the disease vector.
A very good way of putting it.
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
In article <···············@todday.aiai.ed.ac.uk>,
Tim Bradshaw <···@aiai.ed.ac.uk> writes:
> * Erik Naggum wrote:
>> have you really _needed_ a piece of software that runs only on Microsoft?
>
> Yes, often. I've *really needed* something that would read some
> proprietary MS format, like Word, and actually deal with it well
> rather than leaving me thrashing around for a week trying to reverse
> engineer the formatting.
>
> Of course I don't use proprietary document formats, but I can't stop
> people sending me stuff in them. Proprietary formats are a virus,
> MIME is the disease vector.
>
> --tim
Sure, you can't keep people from sending you messages using proprietary
document formats. You can take a stand against them though. Make the sender
convert it into ASCII or just delete it. ("Oh, was that important? I thought
it was some binary SPAM. Please send it again in ASCII.) You DO have a choice!
Mike McDonald
·······@mikemac.com
In article <····················@news.teleport.com>, ·······@mikemac.com
says...
> Sure, you can't keep people from sending you messages using proprietary
> document formats. You can take a stand against them though. Make the sender
> convert it into ASCII or just delete it. ("Oh, was that important? I thought
> it was some binary SPAM. Please send it again in ASCII.) You DO have a choice!
IT departments get treated like shit by corporations. That may sound like
whinging to some, but it's a fact. Ask the support people. Alternately,
just read the Dilbert strip for a few weeks. It's all there. Shit.
Where does a 300 pound gorilla sit? Yes, you have a choice. I just
wouldn't want to be in your shoes when try telling that gorilla which
format to send a document in. When it's only text, this is not too hard,
unless you happen to work for a particularly, err, corporate gorilla.
Alas, not all documents are text. When you're working with multimedia, it
can sometimes get really complicated. You may need Mac _and_ Windows.
No matter what you use, lots of messy - and time consuming - format
conversions may be needed, all because a 300 pound gorilla likes to use
the wrong software.
When that 300 pound gorilla is also hassling you about deadlines, your
choices can get real limited. You may have noticed that there are one or
two 300 pound gorillas in the business world!
Does anyone have an elephant gun I could use?
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
In article <··························@news.demon.co.uk>,
···@wildcard.this.email.address.intentionally.left.crap.demon.co.uk (Martin Rodgers) writes:
> In article <····················@news.teleport.com>, ·······@mikemac.com
> says...
>
>> Sure, you can't keep people from sending you messages using proprietary
>> document formats. You can take a stand against them though. Make the sender
>> convert it into ASCII or just delete it. ("Oh, was that important? I thought
>> it was some binary SPAM. Please send it again in ASCII.) You DO have a choice!
>
> IT departments get treated like shit by corporations. That may sound like
> whinging to some, but it's a fact. Ask the support people. Alternately,
> just read the Dilbert strip for a few weeks. It's all there. Shit.
>
> Where does a 300 pound gorilla sit? Yes, you have a choice. I just
> wouldn't want to be in your shoes when try telling that gorilla which
> format to send a document in. When it's only text, this is not too hard,
> unless you happen to work for a particularly, err, corporate gorilla.
>
> Alas, not all documents are text. When you're working with multimedia, it
> can sometimes get really complicated. You may need Mac _and_ Windows.
> No matter what you use, lots of messy - and time consuming - format
> conversions may be needed, all because a 300 pound gorilla likes to use
> the wrong software.
>
> When that 300 pound gorilla is also hassling you about deadlines, your
> choices can get real limited. You may have noticed that there are one or
> two 300 pound gorillas in the business world!
>
> Does anyone have an elephant gun I could use?
But unless you're conscripted labor or a slave, YOU CHOSE to work for the
300lb gorilla. (Shouldn't that be some number of stones gorilla?) You just
chose that putting up with that "shit" plus your salary was worth it to you.
But you still made the decision.
Mike McDonald
·······@mikemac.com
In article <·····················@news.teleport.com>, ·······@mikemac.com
says...
> But unless you're conscripted labor or a slave, YOU CHOSE to work for the
> 300lb gorilla. (Shouldn't that be some number of stones gorilla?) You just
> chose that putting up with that "shit" plus your salary was worth it to you.
> But you still made the decision.
That's why I said you do indeed have a choice, just not much of a choice.
A great many people choose to work for corporations. The fools!
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
In article <··························@news.demon.co.uk>,
···@wildcard.this.email.address.intentionally.left.crap.demon.co.uk (Martin Rodgers) writes:
> In article <·····················@news.teleport.com>, ·······@mikemac.com
> says...
>
>> But unless you're conscripted labor or a slave, YOU CHOSE to work for the
>> 300lb gorilla. (Shouldn't that be some number of stones gorilla?) You just
>> chose that putting up with that "shit" plus your salary was worth it to you.
>> But you still made the decision.
>
> That's why I said you do indeed have a choice, just not much of a choice.
You have lots of choices! You could be an indepentant consultant who doesn't
work on things he doesn't want to (like someone we "know"!), you could work
for an "inlightened" company, you could become a carpenter, lawyer (soliciter
for some), or a gigalo. Life's full of opportunities. But of course if you
insist "I want the same job but with this one change and to get paid a million
pounds/year", then yes, you're hosed.
> A great many people choose to work for corporations. The fools!
I wouldn't classify them as fools. They made a trade off inorder to satisfy
some of their needs. More power to them!
Mike McDonald
·······@mikemac.com
In article <······················@news.teleport.com>,
·······@mikemac.com says...
> > That's why I said you do indeed have a choice, just not much of a choice.
>
> You have lots of choices! You could be an indepentant consultant who doesn't
> work on things he doesn't want to (like someone we "know"!), you could work
> for an "inlightened" company, you could become a carpenter, lawyer (soliciter
> for some), or a gigalo. Life's full of opportunities. But of course if you
> insist "I want the same job but with this one change and to get paid a million
> pounds/year", then yes, you're hosed.
You could be a lot of things. Many people I know who work with computers
seem to be independent consultants these days, and in some cases because
they got sick of having major technical choices made for them.
What percentage of the world's IT people are independent consultants? I
don't know, but I am curious. There should be more of us, and perhaps
there will be. Is the trend still moving increasingly toward outsourcing
IT work? Even so, choices can still be limited by a sufficiently big
client that refuses to stop using, say, MS software. You only need one
such client to justify using the same tools that they do.
Not that this has much to do with Lisp. It's more of a Dilbert thing.
See the strip where Alice briefly escaped from the office, and worked for
a while at home. See the endless ways in which choices are removed.
> > A great many people choose to work for corporations. The fools!
>
> I wouldn't classify them as fools. They made a trade off inorder to satisfy
> some of their needs. More power to them!
Exactly. A trade off. That's my point. That's why they use MS software.
Some people hate Macs with the same passion that others hate Windows. It
doesn't mean they don't have to deal with Mac software.
The same is true for apps. If you get sent a document that is in a format
that you loath, then you convert it. Not all convertions are simple,
which could be why some people hate apps that they never use themselves.
One man's meat may be another man's poison. This is bigger than MS, and
nothing at all to do with Lisp. A few years ago, we could see loads of
attacks on Lisp from C++ programmers. MS didn't create C++. AT&T did.
This is why I prefer to leave politics out of this newsgroup. I know how
devisive it can be. There will be no simple answer to this problem, and I
doubt we'll find it in here. When it is found, I bet it'll be in the news
headlines. It'll need to be that big to make a difference.
Alternately, perhaps a long series of very small changes is what we'll
get. Small but significant conversions to the "one true way". Hmm.
Whatever happens, I don't believe that MS hatemail belongs in here. We
have plenty of other newsgroups for that kind of thing:
alt.conspiracy.microsoft
alt.destroy.microsoft
alt.microsoft.sucks
alt.sad-people.microsoft.lovers
alt.microsoft.crash.crash.crash
And perhaps even alt.bill-gates.kind.benificent.loving.big-brother and
alt.fan.bill-gates. Not everything in the alt.* heirarchy that looks like
a newsgroup for fans is full of fanmail. There's a tradition of using
such names to trap unwary fans.
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
In article <··························@news.demon.co.uk>,
···@wildcard.this.email.address.intentionally.left.crap.demon.co.uk (Martin Rodgers) writes:
> In article <······················@news.teleport.com>,
> ·······@mikemac.com says...
> One man's meat may be another man's poison. This is bigger than MS, and
> nothing at all to do with Lisp. A few years ago, we could see loads of
> attacks on Lisp from C++ programmers. MS didn't create C++. AT&T did.
>
> This is why I prefer to leave politics out of this newsgroup. I know how
> devisive it can be. There will be no simple answer to this problem, and I
> doubt we'll find it in here. When it is found, I bet it'll be in the news
> headlines. It'll need to be that big to make a difference.
>
> Alternately, perhaps a long series of very small changes is what we'll
> get. Small but significant conversions to the "one true way". Hmm.
>
> Whatever happens, I don't believe that MS hatemail belongs in here.
This has nothing to do with MS or politics or any such thing. It has to do
with empowering yourself and taking charge of your own destiny. If you don't
like working with products from XYZ LTD. or working with Joe Blow, then DON'T.
Life's too short to waste time and effort on those things.
Mike McDonald
·······@mikemac.com
In article <······················@news.teleport.com>,
·······@mikemac.com says...
> This has nothing to do with MS or politics or any such thing. It has to do
> with empowering yourself and taking charge of your own destiny. If you don't
> like working with products from XYZ LTD. or working with Joe Blow, then DON'T.
Agreed. The general assumption here seems to be that MS software is
always wrong. I keep an open mind, mainly because to me they're just
tools. If a tool is a problem, find a better one. When somebody else
insists that you use, or not use, a tool, then they are the problem.
> Life's too short to waste time and effort on those things.
Indeed it is.
--
Please note: my email address is munged; You can never browse enough
Hit the advocates three timesd over the head with the Linux Advocacy
mini-HOWTO <URL:http://sunsite.unc.edu/LDP/HOWTO/mini/Advocacy.html>
* Mike McDonald wrote:
> Sure, you can't keep people from sending you messages using proprietary
> document formats. You can take a stand against them though. Make the
> sender convert it into ASCII or just delete it. ("Oh, was that
> important? I thought it was some binary SPAM. Please send it again
> in ASCII.) You DO have a choice!
No, I don't, or not usefully. These people are clients, who are much
larger organisations than we are, and if we say `no won't read your
mail' they'll just go elsewhere, because they have been infected with
the proprietary-format virus (MS strain), and one of its earliest
effects (before the grotesque swelling, pustules, and weeping sores
which mark the later stages start appearing) is to destroy the part of
their minds which allows them to reason rationally about why open
formats are a good thing. Indeed, they are often under the delusion
that whatever cruft they deal in *is* an open format (Anyone see Bill
Gates' article in The Economist the other week? -- apparently Windows
is an `open standard'...).
--tim
(What newsgroups should this stuff go to?)
In article <···············@todday.aiai.ed.ac.uk>,
Tim Bradshaw <···@aiai.ed.ac.uk> writes:
> * Mike McDonald wrote:
>
>> Sure, you can't keep people from sending you messages using proprietary
>> document formats. You can take a stand against them though. Make the
>> sender convert it into ASCII or just delete it. ("Oh, was that
>> important? I thought it was some binary SPAM. Please send it again
>> in ASCII.) You DO have a choice!
>
> No, I don't, or not usefully. These people are clients, who are much
> larger organisations than we are, and if we say `no won't read your
> mail' they'll just go elsewhere, because they have been infected with
> the proprietary-format virus (MS strain), and one of its earliest
> effects (before the grotesque swelling, pustules, and weeping sores
> which mark the later stages start appearing) is to destroy the part of
> their minds which allows them to reason rationally about why open
> formats are a good thing. Indeed, they are often under the delusion
> that whatever cruft they deal in *is* an open format (Anyone see Bill
> Gates' article in The Economist the other week? -- apparently Windows
> is an `open standard'...).
>
> --tim
>
> (What newsgroups should this stuff go to?)
Who held the gun to your head and made you take them on as clients? You do
have a choice and you made it. You prefered their money over your distaste for
their technical preferences.
Mike McDonald
·······@mikemac.com
·······@mikemac.com (Mike McDonald) writes:
> Sure, you can't keep people from sending you messages using proprietary
> document formats. You can take a stand against them though. Make the sender
> convert it into ASCII or just delete it. ("Oh, was that important? I thought
> it was some binary SPAM. Please send it again in ASCII.) You DO have a
> choice!
Sure, and I can always choose to do something that causes me to lose
my job. Maybe I *like* my job in most respects and therefore put up
with some annoying stuff. Indeed, I may have to put up with it in
order to keep the otherwise likeable job, and as a result I may
need some microslaveware.
-- jd
* Martin Rodgers wrote:
> Like Erik, I too question Microsoft's business ethics. I also question
> the policies of my country's government, as do a lot of other people.
> How do we "vote" for better business ethics without sounding like
> socialists? (A very real problem here in the UK.)
I don't question Microsoft's business ethics. Microsoft are doing
*exactly* what I would do in their position -- make as much money as
they possibly can, regardless of consequence. Because they're a
monopolist in a software market the preferred technique to do this is
destruction of competition, and the preferred technique to do *that*
is rapidly-changing, deliberately-incompatible systems & interfaces.
That's fine, that's what you should *expect* a monopolist to do.
The problem is not Microsoft, the problem is the legislative framework
that is letting them thrive.
--tim
Tim Bradshaw <···@aiai.ed.ac.uk> writes:
> I don't question Microsoft's business ethics. Microsoft are doing
> *exactly* what I would do in their position -- make as much money as
> they possibly can, regardless of consequence.
Remind me never to hire you. (Not that I'm in a position to hire
anybody, but...)
> That's fine, that's what you should *expect* a monopolist to do.
That is what I expect a monopolist to do, but that doesn't make it
_fine_, no way.
> The problem is not Microsoft, the problem is the legislative
> framework that is letting them thrive.
As I see it, the problem is all the customers that let them get away
with it.
Stig Hemmer.
* Stig Hemmer wrote:
> [I wrote]
>> The problem is not Microsoft, the problem is the legislative
>> framework that is letting them thrive.
> As I see it, the problem is all the customers that let them get away
> with it.
Well, sure, that's the same thing, isn't it? The customers should be
electing governments who enact legislation that is effective against
monopolies (:).
--tim
In article <···············@todday.aiai.ed.ac.uk>, ···@aiai.ed.ac.uk
says...
> I don't question Microsoft's business ethics. Microsoft are doing
> *exactly* what I would do in their position -- make as much money as
> they possibly can, regardless of consequence. Because they're a
Indeed. They're hardly unique, in that respect. Nor is this restricted to
purely business practice. Individuals and governments, too.
It's very hard to question it without sounding anti-capitalist. However,
the "zero growth" idea isn't exactly anti-capitalist, nor is business
with an "ethical dimension". Individuals and governments are also setting
good examples.
> monopolist in a software market the preferred technique to do this is
> destruction of competition, and the preferred technique to do *that*
> is rapidly-changing, deliberately-incompatible systems & interfaces.
> That's fine, that's what you should *expect* a monopolist to do.
Agreed. That's why I refered to similar practices in other industries,
like oil, steel, rail, etc.
> The problem is not Microsoft, the problem is the legislative framework
> that is letting them thrive.
I've hinted at this myself, only a few days ago. The problem is bigger
than Microsoft. Removing MS won't change a thing, as one of their rivals
will just take their place.
Tolkien described a fictional form of this problem. ;)
--
Please note: my email address is munged; You can never browse enough
Not coming to you from Glastonbury
···@wildcard.this.email.address.intentionally.left.crap.demon.co.uk (Martin Rodgers) writes:
> I've hinted at this myself, only a few days ago. The problem is bigger
> than Microsoft. Removing MS won't change a thing, as one of their rivals
> will just take their place.
What an obnoxiously defeatist view! Removing MS will change something
even if one of their rivals does take their place (which is far from
certain and may involve a to-be-enjoyed delay).
-- jd
In article <··············@gairsay.aiai.ed.ac.uk>,
····@gairsay.aiai.ed.ac.uk says...
> What an obnoxiously defeatist view! Removing MS will change something
> even if one of their rivals does take their place (which is far from
> certain and may involve a to-be-enjoyed delay).
As I've said before, and I'll say again, this problem is much bigger than
MS. It existed long before MS existed. Look at the history.
--
Please note: my email address is munged; You can never browse enough
"Ahh, aren't they cute" -- Anne Diamond describing drowning dolphins
Tim Bradshaw <···@aiai.ed.ac.uk> writes:
<
< * Martin Rodgers wrote:
<
< > Like Erik, I too question Microsoft's business ethics. I also question
< > the policies of my country's government, as do a lot of other people.
< > How do we "vote" for better business ethics without sounding like
< > socialists? (A very real problem here in the UK.)
<
< I don't question Microsoft's business ethics. Microsoft are doing
< *exactly* what I would do in their position -- make as much money as
< they possibly can, regardless of consequence. Because they're a
< monopolist in a software market the preferred technique to do this is
< destruction of competition, and the preferred technique to do *that*
< is rapidly-changing, deliberately-incompatible systems & interfaces.
< That's fine, that's what you should *expect* a monopolist to do.
If Microsoft is a monopoly, there would be no viable competition. You
cannot be in "microsoft's position", you are a person (if you will
allow me this assumtion). Corporations are machines protected by the
state in which their charter resides.
It is the primary and sole responsibility of all corporate officers to
optimize shareholder value. They usually do this by allowing the
consumer to perceive the full benefits of their product (usually in
relation to some problem or need). They should not recognize
competition, as it is a problem of the corporation and not the consumer
(unless it becomes a problem to the consumer of course, in which case
the corporation is expected to fix the problem).
I see the computer, or at least the Personal Computer, industry as
some kind of oligopoly. Since the operating system is dependent on the
processor software, as well as the bios software (which is not even
found within much of today's hardware). All of which is offered by a
few corporations that openly admit to the sharing of secrets (they use
it for cool looking stickers, labels, and other such means of
identification).
I have tasm - but intel corporation has openly stated that it has
extensions (to something extendible I presume) and support only for
masm in it's new "diagnostic VTune" (who wants recognize a problem - I
want to _see_ what inefficiency has been tuned!) program (and it only
runs under windows95 - which did not come with my intel processor). I
can only imagine how difficult it must have been for the tasm
developers to use the 386 debug registers without any support from the
processor corporation that _adds_ new instructions to a processor with
intentions of making it slower, hotter (think waffle iron), and more
innovative.
For example, the new pentium IIs have better support for 16 bit
applications (like IBM's DOS), and MMX (which is not an acronym). An
idea that the intel engineers must have really struggled with. Who
wanted to use that FMUL instruction anyway? With the new innovative MMX
technology, you can just add the numbers up using an open general
purpose register and ADD all the while checking for some silly flag.
Truly innovative technology! (I would have used a compile with support
for the fp unit).
Their are so many pieces of software that are required to make a
program "operate". Anyone can make new hardware for the pc, but if you
want to have it be plug-and-play, it must have support in the bios and
support for the plug-and-play bios spec; not to mention it must have
the drivers pre-loaded into the operating system. Then the users are
free to plug the new hardware in, and play with it _all_ day until it
finally works. Having five devices all trying figure out which one is
going to seize interrupt request (IRQ) number 7 - while not using
communication (com) port 3 if com port 1 is a mouse - is not a good
thing to do at _every_ boot time (as my experiences have shown to me
at least).
I wish a single company did make the bios, processor, and software for
my computer. It would be easy to complain to them if it didn't work
correctly (especially if I didn't know what part was causing the
machine to not act as the documentation (those books that did not come
with my computer (had to order them from 18 different places instead))
stated) so that they could fix it. With PCs you can complain - but the
responsible party for the fault is rarely known, and certainly never
found responsible; hence it is never corrected. Where are the
benefits?
< The problem is not Microsoft, the problem is the legislative framework
< that is letting them thrive.
The problem with microsoft is they have stolen the credit for having
innovated the elimination of consistency and reliability with regards
to software. It is a very smart company, just not very original.
This all relates to Lisp; I just didn't mention it, sorry.
Steve Gonedes <········@worldnet.att.net> writes:
> It is the primary and sole responsibility of all corporate officers to
> optimize shareholder value.
Is it, really? But for the argument's sake, let's suppose it is so.
Then why do they always go after the small local peaks, instead of
the big global maximum? :-)
Andreas
In article <··············@laphroig.mch.sni.de>,
Andreas Eder <···@laphroig.mch.sni.de> wrote:
>Steve Gonedes <········@worldnet.att.net> writes:
>
>> It is the primary and sole responsibility of all corporate officers to
>> optimize shareholder value.
>
>Is it, really? But for the argument's sake, let's suppose it is so.
>Then why do they always go after the small local peaks, instead of
>the big global maximum? :-)
>
It is, legally, in the United States. (Corporations are "people" in
some respects that help them, not in some that would hurt them.
I don't much care for the laws concerning corporations in the U.S. I
have no knowledge of German business law.)
However, you ask why they always head for the local maxima, instead of
trying for something better.
I don't know. Can somebody explain why 80% of shops are SEI level 1,
why people were using two-digit years at a shop I worked at in '96,
and why people are writing new stuff in COBOL? I think this will
give us a hint as to why execs like to use hill-climbing rather than,
say, simulated annealing or best-first search. (Ever felt you
were working for a company that was selecting its strategy by the
British Museum method - try every one until you get the right one?)
Anyway, to put this on-topic, at least allusively, why don't more
people use Common Lisp?
--
David H. Thornley | These opinions are mine. I
·····@thornley.net | do give them freely to those
http://www.thornley.net/~thornley/david/ | who run too slowly. O-
In my humble opinion, all of this NT grousing is bogus. If the customer
says "we want an NT application to do x", that's just a requirement. So
add $200k to the price and a year to the estimated delivery date. End of
problem.
Regards, Clayton Weaver ······@eskimo.com (Seattle)
--
Regards, Clayton Weaver ······@eskimo.com (Seattle)
"Since this is a technical newsgroup and has mostly done a pretty good job
In article <··············@laphroig.mch.sni.de>,
Andreas Eder <···@laphroig.mch.sni.de> writes:
> Steve Gonedes <········@worldnet.att.net> writes:
>
>> It is the primary and sole responsibility of all corporate officers to
>> optimize shareholder value.
>
> Is it, really? But for the argument's sake, let's suppose it is so.
> Then why do they always go after the small local peaks, instead of
> the big global maximum? :-)
>
>
> Andreas
Ever hear the saying "A bird in the hand is worth two in the bush"?
Mike McDonald
·······@mikemac.com
In article <··············@laphroig.mch.sni.de>, Andreas Eder
<···@laphroig.mch.sni.de> wrote:
>Steve Gonedes <········@worldnet.att.net> writes:
>
>> It is the primary and sole responsibility of all corporate officers to
>> optimize shareholder value.
>
>Is it, really? But for the argument's sake, let's suppose it is so.
>Then why do they always go after the small local peaks, instead of
>the big global maximum? :-)
>
>
>Andreas
Because software is a *subscription* industry. If you acheive the global
maximum, then you sell lots of that release, and very little of any
subsequent release.
Broken products provide an endless revenue stream because customers will
always upgrade in the hope of fixing the many bugs.
It's a sad fact of market economies that products that represent *global*
maxima are not very profitable. (when was the last time you bought a new
salt shaker?).
Raf
In article <··············@laphroig.mch.sni.de>, Andreas Eder
<···@laphroig.mch.sni.de> wrote:
>Steve Gonedes <········@worldnet.att.net> writes:
>
>> It is the primary and sole responsibility of all corporate officers to
>> optimize shareholder value.
>
>Is it, really? But for the argument's sake, let's suppose it is so.
>Then why do they always go after the small local peaks, instead of
>the big global maximum? :-)
>
>
>Andreas
Because software is a subscription industry. If you get it right, your
customers never need to upgrade.
Raf
--
Raffael Cavallaro
Zeno wrote:
> ...
> > managers act weird, too. the Year 2000 problem is little more than the
> > age-old myth that the end of world coincides with new centuries, and my
> > guess there are still so many managers who haven't reacted in time is
> > that they don't actually believe there will be any life on planet earth
> > in new millennium, anyway, so why waste all the money? think about it.
>
> You say that you hammered on my Microsoft affiliation so that I would
> express the sentiments I did. I assume this was so you could post a
> message like this. I found this to be one of the most entertaining
> messages I have ever read. Surely though, it would have been more
> entertaining had it not been at my expense.
Since you are new in this group: here is a word of advice: pay close
attention to what technical advice Erik gives, and filter out the way he
delivers his message if you don't like it. You'll benefit from the
former.
> I would have to think extremely hard, and have the help of some
> mushrooms to understand that managers are buying Microsoft products
> because there will be no life on planet earth in a short few years. I
> would have to think equally hard to understand how all of these people
> with these crumbling projects could blame themselves and everyone but
> Microsoft for the problem, with you being the only one who can see the
> truth of the ugly world-takover of Microsoft. It's a company.
> Nothing more, nothing less.
I would have put it slightly differen: MS has taken over the role that
IBM had in the 70s and early to mid 80s: you can't (or couldn't} be
blamed / fired for buying their products, even though better, more cost
effective products are available. Both companies do | did their best to
exploit this mindset.
> ...
> It would be absolutely ludicrous for either of us to say that we have
> not been molded by our environments, but you seem to think that your
> environment has not affected you. Mass media is not a fad, and
> neither is the automobile. Your solution of nipping these things in
> the bud to save mankind is quite far-fetched to say the least. I
> assume you don't supply Internet connectivity to your customers, or do
> you allow yourself to get sucked into the communication vortex just a
> little bit, when you cannot show your customers what fools they are to
> want to communicate with the rest of the world?
If I understand the above remark correctly, you are saying that you need
MS products to connect to the Internet. Nothing would be further from
the truth.
> ...
> What I have been after is one or two of the testimonials you speak of.
> I am perfectly willing to listen to them, but instead of testimonials
> to the glory of CL, all I hear is why MS is bad. You are absolutely
CL is a language, VB a combination of language with development
environment with presentation environment, MS a company. Right now we
are into comparing apples with oranges with cherries.
As far as the reasons why I think CL (and other LISP like languages) are
superior to (the language component of) VB: much better at letting you
form your own abstractions, much better cod structuring, the run time
system takes care of memory management, very easy (for me) syntax.
Unless your approach to problem structuring isn't hopelessly stuck in a
mindset formed by one language, you'll most likely will (perhaps after
some learning period) produce more reliable code much faster, probably
considerably more reliable. The only one who can answer this for you is
yourself: get hold of some Lisp implementation, play around with it,
see how you take to it. If you never used anything similar, get a good
introductory text to guide you in your experimentation. Best check the
FAQ (it gets posted to this newsgroup regularly) about available
implementations and recommended books. If most of your work involves
GUI: you'll have to settle for one implmentation. Most free Lisps
require some work to integrate some of the available GUIs; and I don't
have the faintest idea what is available inder Win32. The FAQ probably
will help, and I suspect that all the commercial implementations have
something pretty nifty included.
> ...
> High quality software is not impossible, impractical, or insane--it is
> a worthwhile goal to be strived for. Communism and anti-innovation
> *are* impossible, impractical, *and* insane. How can you complain of
> sleeping masses and lack of individualism and, in the same breath,
> promote communism? Anti-innovation is just as ludicrous. People will
> innovate--if they did not, we would all be sitting around in caves,
> unable to figure out that we could get more food by sharpening a stick
> and throwing it at an animal.
>
> Your beliefs that communism and anti-innovation are worthwhile
> endeavors does lend more insight to your hatred of one man for making
> so much more than the others, and to your rejection of new technology.
I am curious about your definition of communism. I find it rather odd
that you read communist propaganda into Erik's post. As for
innovation: you seem to confuse gratuitous incompatibilities with
innovation. I'll abstain to draw any conclusions about your political
orientation from that.
> ...
--
Hartmann Schaffer
Guelph, Ontario, Canada
········@netcom.ca (hs)
* Zeno the Anonymous Poster
| My comments were in response to Erik saying that it was impossible to be
| concerned with MS business ethics without being associated with other
| things considered impossible, impractical, or insane, such as
| high-quality software, communism, or anti-innovation. I am quite sure
| that Erik doesn't think that high-quality software is impossible,
| impractical, or insane, and I took his statements to mean that neither
| are his other two examples, communism and anti-innovation.
|
| I don't know if you really are interested or if your statement was
| rhetorical, but my definition of communism is that it is a belief that
| "From each according to ability, and to each according to need," will
| work. This, I was taught in school, is the most basic of definitions of
| communism, and it precludes anyone making vast amounts more of money than
| others simply because they have the ability to do so. Erik's reference
| to communism in the same example as high-quality software, along with
| incessant repetition about the amount of money Bill Gates earns, leads me
| further to believing that he was speaking of communism as something not
| impossible, impractical, or insane.
you seem to need the word "respectively" in order to make three things
match three other things. let's try this again for improved legitibility:
Microsoft is so good at turning off the alarm clock that it has
become impossible to be concerned about their business ethics without
being associated with something the tacit assumptions elsewhere say
is "impossible" or "impractical" or "insane", like high quality
software, communism, or anti-innovation, RESPECTIVELY.
let me know if you need further improvements in order to understand it.
#:Erik
--
http://www.naggum.no/spam.html is about my spam protection scheme and how
to guarantee that you reach me. in brief: if you reply to a news article
of mine, be sure to include an In-Reply-To or References header with the
message-ID of that message in it. otherwise, you need to read that page.
In article <················@naggum.no>, Erik Naggum <······@naggum.no> wrote:
> become impossible to be concerned about their business ethics without
> being associated with something the tacit assumptions elsewhere say
> is "impossible" or "impractical" or "insane", like high quality
> software, communism, or anti-innovation, RESPECTIVELY.
Oh! destructuring-bind
-dan
····@deltanet.com (Zeno) writes:
> And how they took it over was impressive. Microsoft controlled the PC
> market so well that IBM moved to #10 on the list of top 10 most IBM
> compatible computers and then disappeared off it all together. I do
> not take to task fighting such changes, but rather try to recognize
> them early enough to adapt in time myself.
No, what was "impressive" was not Microsoft way of taking over the
market, but the readiness of the market to be taken over, considering
that they had just had at least 30 years of IBM and centralism, with
everybody cursing and whining like they do now.
And you shouldn't fight the changeover from IBM to Microsoft, but
rather you should try to educate your clients better.
Regs, Pierre.
--
Pierre Mai <····@cs.tu-berlin.de> http://home.pages.de/~trillian/
"Such is life." -- Fiona in "Four Weddings and a Funeral" (UK/1994)
Zeno (····@deltanet.com) asked:
Setting distribution fees (royalties, run-time fees, etc.) aside for the
moment, is it possible to create a small application with Lisp which could
be distributed over the web? Let's say, you wanted to create a small
utility which would search the files on Win95/NT computers and find
duplicates, then provide a list to the user allowing them to delete
duplicate files as they see fit. Can you do this with Lisp? Can you do
this with Allegro Common Lisp? Are small utilities like these necessarily
large programs because of the size of Lisp itself?
If you want small lisp deliverables, I would suggest going with a Scheme
implementation, rather than Common Lisp.
I once had to do a project for an important pro bono client that required
both (a) pretty complex semantics-driven pattern matching and (b) had to run
on (even then) ancient 386/DOS machines with 2 or 4MB ram. I really needed
a lisp to have any hope of being able to accomplish the functional goals,
and it needed to be small (and also free). I fairly quickly found
MIT-Scheme (aka CScheme) which fit my needs perfectly. My deliverable was
an "executable" (saved image) which fit on a single floppy, along with an
install script and a README file.
Most Schemes don't have extensive "libraries" built into them as CL does. If
you don't mind the coding work (and style) that comes along with that, then
Scheme is a very nice vehicle for writing and delivering the kind of program
that you are talking about.
Another possibility is to use Common Lisp, but then use a "tree shaker" to
eliminate all of the parts of the image that are unreachable in your final
product. Unfortunately, one needs to adapt coding practices which make the
work of the tree shaker effective, which are sometimes non-obvious. Allegro
comes with a tree shaker, and fairly good documentation on how to code for
it effectively. Personally, I find it easier to start small (Scheme) and
add functionality as I need it, rather than assume the world (CL) and try to
pare it down later.
Larry
--
Lawrence Hunter, PhD.
National Library of Medicine phone: +1 (301) 496-9303
Bldg. 38A, 9th fl, MS-54 fax: +1 (301) 496-0673
Bethesda. MD 20894 USA email: ······@nlm.nih.gov
On 24 Jun 1998 10:38:39 -0400, Larry Hunter <······@nlm.nih.gov> wrote:
[Snip]
>
>Another possibility is to use Common Lisp, but then use a "tree shaker" to
>eliminate all of the parts of the image that are unreachable in your final
>product.
Are there any freely available "tree-shakers"?
Thanks,
--
Kenneth P. Turvey <·······@pug1.SprocketShop.com>
Ninety percent of the politicians give the other ten percent a bad name.
-- Henry Kissinger
"Howard R. Stearns" <······@elwood.com> writes:
> Accordingly, Eclipse:
> + is $500 per machine, regardless of platform.
> + has a text-based interface and generates human readable, lintable C
> code that is compatible with K&R, ANSI C, and C++ compilers.
> + comes with a C-callable library that contains all the ANSI CL
> utilities. This library is identical on all platforms. This can be
> used by Lisp applications, or directly by hand-written C code.
> + does not require any royalties for applications that don't use eval,
> compile and friends.
>
> Nonetheless, we haven't seen the "huge market" described by some. Can
> someone convince me that we did, in fact, view the situation correctly
> and should not instead charge much more money and royalties?
Although I have argued that ACL's price is not that much of a
problem, I'm equally prepared to try to convince you, that you
should continue ECL's pricing policy. IMHO there are projects and
customers who need the features and pricing ECL is offering, and
ECL _is_ IMHO a very promising[1] product, even for those who
don't need some of the capabilities ECL is sporting, like e.g. me.
So you should IMHO see at least a decent market. As to the "huge
market" seen by some -- I sometimes think that this is just another
instance of the following line of reasoning:
A: Lisp is cool, great whatever, _BUT_ I can't use it, because it
doesn't do X.
B: Ok, here you have an implementation which either does X out of the
box, or can easily be extended to do X.
A: Well yes, but it doesn't provide feature Y...
B: Ok, here you have ....
A: Well yes, but it doesn't interoperate with Tool/BS Z...
B: Ok, but if you do this, you can do ....
A: Ok, Ok, but CL costs way too much...
....
I.e. no matter what features you provide, and what pricing policy you
adopt, there will always be "just another something" that's still
missing, and if you provide that, too, there will be a whole mass of
customers of product XYZ who will be jumping up and down to change
over to your product...
Since ECL has already provided much of the wishlist, the customers
might just be waiting for you to provide a flashy IDE or
(cross-plattform) GUI package. I don't see the problem here, because
using ECL with existing C-based GUIs and even GUI-Builders shouldn't
be much of a problem. Of course CLIM might be nicer from a Lisp POV,
but this is exactly the point: Start using Lisp _now_, and don't wait
for the perfect, ultimate "can-do-everything" tool to appear
magically. Linux wouldn't exist, if more people had thought this way
at the beginning, and most other software, too.
So IMHO ECL is filling an important market place, which maybe has not
yet revealed it's full potential, and it would really be sad, if
Elwood would (have to) change it's pricing strategy.
Regs, Pierre.
Footnotes:
[1] I'd like to say great, or even astonishing here, but I haven't
(yet) had the chance to use ECL in a project, so I'm sadly not in the
position to comment further... ;( But this might change in the near
future... ;)
--
Pierre Mai <····@cs.tu-berlin.de> http://home.pages.de/~trillian/
"Such is life." -- Fiona in "Four Weddings and a Funeral" (UK/1994)
In article <··············@torus.cs.tu-berlin.de>,
Pierre Mai <····@cs.tu-berlin.de> writes:
> "Howard R. Stearns" <······@elwood.com> writes:
>> Nonetheless, we haven't seen the "huge market" described by some. Can
>> someone convince me that we did, in fact, view the situation correctly
>> and should not instead charge much more money and royalties?
>
> Although I have argued that ACL's price is not that much of a
> problem, I'm equally prepared to try to convince you, that you
> should continue ECL's pricing policy. IMHO there are projects and
> customers who need the features and pricing ECL is offering, and
> ECL _is_ IMHO a very promising[1] product, even for those who
> don't need some of the capabilities ECL is sporting, like e.g. me.
Right. If you change to be just like all of the other Lisp vendors, you'll
be just like all the rest. But then why would someone shose your's over the
existing, well established vendors? Be different!
> I.e. no matter what features you provide, and what pricing policy you
> adopt, there will always be "just another something" that's still
> missing, and if you provide that, too, there will be a whole mass of
> customers of product XYZ who will be jumping up and down to change
> over to your product...
It's the customers job to bitch no matter what! And some are REALLY good at
it!
> Start using Lisp _now_, and don't wait
> for the perfect, ultimate "can-do-everything" tool to appear
> magically. Linux wouldn't exist, if more people had thought this way
> at the beginning, and most other software, too.
That's the key thing. Linux already had a source of apps and tools before it
began. Can someone list all of the CL apps that we just can't live without?
Nope, I didn't think so. But there is a bit of the chicken-and-egg syndrome
involved here. You need enough of the system in place to make writing apps
easy and worth while. But you need apps to justify building the system. Take
my favorite waste of time, CLIM. Just imagine that someday I actually get it
working. So what? Can anyone show me a useful app that I could then run? But
without a working CLIM, it's hard for users to write the "killer" app that
justifies the toolkit. And around and around we go!
Mike McDonald
·······@mikemac.com