Hello,
is there any good modern Lisp that works on Unix (MacOSX, Solaris,
Linux) and Windows 2000/XP ? I would also need real operating level
concurrent threading and a good integrated compiler.
I looked a few years ago and there wasn't any thread or windows support
on the better Lisp systems.
>>>>> "llothar" == llothar <·······@web.de> writes:
llothar> Hello, is there any good modern Lisp that works on Unix
llothar> (MacOSX, Solaris, Linux) and Windows 2000/XP ?
Geez, lots. Clisp and Scheme48 come to mind.
--
I hope to get back to work on Arc soon.
-- Paul Graham, quoted in "ACM Crossroads", no date given
Eric Hanchrow wrote:
> Geez, lots. Clisp and Scheme48 come to mind.
CLISP doesn't have thread support and Scheme is not Lisp.
--
Frank Buss, ··@frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de
How about ECL? http://ecls.sourceforge.net
It runs on all the platforms you mention, plus some, and has
mutlithreading support.
http://ecls.sourceforge.net/cgi-bin/view/Main/MultiProcessing
llothar wrote:
> Hello,
>
> is there any good modern Lisp that works on Unix (MacOSX, Solaris,
> Linux) and Windows 2000/XP ? I would also need real operating level
> concurrent threading and a good integrated compiler.
>
> I looked a few years ago and there wasn't any thread or windows support
> on the better Lisp systems.
On 2006-08-02, ··············@gmail.com <··············@gmail.com> wrote:
> How about ECL? http://ecls.sourceforge.net
>
> It runs on all the platforms you mention, plus some, and has
> mutlithreading support.
> http://ecls.sourceforge.net/cgi-bin/view/Main/MultiProcessing
Does it work with Slime?
thnx
tj
--
Tim Johnson <···@johnsons-web.com>
http://www.alaska-internet-solutions.com
Tim> Does it work with Slime? thnx tj
Slime proper works only with Common Lisp, but there's a Slime48 which
works with Scheme48.
--
The whole point of loud music is to make it possible
to date without talking.
-- Roger Ebert
Eric Hanchrow wrote:
> Tim> Does it work with Slime? thnx tj
>
> Slime proper works only with Common Lisp, but there's a Slime48 which
> works with Scheme48.
Errr, since he was asking about ECL, the answer is "yes" since ECL is
Common Lisp.
··············@gmail.com wrote:
> Eric Hanchrow wrote:
> > Tim> Does it work with Slime? thnx tj
> >
> > Slime proper works only with Common Lisp, but there's a Slime48 which
> > works with Scheme48.
>
> Errr, since he was asking about ECL, the answer is "yes" since ECL is
> Common Lisp.
Although I believe that the answer is still "yes", it is not because
ECL is written in Common Lisp. Each Lisp implementation that Slime
supports needs backend code that is specific to that implementation of
CL. Many things that Slime needs to do fall into the void of
implementation-defined behaviour.
Look at Slime's swank-*.lisp files for a list of backends.
Cheers
Brad
On 2006-08-02, bradb <··············@gmail.com> wrote:
> Look at Slime's swank-*.lisp files for a list of backends.
swank-ecl.lisp is there...
Thanks
Good news!
tj
--
Tim Johnson <···@johnsons-web.com>
http://www.alaska-internet-solutions.com
Tim Johnson wrote:
> On 2006-08-02, bradb <··············@gmail.com> wrote:
> > Look at Slime's swank-*.lisp files for a list of backends.
> swank-ecl.lisp is there...
What's the weakness of ECL ??
Currently I am mostly using GCL...
Is ECL better than GCL??
Thanks
············@gmail.com wrote:
> Tim Johnson wrote:
> > On 2006-08-02, bradb <··············@gmail.com> wrote:
> > > Look at Slime's swank-*.lisp files for a list of backends.
> > swank-ecl.lisp is there...
> What's the weakness of ECL ??
>
> Currently I am mostly using GCL...
> Is ECL better than GCL??
>
> Thanks
The way I understand it, almost any other Lisp implementation is better
than GCL. If ECL has any weaknesses, they would be that it is not as
fast as it could be, and that it may not be quite 100% conformant to
ANSI with respect to CLOS and other aspects. Though I think I am being
overly harsh, ECL is really pretty nice - and if you need to embed it
in a C program it is VERY nice. Personally I use SBCL on OS X Intel as
my Lisp of choice.
Cheers
Brad
············@gmail.com wrote:
> Tim Johnson wrote:
> > On 2006-08-02, bradb <··············@gmail.com> wrote:
> > > Look at Slime's swank-*.lisp files for a list of backends.
> > swank-ecl.lisp is there...
> What's the weakness of ECL ??
>
> Currently I am mostly using GCL...
> Is ECL better than GCL??
It is possible to make GCL dump an image that is a single executable.
GCL has a newer compiler and probably produces faster code. GCL starts
faster.
ECL has more up to date docs, is easier to build, emits clearer C code
and can be embedded.
I use ECL currently.
llothar wrote:
> Hello,
>
> is there any good modern Lisp that works on Unix (MacOSX, Solaris,
> Linux) and Windows 2000/XP ? I would also need real operating level
> concurrent threading and a good integrated compiler.
Lispworks and ACL runs on both Windows and Linux. Both have thread- and
GUI(window?)-support. There are trial-versions of these available.
The free SBCL is in the process of being ported to Windows, but still
not usable I think.
I do not know about the Mac- or Solaris-support of any of these.
--
mvh, Lars Rune Nøstdal
http://lars.nostdal.org/
··············@gmail.com wrote:
> Hey, does this mean an ArachnoLisp is around the corner? :)
No i'm just getting tired of Eiffel and look if there is maybe a some
long term substitution.
>From Eiffel to a good Lisp would be an evolution, you can't say this
about many other languages.
I'm also looking at Dylan.
llothar wrote:
> ··············@gmail.com wrote:
> > Hey, does this mean an ArachnoLisp is around the corner? :)
>
> No i'm just getting tired of Eiffel and look if there is maybe a some
> long term substitution.
> From Eiffel to a good Lisp would be an evolution, you can't say this
> about many other languages.
>
> I'm also looking at Dylan.
You might want to take a look at Bigloo.
http://www-sop.inria.fr/mimosa/fp/Bigloo/
It's Scheme, not Common Lisp, but has support for lots of systems and
includes threads. It compiles to native code, the speed of which it's
admirers boast somewhat obnoxiously. :)
··············@gmail.com wrote:
> It's Scheme, not Common Lisp, but has support for lots of systems and
> includes threads. It compiles to native code, the speed of which it's
> admirers boast somewhat obnoxiously. :)
No it does not have native threads, just unuseable cooperate threads.
Seems that there is only the expensive way of LispWorks or ACL (summing
up to 5000 US$). Free Lisp systems still are do not meet my goals.
llothar wrote:
> ··············@gmail.com wrote:
>
> > It's Scheme, not Common Lisp, but has support for lots of systems and
> > includes threads. It compiles to native code, the speed of which it's
> > admirers boast somewhat obnoxiously. :)
>
> No it does not have native threads, just unuseable cooperate threads.
Why does it matter how the threads are implemented internally?
>
> Why does it matter how the threads are implemented internally?
>
Because in a few years all systems have more then one CPU kernel.
I'm not starting a new long term project on a system with this huge
restriction.
llothar wrote:
> >
> > Why does it matter how the threads are implemented internally?
> >
>
> Because in a few years all systems have more then one CPU kernel.
> I'm not starting a new long term project on a system with this huge
> restriction.
Ah, I see you want them for performance.
The current situation regarding threads is a problem. But if your
project is long term you may be able to use SBCL's Windows port when
it's complete.
I would point out that performance of different Lisps varies greatly.
The difference between 1 or 2 fully utilized cores is probably less
than that between different Lisp compilers.
llothar wrote:
> >
> > Why does it matter how the threads are implemented internally?
> >
>
> Because in a few years all systems have more then one CPU kernel.
> I'm not starting a new long term project on a system with this huge
> restriction.
Geez I envy either your coding skills or development resources, if you
can afford to sustain long term projects with "woulda coulda shoulda"
requirements projected several years into the future. This sounds like
seriously overthinking things unless you have some very specific
project in mind with very specific requirements.
Cheers,
Brandon Van Every
> Geez I envy either your coding skills or development resources, if you
Thanks, yes i like to make a 10 year plan when developing software.
I'm already in year 5 and still in late beta. Good vertical market
software always
need 10 years to be called good.
> can afford to sustain long term projects with "woulda coulda shoulda"
> requirements projected several years into the future. This sounds like
> seriously overthinking things unless you have some very specific
> project in mind with very specific requirements.
No. But now the direction in CPU development is so clear to see that
there is any uncertainty about multicore CPU's. And there are solutions
which already support this. So if i intend to do a huge change, why
should i restrict myself so much ?
I wrote in another below that i don't think that CL will boost
productivity over Eiffel (i just hope it increase it).
If i look into the language market i see that this is a huge problem
(and a chance for a newcomer) because no script language (Perl, PHP,
Ruby, Python, TCL, LUA) can handle this well. And also very few
interactive languages/implementation like Lisp, Scheme or Smalltalk.
llothar wrote:
> ··············@gmail.com wrote:
>
> > It's Scheme, not Common Lisp, but has support for lots of systems and
> > includes threads. It compiles to native code, the speed of which it's
> > admirers boast somewhat obnoxiously. :)
>
> No it does not have native threads, just unuseable cooperate threads.
>
> Seems that there is only the expensive way of LispWorks or ACL (summing
> up to 5000 US$). Free Lisp systems still are do not meet my goals.
These commercial implementation might be good, but it's so expensive
that I don't think most of lisp programmers can offer. For a simply
$50-$100 I think it's a lot better. Or maybe there should be some
implementation in the middle...
> > Seems that there is only the expensive way of LispWorks or ACL (summing
> > up to 5000 US$). Free Lisp systems still are do not meet my goals.
> These commercial implementation might be good, but it's so expensive
> that I don't think most of lisp programmers can offer. For a simply
> $50-$100 I think it's a lot better. Or maybe there should be some
> implementation in the middle...
Well this is the bad effect of Free Open-Source.
OSS killed the middle segment of software because the low cost
solutions can't be competetive - even if they are a little bit better
then the OSS. And it left the segment of people who pays for it so
small that other companies simply can't afford to take less money for
the solutions.
OSS everywhere is not a real advantage for the Software World, it
results in much worse but cost-free software. It is not good - but
people don't see it at the moment.
llothar wrote:
> > > Seems that there is only the expensive way of LispWorks or ACL (summing
> > > up to 5000 US$). Free Lisp systems still are do not meet my goals.
> > These commercial implementation might be good, but it's so expensive
> > that I don't think most of lisp programmers can offer. For a simply
> > $50-$100 I think it's a lot better. Or maybe there should be some
> > implementation in the middle...
>
> Well this is the bad effect of Free Open-Source.
>
> OSS killed the middle segment of software because the low cost
> solutions can't be competetive - even if they are a little bit better
> then the OSS. And it left the segment of people who pays for it so
> small that other companies simply can't afford to take less money for
> the solutions.
>
> OSS everywhere is not a real advantage for the Software World, it
> results in much worse but cost-free software. It is not good - but
> people don't see it at the moment.
Isn't there a solution to this? Something like `Lazy Payment' (based on
how `lazy evaluation' in programming works, or somethinglikethat).
The customer wants changes in some software, but this cannot happen
because the software is "free". When it comes to proprietary software,
the users can give feedback to the developers telling them what they
need in the future - and the developers can implement these because all
the customer have already shared the costs (mostly time) of
implementing the current one (+ some).
Only big companies (IBM, Novell, etc.) can afford to hire a developer
full-time in the new "free model", the consumer cannot do this - but I
think many really-really wants this. (I do!)
I think some formal feature, like CVS and bugtracking etc., in
development of free software should exist. So the users can poll by
pushing money on the things they want implemented. Then payment to the
developers are commited when features are completed (lazy payment).
(ok, this was written in a hurry )
--
mvh, Lars Rune Nøstdal
http://lars.nostdal.org/
Lars Rune Nøstdal wrote:
> llothar wrote:
> > > > Seems that there is only the expensive way of LispWorks or ACL (summing
> > > > up to 5000 US$). Free Lisp systems still are do not meet my goals.
> > > These commercial implementation might be good, but it's so expensive
> > > that I don't think most of lisp programmers can offer. For a simply
> > > $50-$100 I think it's a lot better. Or maybe there should be some
> > > implementation in the middle...
> >
> > Well this is the bad effect of Free Open-Source.
> >
> > OSS killed the middle segment of software because the low cost
> > solutions can't be competetive - even if they are a little bit better
> > then the OSS. And it left the segment of people who pays for it so
> > small that other companies simply can't afford to take less money for
> > the solutions.
> >
> > OSS everywhere is not a real advantage for the Software World, it
> > results in much worse but cost-free software. It is not good - but
> > people don't see it at the moment.
>
> Isn't there a solution to this? Something like `Lazy Payment' (based on
> how `lazy evaluation' in programming works, or somethinglikethat).
>
> The customer wants changes in some software, but this cannot happen
> because the software is "free". When it comes to proprietary software,
> the users can give feedback to the developers telling them what they
> need in the future - and the developers can implement these because all
> the customer have already shared the costs (mostly time) of
> implementing the current one (+ some).
>
> Only big companies (IBM, Novell, etc.) can afford to hire a developer
> full-time in the new "free model", the consumer cannot do this - but I
> think many really-really wants this. (I do!)
>
> I think some formal feature, like CVS and bugtracking etc., in
> development of free software should exist. So the users can poll by
> pushing money on the things they want implemented. Then payment to the
> developers are commited when features are completed (lazy payment).
>
> (ok, this was written in a hurry )
>
> --
> mvh, Lars Rune Nøstdal
> http://lars.nostdal.org/
I'd pay a sum each month in support for SBCL and other Lisp-stuff.
Maybe a set of defacto tools; something like this `SBCL Studio'-thing I
saw mentioned a while ago.
* Light threads (Erlang-style!)
* Windows-port (but maybe something like a remote foreign function
interface (to GTK) via sockets would be cooler/better? I dunno)
* Small executable deliverables (tree shaker?)
* Your Idea Here?
How many users would be willing to pay for this - how many are we? :)
--
mvh, Lars Rune Nøstdal
http://lars.nostdal.org/
Lars Rune Nøstdal wrote:
> Lars Rune Nøstdal wrote:
> > llothar wrote:
> > > > > Seems that there is only the expensive way of LispWorks or ACL (summing
> > > > > up to 5000 US$). Free Lisp systems still are do not meet my goals.
> > > > These commercial implementation might be good, but it's so expensive
> > > > that I don't think most of lisp programmers can offer. For a simply
> > > > $50-$100 I think it's a lot better. Or maybe there should be some
> > > > implementation in the middle...
> > >
> > > Well this is the bad effect of Free Open-Source.
> > >
> > > OSS killed the middle segment of software because the low cost
> > > solutions can't be competetive - even if they are a little bit better
> > > then the OSS. And it left the segment of people who pays for it so
> > > small that other companies simply can't afford to take less money for
> > > the solutions.
> > >
> > > OSS everywhere is not a real advantage for the Software World, it
> > > results in much worse but cost-free software. It is not good - but
> > > people don't see it at the moment.
> >
> > Isn't there a solution to this? Something like `Lazy Payment' (based on
> > how `lazy evaluation' in programming works, or somethinglikethat).
> >
> > The customer wants changes in some software, but this cannot happen
> > because the software is "free". When it comes to proprietary software,
> > the users can give feedback to the developers telling them what they
> > need in the future - and the developers can implement these because all
> > the customer have already shared the costs (mostly time) of
> > implementing the current one (+ some).
> >
> > Only big companies (IBM, Novell, etc.) can afford to hire a developer
> > full-time in the new "free model", the consumer cannot do this - but I
> > think many really-really wants this. (I do!)
> >
> > I think some formal feature, like CVS and bugtracking etc., in
> > development of free software should exist. So the users can poll by
> > pushing money on the things they want implemented. Then payment to the
> > developers are commited when features are completed (lazy payment).
> >
> > (ok, this was written in a hurry )
> >
> > --
> > mvh, Lars Rune Nøstdal
> > http://lars.nostdal.org/
>
> I'd pay a sum each month in support for SBCL and other Lisp-stuff.
> Maybe a set of defacto tools; something like this `SBCL Studio'-thing I
> saw mentioned a while ago.
>
> * Light threads (Erlang-style!)
> * Windows-port (but maybe something like a remote foreign function
> interface (to GTK) via sockets would be cooler/better? I dunno)
> * Small executable deliverables (tree shaker?)
> * Your Idea Here?
>
> How many users would be willing to pay for this - how many are we? :)
I like this idea a lot. Personally SBCL is making me no money right
now. If I was making enough money (say ~80K USD, enough that it makes
sense to give up my day job) from SBCL then I would happily chip in a
couple of hundred dollars per month as a "thanks for all the fish"
donation.
Cheers
Brad
> I'd pay a sum each month in support for SBCL and other Lisp-stuff.
> Maybe a set of defacto tools; something like this `SBCL Studio'-thing I
> saw mentioned a while ago.
>
> * Light threads (Erlang-style!)
> * Windows-port (but maybe something like a remote foreign function
> interface (to GTK) via sockets would be cooler/better? I dunno)
> * Small executable deliverables (tree shaker?)
> * Your Idea Here?
What you mentioned + full networking support (tcp/udp/raw socket) and
better doc for extension modules.
> How many users would be willing to pay for this - how many are we? :)
I will.
I'd rather have a single standard compliant, feature rich
implementation for all the platforms rather than openmcl for osx, sbcl
for linux, clisp for win32 (and cmucl for freebsd?).
I wasted so much time figuring out how to use a library on my platform
(because the author use a different lisp on different platform and
almost all of them won't have time to test it on others) that it took
away the fun of using lisp in the first place. (examples : elephant,
ucw, etc.)
Lars Rune N�stdal wrote:
> I'd pay a sum each month in support for SBCL and other Lisp-stuff.
The hypnosis reaches its zenith. Not only are we willing to spend our
"free time" fighting with inferior "free" tools rather than pay $$$ to a
commercial vendor, we are also willing to pay $$$ to avoid paying $$$
to... oh, never mind.
:)
kt
--
Cells: http://common-lisp.net/project/cells/
"I'll say I'm losing my grip, and it feels terrific."
-- Smiling husband to scowling wife, New Yorker cartoon
Ken Tilton wrote:
> Lars Rune Nøstdal wrote:
>
> > I'd pay a sum each month in support for SBCL and other Lisp-stuff.
>
> The hypnosis reaches its zenith. Not only are we willing to spend our
> "free time" fighting with inferior "free" tools rather than pay $$$ to a
> commercial vendor, we are also willing to pay $$$ to avoid paying $$$
> to... oh, never mind.
>
> :)
Yeah - or if we'd pay $$$ for the open tools as we do for our non-open
tools we'd have the benefits of both open tools (fast on-site
bug-fixes, flexibility, "code is the best doc" etc.) and non-open tools
(growth, support, etc.) at the same time.
"Open" is the technical benefit (I really like source code), and "free"
can sometimes be bad (like mentioned in recent threads) - but not
always; hence the combination.
Guess it depends on wheter people _understand_ that they need to feed
their dependencies to see them grow as they do (or need). People will
have to really get this into their heads and this might be hard to do
the way things are, or have been going for while now.
Maybe someone should try this - maybe via some "defacto" public hub
like `common-lisp.org' or something. Get rid of the google-ads on
planet.lisp.org etc. and replace them with "You want feature X in
Lisp-software Y? Donate Money for Beer Now!! \o/"
..I dunno; in my mind something like this should work - but I might be
naive .. damn - but it would be cool if stuff like this really did
work.. :)
--
mvh, Lars Rune Nøstdal
http://lars.nostdal.org/
Ken Tilton <·········@gmail.com> writes:
>
>> I'd pay a sum each month in support for SBCL and other Lisp-stuff.
>
> The hypnosis reaches its zenith. Not only are we willing to spend our
> "free time" fighting with inferior "free" tools rather than pay $$$ to
> a commercial vendor, we are also willing to pay $$$ to avoid paying
> $$$ to... oh, never mind.
But remember, free software is not about price; it's about freedom. And
many of us are willing to pay quite a lot of money to preserve our freedom.
--
Robert Uhl <http://public.xdi.org/=ruhl>
Ken Thompson claims that he started developing Unix so he could play
Space War, but the end product shows he was really much more interested
in cheating at Scrabble.
--Steve VanDevender
Robert Uhl wrote:
> Ken Tilton <·········@gmail.com> writes:
>
>>>I'd pay a sum each month in support for SBCL and other Lisp-stuff.
>>
>>The hypnosis reaches its zenith. Not only are we willing to spend our
>>"free time" fighting with inferior "free" tools rather than pay $$$ to
>>a commercial vendor, we are also willing to pay $$$ to avoid paying
>>$$$ to... oh, never mind.
>
>
> But remember, free software is not about price; it's about freedom. And
> many of us are willing to pay quite a lot of money to preserve our freedom.
Oh, yes, i understand, it is so important to have control over how LOOP
is implemented, and when the black helicopters get a little lower we'll
still have the source for... doh! I am talking to a hypnotic!
kt
--
Cells: http://common-lisp.net/project/cells/
"I'll say I'm losing my grip, and it feels terrific."
-- Smiling husband to scowling wife, New Yorker cartoon
Ken Tilton wrote:
> Robert Uhl wrote:
> > Ken Tilton <·········@gmail.com> writes:
> >
> >>>I'd pay a sum each month in support for SBCL and other Lisp-stuff.
> >>
> >>The hypnosis reaches its zenith. Not only are we willing to spend our
> >>"free time" fighting with inferior "free" tools rather than pay $$$ to
> >>a commercial vendor, we are also willing to pay $$$ to avoid paying
> >>$$$ to... oh, never mind.
> >
> >
> > But remember, free software is not about price; it's about freedom. And
> > many of us are willing to pay quite a lot of money to preserve our freedom.
>
> Oh, yes, i understand, it is so important to have control over how LOOP
> is implemented, and when the black helicopters get a little lower we'll
> still have the source for... doh! I am talking to a hypnotic!
"Give a hand, take an arm"? .. "Inch .. miles"? Do you really want a
closed implementation?
If a closed implementation of Lisp+libraries had a major share of the
market, what would stop the company behind it from doing the
"Microsoft-thing" on us, our products and our market/customers?
The software you have spent years developing could be implemented by
them "in a flash" because they now have all the money and resources
being a monopoly. They own the platform and do not have to bother with
good quality and real growth any longer because they own all our data
in form of closed standards too. On top of this they could steal all
your customers and destroy any potential competition by including your
product for "free"#1 in the default install on every machine ever sold.
This isn't fair play IMHO - and it stops new and what could be better
players from ever entering the market. Why wouldn't this happen
again-and-again with closed software? Is this paranoid thinking? Well,
I might have done the same thing if I was Microsoft or that
Lisp-company. :)
In contrast; this would never happen if the software was open. The
"owners" would be whoever develops and uses it. If someone is doing
something "evil" the project would fork and users would simply stop
using it - like xorg, which was a successful fork from the stalling
xfree.
If an open implementation of Lisp+libraries had a major share of the
market, and they decided to include your software that you have spent
years developing in their default install it wouldn't hurt you at all
because income would follow whoever freely chooses to develop, support
and/or maintain the software - and possibly also whoever supplies data
to/via that software. Everyone (well almost, but who cares about them)
would profit; there would always be jobs because there is always
something to improve - and everyone has equal opportunity. Technically
this would be superior too. The competition would be amongst the best
developers and also suppliers of data; not the best ... marketing,
embrace-and-extend-, lie-machines ...
Working with Microsoft products is not very comfortable IMHO; but they
get away with it because they have (had?) a monopoly. In my eyes this
also hurts progress; maybe even more than the problems with free/open
software now.
..heck; maybe Microsoft even is to blame for what should have been
Lisp-world instead of a C++- and VB-world..
I also think we need to think about timing and "head starts" if we
should judge this fairly. "What is the point in paying for inferiour
open software when you can pay the same price for superiour closed
software?" This argument would not stand if it wasn't for timing,
history and our stupidity.
..maybe I'm describing a world that never exists or could exist; it's
hard to predict what could have been if things where different - but at
least I'm hoping the losers in the future market will be those who do
not dare.. :)
#1: Oh, Microsoft _never_ gives anything away for free; it's all
included in the price-tag hidden somewhere in the equation (someone
mentioned this back here I think).
--
mvh, Lars Rune Nøstdal
http://lars.nostdal.org/
Lars Rune Nøstdal wrote:
> Ken Tilton wrote:
> > Robert Uhl wrote:
> > > Ken Tilton <·········@gmail.com> writes:
> > >
> > >>>I'd pay a sum each month in support for SBCL and other Lisp-stuff.
> > >>
> > >>The hypnosis reaches its zenith. Not only are we willing to spend our
> > >>"free time" fighting with inferior "free" tools rather than pay $$$ to
> > >>a commercial vendor, we are also willing to pay $$$ to avoid paying
> > >>$$$ to... oh, never mind.
> > >
> > >
> > > But remember, free software is not about price; it's about freedom. And
> > > many of us are willing to pay quite a lot of money to preserve our freedom.
> >
> > Oh, yes, i understand, it is so important to have control over how LOOP
> > is implemented, and when the black helicopters get a little lower we'll
> > still have the source for... doh! I am talking to a hypnotic!
>
> "Give a hand, take an arm"? .. "Inch .. miles"? Do you really want a
> closed implementation?
>
> If a closed implementation of Lisp+libraries had a major share of the
> market, what would stop the company behind it from doing the
> "Microsoft-thing" on us, our products and our market/customers?
>
> The software you have spent years developing could be implemented by
> them "in a flash" because they now have all the money and resources
> being a monopoly.
The software your commercial company have spent years developing could
be implemented by FS community "in a flash" because they now have all
the resources and can provide the same software for free.
> They own the platform and do not have to bother with
> good quality and real growth any longer because they own all our data
> in form of closed standards too. On top of this they could steal all
> your customers and destroy any potential competition by including your
> product for "free"#1 in the default install on every machine ever sold.
>
>
> This isn't fair play IMHO - and it stops new and what could be better
> players from ever entering the market. Why wouldn't this happen
> again-and-again with closed software? Is this paranoid thinking? Well,
> I might have done the same thing if I was Microsoft or that
> Lisp-company. :)
>
FS could also stop what could be new and better software developed by a
commercial company too. Think VHS video, free software doesn't have to
be better, it just have to be good enough and free to kill another
better commercial product.
The reason many free software has not killed the commercial counterpart
is because most free software lacks in usability design/user interface
and supports.
> In contrast; this would never happen if the software was open. The
> "owners" would be whoever develops and uses it. If someone is doing
> something "evil" the project would fork and users would simply stop
> using it - like xorg, which was a successful fork from the stalling
> xfree.
>
> If an open implementation of Lisp+libraries had a major share of the
> market, and they decided to include your software that you have spent
> years developing in their default install it wouldn't hurt you at all
> because income would follow whoever freely chooses to develop, support
> and/or maintain the software - and possibly also whoever supplies data
> to/via that software.
Why? What if another commercial company pick up that free install
package and sell supports for the whole package. Who would buy support
from me? -- they will still be using my software, but from what they
see, they are using the whole installation, so the first thing when
they search for "product support for lisp package X" they would not
find me, but that commercial company.
> Everyone (well almost, but who cares about them)
> would profit; there would always be jobs because there is always
> something to improve - and everyone has equal opportunity. Technically
> this would be superior too. The competition would be amongst the best
> developers and also suppliers of data; not the best ... marketing,
> embrace-and-extend-, lie-machines ...
>
> Working with Microsoft products is not very comfortable IMHO; but they
> get away with it because they have (had?) a monopoly. In my eyes this
> also hurts progress; maybe even more than the problems with free/open
> software now.
>
Many Linux tools lacks in most user friendly way to configure the
system -- Yes, I think a configuration dialog or control panel is
better than editing config file by hand -- but they get away with it
partly because they are free. In my eyes this also hurts progress.
> ..heck; maybe Microsoft even is to blame for what should have been
> Lisp-world instead of a C++- and VB-world..
>
> I also think we need to think about timing and "head starts" if we
> should judge this fairly. "What is the point in paying for inferiour
> open software when you can pay the same price for superiour closed
> software?" This argument would not stand if it wasn't for timing,
> history and our stupidity.
>
> ..maybe I'm describing a world that never exists or could exist; it's
> hard to predict what could have been if things where different - but at
> least I'm hoping the losers in the future market will be those who do
> not dare.. :)
>
> #1: Oh, Microsoft _never_ gives anything away for free; it's all
> included in the price-tag hidden somewhere in the equation (someone
> mentioned this back here I think).
>
GPL library is NOT free, it has the cost of ALL your code that link to
it.
Linux is has the cost of wanting more market share, which is no
different than the strategy Microsft use to get Windows on every
computer, only that Windows is easier for most people and Windows is
not free because they are from commercial company.
But in the eyes of user, not developer, hidden cost is well hidden in
both case of Microsoft and Linux.
Only BSD/MIT/Public-domain code is real free.
I'll not response to this thread to start all the Free software debate
again.
> --
> mvh, Lars Rune Nøstdal
> http://lars.nostdal.org/
Pisin Bootvong wrote:
> Lars Rune Nøstdal wrote:
> > Ken Tilton wrote:
> > > Robert Uhl wrote:
> > > > Ken Tilton <·········@gmail.com> writes:
> > > >
> > > >>>I'd pay a sum each month in support for SBCL and other Lisp-stuff.
> > > >>
> > > >>The hypnosis reaches its zenith. Not only are we willing to spend our
> > > >>"free time" fighting with inferior "free" tools rather than pay $$$ to
> > > >>a commercial vendor, we are also willing to pay $$$ to avoid paying
> > > >>$$$ to... oh, never mind.
> > > >
> > > >
> > > > But remember, free software is not about price; it's about freedom. And
> > > > many of us are willing to pay quite a lot of money to preserve our freedom.
> > >
> > > Oh, yes, i understand, it is so important to have control over how LOOP
> > > is implemented, and when the black helicopters get a little lower we'll
> > > still have the source for... doh! I am talking to a hypnotic!
> >
> > "Give a hand, take an arm"? .. "Inch .. miles"? Do you really want a
> > closed implementation?
> >
> > If a closed implementation of Lisp+libraries had a major share of the
> > market, what would stop the company behind it from doing the
> > "Microsoft-thing" on us, our products and our market/customers?
> >
> > The software you have spent years developing could be implemented by
> > them "in a flash" because they now have all the money and resources
> > being a monopoly.
>
> The software your commercial company have spent years developing could
> be implemented by FS community "in a flash" because they now have all
> the resources and can provide the same software for free.
No, the company would never do this because they would build on the
already existing FS instead. Isn't that great? :)
> > They own the platform and do not have to bother with
> > good quality and real growth any longer because they own all our data
> > in form of closed standards too. On top of this they could steal all
> > your customers and destroy any potential competition by including your
> > product for "free"#1 in the default install on every machine ever sold.
> >
> >
> > This isn't fair play IMHO - and it stops new and what could be better
> > players from ever entering the market. Why wouldn't this happen
> > again-and-again with closed software? Is this paranoid thinking? Well,
> > I might have done the same thing if I was Microsoft or that
> > Lisp-company. :)
> >
>
> FS could also stop what could be new and better software developed by a
> commercial company too. Think VHS video, free software doesn't have to
> be better, it just have to be good enough and free to kill another
> better commercial product.
Is this bad? The developers of the CS should instead help work on the
FS when users needs it to be better and are willing to pay to make it
happen faster.
I do not want closed software.
> The reason many free software has not killed the commercial counterpart
> is because most free software lacks in usability design/user interface
> and supports.
It can be solved by making people understand that they should
pay/donate for FS if they want changes; this is the same as good enough
-> better above. I mentioned something about `lazy payment' in a post
further up here.
> > In contrast; this would never happen if the software was open. The
> > "owners" would be whoever develops and uses it. If someone is doing
> > something "evil" the project would fork and users would simply stop
> > using it - like xorg, which was a successful fork from the stalling
> > xfree.
> >
> > If an open implementation of Lisp+libraries had a major share of the
> > market, and they decided to include your software that you have spent
> > years developing in their default install it wouldn't hurt you at all
> > because income would follow whoever freely chooses to develop, support
> > and/or maintain the software - and possibly also whoever supplies data
> > to/via that software.
>
> Why? What if another commercial company pick up that free install
> package and sell supports for the whole package. Who would buy support
> from me? -- they will still be using my software, but from what they
> see, they are using the whole installation, so the first thing when
> they search for "product support for lisp package X" they would not
> find me, but that commercial company.
If the customers can't find you the company using your software will
when they need to - and they most likely will. If not - the best people
and software will win; this is fair IMHO.
> > Everyone (well almost, but who cares about them)
> > would profit; there would always be jobs because there is always
> > something to improve - and everyone has equal opportunity. Technically
> > this would be superior too. The competition would be amongst the best
> > developers and also suppliers of data; not the best ... marketing,
> > embrace-and-extend-, lie-machines ...
> >
> > Working with Microsoft products is not very comfortable IMHO; but they
> > get away with it because they have (had?) a monopoly. In my eyes this
> > also hurts progress; maybe even more than the problems with free/open
> > software now.
> >
>
> Many Linux tools lacks in most user friendly way to configure the
> system -- Yes, I think a configuration dialog or control panel is
> better than editing config file by hand -- but they get away with it
> partly because they are free. In my eyes this also hurts progress.
Progress should be or is based on needs - I do not need nor want a GUI
on my server.
On the FS platform's desktop someone still needs to digg the first
ditch - this is what Novell and Ubuntu is trying I think. They are
trying to digg a ditch and crack a hole in the wall so some of the
market will spill out and generate a flow of need, interest expectation
and progress towards a platform everyone has an equal oppurtunity to
compete freely on.
..and it would not suck, because you and anyone else would be able to
fix bugs; or pay _someone else_ (free competition) to fix bugs.. :)
This isn't hurting progress for anyone I care about; it's improving
progress on the FS desktop at the same time as they are trying to move
away from an OS and a company that is stalling; thus also progress.
Look at IE for instance; no progress whatsoever for years-and-years.
> > #1: Oh, Microsoft _never_ gives anything away for free; it's all
> > included in the price-tag hidden somewhere in the equation (someone
> > mentioned this back here I think).
> >
>
> GPL library is NOT free, it has the cost of ALL your code that link to
> it.
> Linux is has the cost of wanting more market share, which is no
> different than the strategy Microsft use to get Windows on every
> computer, only that Windows is easier for most people and Windows is
> not free because they are from commercial company.
The difference is that with GPL, Linux, FS and Open Standards we always
have a chance if something better than Linux or whatever comes along.
> But in the eyes of user, not developer, hidden cost is well hidden in
> both case of Microsoft and Linux.
True.
--
mvh, Lars Rune Nøstdal
http://lars.nostdal.org/
For free implementations...
Currently there isn't a good one like gcc.
I think SBCL might be good, but it's still under development.
And for the commercial ones,
check the Microsoft's latest Visual Studio 2005
http://msdn.microsoft.com/vstudio/products/compare/default.aspx
the free (express) version is better than the commercial version free
version I think.
With the heap size limitation I think it hurts a lot.
And what I mean about the $50-$100 one, I don't mean the full scale
enterprise version, but there should be something like standard
version, professional version, and something like that in the middle,
which offers a cheaper way to have the non-heap limited ones.
Lars Rune Nøstdal wrote:
> "Give a hand, take an arm"? .. "Inch .. miles"? Do you really want a
> closed implementation?
I'm absolutely happy with one, since you ask.
>
> If a closed implementation of Lisp+libraries had a major share of the
> market, what would stop the company behind it from doing the
> "Microsoft-thing" on us, our products and our market/customers?
>
> The software you have spent years developing could be implemented by
> them "in a flash" because they now have all the money and resources
> being a monopoly. They own the platform and do not have to bother with
> good quality and real growth any longer because they own all our data
> in form of closed standards too. On top of this they could steal all
> your customers and destroy any potential competition by including your
> product for "free"#1 in the default install on every machine ever sold.
And there you are. What you're after is open standards and
anti-monopoly legislation that's enforced. Don't confuse this with
some open-source religious nuttery. Oh, you already have. Oh well,
you'll end up buried in a pit with all the other cultists.
--tim
Tim Bradshaw wrote:
> > If a closed implementation of Lisp+libraries had a major share of the
> > market, what would stop the company behind it from doing the
> > "Microsoft-thing" on us, our products and our market/customers?
> >
> > The software you have spent years developing could be implemented by
> > them "in a flash" because they now have all the money and resources
> > being a monopoly. They own the platform and do not have to bother with
> > good quality and real growth any longer because they own all our data
> > in form of closed standards too. On top of this they could steal all
> > your customers and destroy any potential competition by including your
> > product for "free"#1 in the default install on every machine ever sold.
>
> And there you are. What you're after is open standards and
> anti-monopoly legislation that's enforced.
Yes, that would be nice/better. But will this MS or some other MS do
this "in good time"?
I'd still miss the source code and flexibility surrounding that
however. I guess I am a bit nutty. :)
--
mvh, Lars Rune Nøstdal
http://lars.nostdal.org/
Tim Bradshaw wrote:
>
> And there you are. What you're after is open standards and
> anti-monopoly legislation that's enforced.
In some parts of the world, like the USA, the problem cannot be solved
legally. So we return to technical solutions.
Cheers,
Brandon Van Every
From: Christopher C. Stacy
Subject: Re: Any good free cross Plattform lisps ?
Date:
Message-ID: <yzlejvwlhsj.fsf@OSX663.local>
"Lars Rune N�stdal" <···········@gmail.com> writes:
> If a closed implementation of Lisp+libraries had a major share of the
> market, what would stop the company behind it from doing the
> "Microsoft-thing" on us, our products and our market/customers?
I am not sure what the "Microsoft-thing" is,
but anyway the answer is probably "a contract".
Lars Rune Nøstdal wrote:
>
> ..heck; maybe Microsoft even is to blame for what should have been
> Lisp-world instead of a C++- and VB-world..
My impression is that historically, PCs in Lisp's heydey didn't have
the performance necessary for good garbage collection. And so, C++
being faster, it won out. A decade later, the performance of garbage
collectors is fine and even mainstream, as Java and C# demonstrate.
But C++ got the standards and is deeply entrenched. I suppose a
language needs to be the right solution for its moment in history.
Cheers,
Brandon Van Every
Ken Tilton wrote:
> Robert Uhl wrote:
> > Ken Tilton <·········@gmail.com> writes:
> >
> >>>I'd pay a sum each month in support for SBCL and other Lisp-stuff.
> >>
> >>The hypnosis reaches its zenith. Not only are we willing to spend our
> >>"free time" fighting with inferior "free" tools rather than pay $$$ to
> >>a commercial vendor, we are also willing to pay $$$ to avoid paying
> >>$$$ to... oh, never mind.
> >
> >
> > But remember, free software is not about price; it's about freedom. And
> > many of us are willing to pay quite a lot of money to preserve our freedom.
>
> Oh, yes, i understand, it is so important to have control over how LOOP
> is implemented, and when the black helicopters get a little lower we'll
> still have the source for... doh! I am talking to a hypnotic!
Regardless of the possibility of your vendor stealing your market there
are other problems.
An important problem is: What if they discontinue the product, or go
out of business?
Microsoft have discontinued VB6 despite a huge amount of software being
written using it. Probably third parties will step in and save those
people by provide compatible implementations. But the situation is
often worse for less popular tools.
Open source folks can certainly stop working on their products, and it
can be very bad when they do. But at least if the source code is
available then you can try to fix problems yourself.
This message is in MIME format. The first part should be readable text,
while the remaining parts are likely unreadable without MIME-aware tools.
--8323328-2096602479-1154636804=:24883
Content-Type: TEXT/PLAIN; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: QUOTED-PRINTABLE
On Thu, 3 Aug 2006, Ken Tilton wrote:
> Lars Rune N=F8stdal wrote:
>
>> I'd pay a sum each month in support for SBCL and other Lisp-stuff.
>
> The hypnosis reaches its zenith. Not only are we willing to spend our "fr=
ee=20
> time" fighting with inferior "free" tools rather than pay $$$ to a commer=
cial=20
> vendor, we are also willing to pay $$$ to avoid paying $$$ to... oh, neve=
r=20
> mind.
>
I *want* to refute this straw man, but it wouldn't help, would it? It=20
should be perfectly clear by now that this debate is never going to get=20
anywhere. How about I agree not to fly off the handle every time somebody=
=20
mentions having a problem with commercial software, and you do the same=20
with free software? Wouldn't that be better for everyone?
--8323328-2096602479-1154636804=:24883--
Nathan Baum wrote:
> On Thu, 3 Aug 2006, Ken Tilton wrote:
> > Lars Rune Nøstdal wrote:
> >
> >> I'd pay a sum each month in support for SBCL and other Lisp-stuff.
> >
> > The hypnosis reaches its zenith. Not only are we willing to spend our "free
> > time" fighting with inferior "free" tools rather than pay $$$ to a commercial
> > vendor, we are also willing to pay $$$ to avoid paying $$$ to... oh, never
> > mind.
> >
>
> I *want* to refute this straw man,
So do I :-)
> but it wouldn't help, would it? It
> should be perfectly clear by now that this debate is never going to get
> anywhere. How about I agree not to fly off the handle every time somebody
> mentions having a problem with commercial software, and you do the same
> with free software? Wouldn't that be better for everyone?
Sadly, the religous idiots on both sides seem to have nothing
better to do than throw around already refuted arguments.
Besides, KT is not going to miss an opportunity to re-post
his empty hypothesis :-)
goose,
"Lars Rune N�stdal" <···········@gmail.com> writes:
> I'd pay a sum each month in support for SBCL and other Lisp-stuff.
> Maybe a set of defacto tools; something like this `SBCL Studio'-thing I
> saw mentioned a while ago.
>
> * Light threads (Erlang-style!)
> * Windows-port (but maybe something like a remote foreign function
> interface (to GTK) via sockets would be cooler/better? I dunno)
> * Small executable deliverables (tree shaker?)
> * Your Idea Here?
>
> How many users would be willing to pay for this - how many are we? :)
I would pay to complete threads addition to clisp, or I'll eventually
do it myself...
--
__Pascal Bourguignon__ http://www.informatimago.com/
Our enemies are innovative and resourceful, and so are we. They never
stop thinking about new ways to harm our country and our people, and
neither do we. -- Georges W. Bush
Lars Rune Nøstdal wrote:
>
> The customer wants changes in some software, but this cannot happen
> because the software is "free". When it comes to proprietary software,
> the users can give feedback to the developers telling them what they
> need in the future - and the developers can implement these because all
> the customer have already shared the costs (mostly time) of
> implementing the current one (+ some).
People make feature requests to open source developers all the time,
and they act on them. Granted, the requests are usually made by
developers, not consumers.
Cheers,
Brandon Van Every
From: Ken Tilton
Subject: Re: Any good free cross Plattform lisps ?
Date:
Message-ID: <qIpAg.864$Kj4.855@fe10.lga>
llothar wrote:
>>>Seems that there is only the expensive way of LispWorks or ACL (summing
>>>up to 5000 US$). Free Lisp systems still are do not meet my goals.
>>
>> These commercial implementation might be good, but it's so expensive
>>that I don't think most of lisp programmers can offer. For a simply
>>$50-$100 I think it's a lot better. Or maybe there should be some
>>implementation in the middle...
>
>
> Well this is the bad effect of Free Open-Source.
>
> OSS killed the middle segment of software because the low cost
> solutions can't be competetive - even if they are a little bit better
> then the OSS. And it left the segment of people who pays for it so
> small that other companies simply can't afford to take less money for
> the solutions.
>
> OSS everywhere is not a real advantage for the Software World, it
> results in much worse but cost-free software. It is not good - but
> people don't see it at the moment.
>
I had to check the author to make sure I hadn't written this.
:)
kt
--
Cells: http://common-lisp.net/project/cells/
"I'll say I'm losing my grip, and it feels terrific."
-- Smiling husband to scowling wife, New Yorker cartoon
From: Ken Tilton
Subject: Re: Any good free cross Plattform lisps ?
Date:
Message-ID: <ESpAg.865$Kj4.179@fe10.lga>
llothar wrote:
>>>Seems that there is only the expensive way of LispWorks or ACL (summing
>>>up to 5000 US$). Free Lisp systems still are do not meet my goals.
>>
>> These commercial implementation might be good, but it's so expensive
>>that I don't think most of lisp programmers can offer. For a simply
>>$50-$100 I think it's a lot better. Or maybe there should be some
>>implementation in the middle...
>
>
> Well this is the bad effect of Free Open-Source.
>
> OSS killed the middle segment of software because the low cost
> solutions can't be competetive - even if they are a little bit better
> then the OSS. And it left the segment of people who pays for it so
> small that other companies simply can't afford to take less money for
> the solutions.
I should add that my first win32 Lisp was a 1998 ACL "Personal Edition"
(not to be confused with later Trial/Express editions) $495 or so back
then. That is not $50-100, but then you have the other problem driving
prices up: a miniscule market.
Apparently I could buy Goose's house for what I paid last year for a
couple of years use of ACL pro. Easily worth it, though, and the IDE
gets better all the time.
Thanks, Stallman!
kt
--
Cells: http://common-lisp.net/project/cells/
"I'll say I'm losing my grip, and it feels terrific."
-- Smiling husband to scowling wife, New Yorker cartoon
Ken Tilton wrote:
> [snip]
> I should add that my first win32 Lisp was a 1998 ACL "Personal Edition"
> (not to be confused with later Trial/Express editions) $495 or so back
> then. That is not $50-100, but then you have the other problem driving
> prices up: a miniscule market.
>
> [snip]
>
> Thanks, Stallman!
>
I don't want to start a religious/political debate here, but with a
miniscule market like Lisp I would think that Open Source Software
would actually help create a larger market, and in turn increase profit
for good commercial offerings. I mean, if we're talking market, let's
talk market and skip the politics for now.
Look, it's like this. There is a whole category of people who would try
most anything for free, but wouldn't pay a dime for it if told they had
to. If their only choice is to buy it or not, they just won't. Period.
They even shun trial versions because they know they aren't getting the
"full deal". These people would never buy one of the commercial
products anyway, so for this group of people OSS is not taking anything
away from the vendors. But if even one of these people gets so hooked
on how fantastic a language like Lisp is that they decide they can no
longer live without something the commercial people offer, then they
just might pay, whereas without OSS they would have never even
considered it.
And there are also some people who MIGHT pay good money for something.
OSS doesn't take much away from the vendors here either, because that
word "MIGHT" simply means they need convincing, i.e. some compelling
reason to purchase. I'm one of these people. I have paid for a
commercial product after using a couple of OSS alternatives for several
years and then wanting the expanded support and features of the
commercial implementation.
The only thing vendors have to fear from OSS is that the OSS
implementations will match all the features of the commercial product.
But of course the vendors can still offer first-rate commercial
support, which the OSS versions can't (community support
notwithstanding), so even here there is still a market.
Users don't have anything to fear from OSS either, because if they
don't want crap software (I'm not saying all OSS is crap -- it
obviously isn't -- but that's the argument here) they can always pay
for something decent. If they are content to live with crap, then
that's their choice.
And of course, my opinion is worth what you just paid to read it. :)
··············@gmail.com wrote:
> The only thing vendors have to fear from OSS is that the OSS
> implementations will match all the features of the commercial product.
Not really. If you had the choice of flying across-country free in coach
class, or pay for a first class flight, which would you do?
A commercial product has to not only have more features, but enough
*extra* features to make the price of developing *all* the features
attactive to buyers. If it costs you $10K to develop the compiler and
$1K to develop the optimizer that the OSS version doesn't have, you have
to recover $11K on a product that arguably provides $1K of value beyond
the OSS version, and hope that nobody starts giving away that other $1K
part.
--
Darren New / San Diego, CA, USA (PST)
This octopus isn't tasty. Too many
tentacles, not enough chops.
Darren New wrote:
> A commercial product has to not only have more features, but enough
> *extra* features to make the price of developing *all* the features
> attactive to buyers. If it costs you $10K to develop the compiler and
> $1K to develop the optimizer that the OSS version doesn't have, you have
> to recover $11K on a product that arguably provides $1K of value beyond
> the OSS version, and hope that nobody starts giving away that other $1K
> part.
>
You're making a couple of invalid assumptions here. Assumption 1) that
the commercial compiler's closed source code is available to be given
away; and assumption 2) that OSS makes any difference at all in your
scenario. The same situation would be true if you were competing in
strictly a commercial market. If it costs $11K to develop an optimizing
compiler with which you are trying to compete against your commercial
competitor's non-optimizing compiler, you are only offering an extra
10% (say) in value over your competitor's product for which your
potential customer has already paid, but you are still asking them to
pay full price on your product (discount incentives aside).
But guess what? This is how a competitive marketplace works, and no one
complained before OSS.
··············@gmail.com wrote:
> You're making a couple of invalid assumptions here.
No I'm not.
> Assumption 1) that
> the commercial compiler's closed source code is available to be given
> away;
I have no idea why you think I think a commercial compiler's close
source is available.
> and assumption 2) that OSS makes any difference at all in your
> scenario.
Well, since that's what I was addressing, I'm not sure where you're
coming from either. Of course, it applies to any software given away
freely, whether open source or not.
> The same situation would be true if you were competing in
> strictly a commercial market. If it costs $11K to develop an optimizing
> compiler with which you are trying to compete against your commercial
> competitor's non-optimizing compiler, you are only offering an extra
> 10% (say) in value over your competitor's product for which your
> potential customer has already paid, but you are still asking them to
> pay full price on your product (discount incentives aside).
Yep. But for someone who hasn't already purchased one, it's a choice
between $10K for the non-optimizer and $11K for the optimizing compiler.
So as long as I can profit with the optimizer costing $1K, I'm good.
So the statement implying that you only need to provide features in your
product that the free products don't have in order to compete is simply
wrong. That's what I was addressing. Any other side conversations are
something different.
> But guess what? This is how a competitive marketplace works, and no one
> complained before OSS.
Of course they did. Microsoft got sued repeatedly for giving away free
software, because people who made products slightly better couldn't sell
them. It's why governments implement tarrifs against countries that are
dumping.
--
Darren New / San Diego, CA, USA (PST)
This octopus isn't tasty. Too many
tentacles, not enough chops.
Darren New wrote:
> I have no idea why you think I think a commercial compiler's close
> source is available.
Did you not say in your previous post "... you have to recover $11K on
a product that arguably provides $1K of value beyond the OSS version,
and hope that nobody starts giving away that other $1K part"?
Surely you didn't mean giving away the executable, or do you want to
through pirating into the mix as well?
[ignoring other statements which just repeat the issue]
> So the statement implying that you only need to provide features in your
> product that the free products don't have in order to compete is simply
> wrong.
That's not what I said.
> > But guess what? This is how a competitive marketplace works, and no one
> > complained before OSS.
>
> Of course they did. Microsoft got sued repeatedly for giving away free
> software, because people who made products slightly better couldn't sell
> them. It's why governments implement tarrifs against countries that are
> dumping.
>
Wow. You so completely miss my point that I must reexamine my ability
to communicate. I never said there couldn't be abuse in the market.
What I said was that OSS could not hurt the market, at list a niche
market like Lisp. I have no idea why Antitrust law should be raised in
this context.
··············@gmail.com wrote:
> Darren New wrote:
>
>>I have no idea why you think I think a commercial compiler's close
>>source is available.
>
>
> Did you not say in your previous post "... you have to recover $11K on
> a product that arguably provides $1K of value beyond the OSS version,
> and hope that nobody starts giving away that other $1K part"?
Yes.
> Surely you didn't mean giving away the executable, or do you want to
> through pirating into the mix as well?
No. The reading comprehension skills of some of the folks here are
amazingly bad.
Obviously, they wouldn't be giving away *your* part, but rather
implementing something that makes it not worth that $1K to buy the
commercial product. I.e., by adding an optimizer to the free compiler
that takes away your competitive advantage.
>>So the statement implying that you only need to provide features in your
>>product that the free products don't have in order to compete is simply
>>wrong.
>
> That's not what I said.
You said
"""
The only thing vendors have to fear from OSS is that the OSS
implementations will match all the features of the commercial product.
"""
How does what you said differ from what I said? You said commercial
vendors have only to fear an OSS implementation with the same features.
I said you implied commercial vendors only need features the OSS
products don't have.
In any case, what you said isn't true. Vendors have to fear OSS
implementations will match *enough* of the features of the commercial
product that the vendors can't compete. That's my point. It's a simple
matter of calculating the costs versus benefits, which nobody seems to
be paying any attention to here.
An airline that only offered first-class travel would lose to an airline
that offered free coach class, even though there are features you get
with first-class travel that you don't get on coach class. Or do you
think that by offering free booze, I could get enough people flying my
airline that I could compete with free flights?
>>>But guess what? This is how a competitive marketplace works, and no one
>>>complained before OSS.
>>
>>Of course they did. Microsoft got sued repeatedly for giving away free
>>software, because people who made products slightly better couldn't sell
>>them. It's why governments implement tarrifs against countries that are
>>dumping.
>
> Wow. You so completely miss my point that I must reexamine my ability
> to communicate. I never said there couldn't be abuse in the market.
Why is it abusive to the market for Microsoft to give away IIS, but not
the Apache foundation giving away Apache or RedHat giving away Apache?
Illegal, in this case, yes, but why "abusive"? Who were they abusing by
giving away IIS, and why isn't RedHat abusing the same people?
> What I said was that OSS could not hurt the market,
And I'm disagreeing. It can hurt the market in *exactly* the same way
that Microsoft giving away components hurt the market. It just happens
to not be illegal because there's nobody with any money to sue.
Heck, everyone cheers when OSS hurts Microsoft in the marketplace.
--
Darren New / San Diego, CA, USA (PST)
This octopus isn't tasty. Too many
tentacles, not enough chops.
Darren New wrote:
> > Surely you didn't mean giving away the executable, or do you want to
> > through pirating into the mix as well?
>
> No. The reading comprehension skills of some of the folks here are
> amazingly bad.
Now, now..
>
> Obviously, they wouldn't be giving away *your* part, but rather
> implementing something that makes it not worth that $1K to buy the
> commercial product. I.e., by adding an optimizer to the free compiler
> that takes away your competitive advantage.
Gee, then you didn't have to invest $10K into making "your" compiler in
the first place since you were using someone else's work. I thought we
were talking about commercial products versus open source, not
open-source-and-adding-commercial-features vs open source. I guess the
track changed when I wasn't looking.
> >>So the statement implying that you only need to provide features in your
> >>product that the free products don't have in order to compete is simply
> >>wrong.
> >
> > That's not what I said.
>
> You said
> """
> The only thing vendors have to fear from OSS is that the OSS
> implementations will match all the features of the commercial product.
> """
And that's not what you said that I said (say that 3 times real fast).
My point here is that commercial vendors will simply not have a
SOFTWARE product to sell if the open source products offer exactly the
same features that theirs do. I never said they couldn't compete, and
in fact implied that they still could by offering professional support.
It is still a viable option to sell a license for $3000 which includes
hand-holding support even though the software itself offers no features
more than it's open source counterpart.
> In any case, what you said isn't true. Vendors have to fear OSS
> implementations will match *enough* of the features of the commercial
> product that the vendors can't compete. That's my point. It's a simple
> matter of calculating the costs versus benefits, which nobody seems to
> be paying any attention to here.
See my statement above.
> An airline that only offered first-class travel would lose to an airline
> that offered free coach class, even though there are features you get
> with first-class travel that you don't get on coach class. Or do you
> think that by offering free booze, I could get enough people flying my
> airline that I could compete with free flights?
Tell that to AdaCore. http://www.adacore.com
> And I'm disagreeing.
Okay, we'll agree to disagree. Deal? :)
··············@gmail.com wrote:
>>Obviously, they wouldn't be giving away *your* part, but rather
>>implementing something that makes it not worth that $1K to buy the
>>commercial product. I.e., by adding an optimizer to the free compiler
>>that takes away your competitive advantage.
>
> Gee, then you didn't have to invest $10K into making "your" compiler in
> the first place since you were using someone else's work.
I can see speaking in general terms is a little too difficult.
Postulate two items:
1) A GPL compiler that does a really bad job of optimization. (True,
earlier I referred to OSS, and a BSD-licensed compiler wouldn't cause
this problem, so I was a bit sloppy there.)
2) A commercial compiler that does a really good job of optimization.
In order for (2) to be produced, the commercial company has to actually
sell it for more money than it costs to create it. That's what
"commercial" means. But since (1) is free, nobody will buy (2) unless
the optimization is worth the entire price being charged for (2). But
since (1) is GPLed, the commercial company can't simply build on (1) and
add the optimization and sell it. Hence, the price charged for (2) has
to include both the cost of developing the optimizer and the cost of
developing the features that are the same as (1) provides.
> And that's not what you said that I said (say that 3 times real fast).
> My point here is that commercial vendors will simply not have a
> SOFTWARE product to sell if the open source products offer exactly the
> same features that theirs do.
You're still missing the point. They won't have a SOFTWARE product to
sell if they have a whole *bunch* of features that the free software
has, if those *additional* features aren't worth the price being charged.
> See my statement above.
I'm going to stop now, because you're ignoring what I'm trying to say,
either intentionally or out of inability to comprehend simple English.
> Tell that to AdaCore. http://www.adacore.com
Yes. So? They're selling the free software, support, and (I'm guessing)
non-GPLed development environments and libraries, which isn't what I'm
talking about. The entire price someone pays is for the development
environment, not the compiler. And this is a perfect case of where the
"features" the provide are *far* beyond that which the free compiler
provides.
I'm not saying commercial programs can't compete. I'm saying you're
wrong in stating that commercial vendors only have to worry about OSS
projects that match their capabilities.
>>And I'm disagreeing.
> Okay, we'll agree to disagree. Deal? :)
Heh.
And, by the way, I still wonder why you think Microsoft is abusive to
the market and Red Hat isn't.
--
Darren New / San Diego, CA, USA (PST)
This octopus isn't tasty. Too many
tentacles, not enough chops.
Since we're going back and forth a little too much here, I'll just
comment on one thing and still agree to disagree about the other stuff
(comments about my inability to comprehend English notwithstanding).
Darren New wrote:
> > Tell that to AdaCore. http://www.adacore.com
>
> Yes. So? They're selling the free software, support, and (I'm guessing)
> non-GPLed development environments and libraries, which isn't what I'm
> talking about. The entire price someone pays is for the development
> environment, not the compiler. And this is a perfect case of where the
> "features" the provide are *far* beyond that which the free compiler
> provides.
>
This simply is not the case. All of their software is GPLed, not just
the compiler. The ide is called GPS and is also available free. The
neat thing is that contribute GREATLY to the GPL source, and the Ada
community loves them for that. What they figured out to do (and there
admittedly has been some discussion about whether this complies with
the GPL) is to release the source of the last version when the new
version comes out.
On Fri, 4 Aug 2006, Darren New wrote:
> ··············@gmail.com wrote:
>> Darren New wrote:
>>
>> > I have no idea why you think I think a commercial compiler's close
>> > source is available.
>>
>>
>> Did you not say in your previous post "... you have to recover $11K on
>> a product that arguably provides $1K of value beyond the OSS version,
>> and hope that nobody starts giving away that other $1K part"?
>
> Yes.
>
>> Surely you didn't mean giving away the executable, or do you want to
>> through pirating into the mix as well?
>
> No. The reading comprehension skills of some of the folks here are amazingly
> bad.
>
> Obviously, they wouldn't be giving away *your* part, but rather implementing
> something that makes it not worth that $1K to buy the commercial product.
> I.e., by adding an optimizer to the free compiler that takes away your
> competitive advantage.
>
>> > So the statement implying that you only need to provide features in your
>> > product that the free products don't have in order to compete is simply
>> > wrong.
>>
>> That's not what I said.
>
> You said
> """
> The only thing vendors have to fear from OSS is that the OSS
> implementations will match all the features of the commercial product.
> """
>
> How does what you said differ from what I said? You said commercial vendors
> have only to fear an OSS implementation with the same features. I said you
> implied commercial vendors only need features the OSS products don't have.
>
> In any case, what you said isn't true. Vendors have to fear OSS
> implementations will match *enough* of the features of the commercial product
> that the vendors can't compete. That's my point. It's a simple matter of
> calculating the costs versus benefits, which nobody seems to be paying any
> attention to here.
>
> An airline that only offered first-class travel would lose to an airline that
> offered free coach class, even though there are features you get with
> first-class travel that you don't get on coach class. Or do you think that by
> offering free booze, I could get enough people flying my airline that I could
> compete with free flights?
Depends how much free booze I get. And how good it is. We're talking
about vodka, right?
Mmm. Vodka.
>> > > But guess what? This is how a competitive marketplace works, and no one
>> > > complained before OSS.
>> >
>> > Of course they did. Microsoft got sued repeatedly for giving away free
>> > software, because people who made products slightly better couldn't sell
>> > them. It's why governments implement tarrifs against countries that are
>> > dumping.
>>
>> Wow. You so completely miss my point that I must reexamine my ability
>> to communicate. I never said there couldn't be abuse in the market.
>
> Why is it abusive to the market for Microsoft to give away IIS, but not the
> Apache foundation giving away Apache or RedHat giving away Apache? Illegal,
> in this case, yes, but why "abusive"? Who were they abusing by giving away
> IIS, and why isn't RedHat abusing the same people?
>
>> What I said was that OSS could not hurt the market,
>
> And I'm disagreeing. It can hurt the market in *exactly* the same way that
> Microsoft giving away components hurt the market. It just happens to not be
> illegal because there's nobody with any money to sue.
>
> Heck, everyone cheers when OSS hurts Microsoft in the marketplace.
>
> --
> Darren New / San Diego, CA, USA (PST)
> This octopus isn't tasty. Too many
> tentacles, not enough chops.
>
Darren New wrote:
<snipped>
>
> Why is it abusive to the market for Microsoft to give away IIS, but not
> the Apache foundation giving away Apache or RedHat giving away Apache?
Because it is illegal to use a monopoly to leverage
a product into a different market.
> Illegal, in this case, yes, but why "abusive"? Who were they abusing by
> giving away IIS, and why isn't RedHat abusing the same people?
One of those two companies is using a monopoly in a
certain market segment to leverage a product into
a completely different market while the other is not.
I leave it to you to figure out which is which.
>
> > What I said was that OSS could not hurt the market,
>
> And I'm disagreeing. It can hurt the market in *exactly* the same way
> that Microsoft giving away components hurt the market.
The reason that it is illegal to abuse a monopoly
position is the very same reason that it is illegal
to shoot at someone with a loaded firearm; someone could
get hurt. OTOH, just pointing you finger and saying
"BANG" won't cause substantial damage.
> It just happens
> to not be illegal because there's nobody with any money to sue.
No, its not illegal for very different reasons; consult
your laywer and ask about anti-racketeering laws,
competition and trade laws, anti-trust laws, etc.
It was illegal long before OSS (or even computers)
made an appearance; now is exactly the wrong time
to whine about it - you've had around 100 years to
complain about those laws.
goose,
··············@gmail.com wrote:
>
> But guess what? This is how a competitive marketplace works, and no one
> complained before OSS.
RMS certainly did. IIRC his politics formed when companies started
charging for software, when previously it was included free with the
hardware.
Cheers,
Brandon Van Every
From: Christopher C. Stacy
Subject: Re: Any good free cross Plattform lisps ?
Date:
Message-ID: <yzllkq2trp3.fsf@OSX663.local>
"Mallor" <···········@gmail.com> writes:
> ··············@gmail.com wrote:
>> But guess what? This is how a competitive marketplace works,
>> and no one complained before OSS.
>
> RMS certainly did. IIRC his politics formed when companies started
> charging for software, when previously it was included free with the
> hardware.
1. What historical event are you referring to?
2. Why did you change your Usenet posting name?
Christopher C. Stacy wrote:
> "Mallor" <···········@gmail.com> writes:
> > ··············@gmail.com wrote:
> >> But guess what? This is how a competitive marketplace works,
> >> and no one complained before OSS.
> >
> > RMS certainly did. IIRC his politics formed when companies started
> > charging for software, when previously it was included free with the
> > hardware.
>
> 1. What historical event are you referring to?
He should be refering to the printer incident where the printer
software (which used to be free) became closed source (un-free) and
this posed problems as they had to wait for a tech.support monkey to
fix any problems they had.
But what he's refering to is probably the fact that LISP companies who
were in competition screwed over the users of their software as they
had to choose one side (Symbolics) or another (Lisp Machines Inc.) and
thus Stallman became more entrenched in the idea that openness, sharing
of code, etc. is a good thing for everyone. But then again, I could be
wrong.
-Rudolf
From: Christopher C. Stacy
Subject: Re: Any good free cross Plattform lisps ?
Date:
Message-ID: <yzl1wrtdnmx.fsf@OSX663.local>
Nathan Baum <···········@btinternet.com> writes:
> You were present at all the events of Richard's life so far?
I don't know why you would ask that absurd question or why you jumped
into the discussion, so maybe we're talking about two different things.
Or maybe you're just some weird asshole, or a Free Software religous nut.
I don't really care, except that I don't have the time to waste on idiots
and trolls on Usenet newsgroups, so to prevent that, I'll make my complete
response here in this one message, and then stop reading this thread.
I wouldn't want you, in your wildest imagination, to think that I am
engaging you in a conversation, because I don't really give a flying
fuck what most people (certainly not you) think about it. But I will
just put a few of my thoughts on the record so that when other people
come along someday and read the archives, they can have at least a
little bit of firsthand information with which to better evaluate
things themselves, if they're interested.
The assertion raised (by "Captain Kirk") is that the marketplace works
the same as it always did before Free Software (or more generally,
"open source software", although people don't seem to be agree on
exactly what those terms mean). I doubt that's true, but without
really getting deep into that debate, I am addressing just two
statements that were made in response.
Brandon Every ("Malllor") wrote that RMS complained about how the
software marketplace work prior free software:
> RMS certainly did. IIRC his politics formed when companies started charging
> for software, when previously it was included free with the hardware.
First of all, that's a puzzling statement, because "free with the
hardware" means that you are most certainly being charged for it.
The cost of the hardware includes the cost of the bundled software.
Not all companies bundled the software, though: sometimes software
was charged for as a line-item. (And of course there was also
companies that simply licensed software, just like today.)
I do not recognize from my life experience some golden era where
companies gave away their software. The (unspecified) "companies"
did charge money for their software. And (I assume this is relevent)
for the most part, it was all proprietary software for which you
usually didn't get the source. Also, note that most software of
te sort we're talking about was not portable and was tied to the
specific hardware. And if it was portable software, you were
only licensed to run it on the hardware you bought it with.
The part about RMS getting upset about that situation was also
puzzling to me because I was at the AI Lab when RMS was there,
and I don't at all recognize the situation being described.
Since the situation didn't exist, it's hard to recall anyone getting
upset about it. So I asked what was being referred to, and Rudolf
"OMouse" jumped in and clarified that it was about the Lisp Machines:
> the fact that LISP companies who were in competition screwed over
> the users of their software as they had to choose one side
> (Symbolics) or another (Lisp Machines Inc.) and thus Stallman
> became more entrenched in the idea that openness, sharing of code,
> etc. is a good thing for everyone.
I was at the AI Lab as a Lisp Machine developer, and then later
also as a Symbolics customer there. Then after I left MIT,
at various places I was just a customer of the Lisp companies.
And later I also worked at various times for both companies
(mostly Symbolics, and also briefly for a resurrection of the
failed LMI which was being run by Richard Greenblatt,
who I originally worked for at MIT).
At no time did it seem to me like Symbolics (which RMS paints
as his enemy, and Evil) was "screwing over their users".
I decline to comment specifically here on the complimentary
aspect of the story by characterizing the actions that RMS took.
None of the above statements by "Captain Kirk" or Brandon Every make
much sense to me. The argument seems to be mixing up the ideas of
open source (that is, customers can see the source code and modify it,
maybe for limited purposes), no-cost ("free beer") software,
and GPL-type-restricted software. The reason I got into the
discussion was to answer as to the specific historical facts
about whether companies didn't charge for software and whether
the Lisp Machine companies screwed over their users.
Nathan Baum <···········@btinternet.com> writes:
> Symbolics did not, in fact, take code which RMS and others had written
> in MIT, make improvements and deliver them back to MIT under a
> proprietary license which prohibited sharing and further improvements?
Symbolics was a company that consisted of most of the hackers who
had written some software as employees of MIT. MIT sold Symbolics
a license that allowed them to use that very software and to make
proprietary modifications to it. At first, Symbolics shared back
their modifications with MIT. But after a while, due to certain
incidents with RMS, they decided that they could no longer do that.
So they continued to provide the source code to MIT (and all their
other customers, as they had always done), but no longer allowed
it to be "shared" (that is, given to their competitors by RMS).
Customers always had all the source code and were free to modify it
for their own purposes for use on those machines. (In other words,
you bought some hardware, and you also bought some software that was
"open" but legally tied to the machines.)
An analogy in today's world would be like Microsoft and Intel getting
together so that the entire source code for Windows and all the MS
applications (Office, Media Player, Studio, internal development tools,
etc.) would be provided for "free" with every licensed Wintel machine.
You would be free to modify all that software to your heart's content,
but you're not supposed to run it on AMD (absent some other deal),
and you can only share your modified versions of the Microsoft
software with other Microsoft customers. Or similarly, like Apple
doing what it does today, except they give you all their source code.
The policy that the Lisp Machine companies adopted was much more
liberal and "open" than what other companies allowed back then or today,
and it was geared towards giving the customers greater control over
their systems. It's one of the big reasons why users loved it.
Nathan Baum <···········@btinternet.com> writes:
>>>> Which universe did that happen in?
>>>
>>> This one. You seem to be implying it didn't, which is odd considering
>>> the facts are widely available from a number of sources, just a few
>>> keystrokes away from Google. Why the doubt?
> We can conclude that everyone but you is part of some grand conspiracy?
I don't know if there is any conspiracy or how grand it might be,
but there has only really been one version of the history told,
almost exclusively from the standpoint of one individual person,
and it's been refined and retold and repeated all over the place.
The people responsible for that version and who rabidly repeat it
will admit right up front that it advances their adgenda (which
they refer to as a "movement" or a "religion").
I would leave it up to individuals as to decide whether
to believe everything they read on the Internet.
Meanwhile, I think it's time for another comp.lang.lisp vacation for me!
Christopher C. Stacy <······@news.dtpq.com> wrote:
+---------------
| [Symbolics's] Customers always had all the source code and were
| free to modify it for their own purposes for use on those machines.
| (In other words, you bought some hardware, and you also bought some
| software that was "open" but legally tied to the machines.)
+---------------
For another historical data point, note that exactly the same
thing was true for the PDP-10 Monitor & CUSPS [kernel & utilities]
up through [IIRC] at least 5-Series TOPS-10 [mid-1970's?], though
one exception was the new Fortran-10 compiler [written in BLISS-10].
[The older F40 had been "open" like the rest of the code.]
[I mention this only because the PDP-10 was widely used for Lisp.]
+---------------
| An analogy in today's world would be like Microsoft and Intel getting
| together so that the entire source code for Windows and all the MS
| applications (Office, Media Player, Studio, internal development tools,
| etc.) would be provided for "free" with every licensed Wintel machine.
| You would be free to modify all that software to your heart's content,
| but you're not supposed to run it on AMD (absent some other deal),
| and you can only share your modified versions of the Microsoft
| software with other Microsoft customers.
+---------------
Yup! That was exactly the case with TOPS-10...
-Rob
-----
Rob Warnock <····@rpw3.org>
627 26th Avenue <URL:http://rpw3.org/>
San Mateo, CA 94403 (650)572-2607
On Mon, 7 Aug 2006, Christopher C. Stacy wrote:
> Nathan Baum <···········@btinternet.com> writes:
>> You were present at all the events of Richard's life so far?
>
> I don't know why you would ask that absurd question
It was a rhetorical question, you see. The implication is that simply
"being at the AI lab" does not equate "knowing everything that ever
happened to RMS in the AI lab."
> or why you jumped into the discussion,
This is Usenet. If you want a discussion in which only those people you
invite can participate, you're using the wrong forum.
> I do not recognize from my life experience some golden era where
> companies gave away their software.
And that clearly means it never happened. Anywhere.
Curiously, it certainly appears to be RMS's belief that he had come to
expect source to be available, that he could make improvements, and that
he could distribute these improvements to other users of the software.
Unless RMS is clinically insane, the only reason I can think of for him to
have that expectation is that, in RMS's experience at least, that state of
affairs was the norm.
You appear to be purposely misrepresenting the philosophy of the Free
Software movement. You're getting caught up on how much *money* the
software cost, but that isn't the kind of freedom which RMS was upset at
losing.
RMS himself made money (although not much, I think) from selling copies of
Emacs, so we can conclude that either he had no problem with companies
*charging* for copies of software, or he's is a hypocritical bastard who
would abandon his principles on a whim for a quick buck. Although I get
the impression you'd be quite willing to believe the latter. For me, it
seems that reality paints a different picture.
> Symbolics was a company that consisted of most of the hackers who
> had written some software as employees of MIT. MIT sold Symbolics
> a license that allowed them to use that very software and to make
> proprietary modifications to it. At first, Symbolics shared back
> their modifications with MIT. But after a while, due to certain
> incidents with RMS, they decided that they could no longer do that.
> So they continued to provide the source code to MIT (and all their
> other customers, as they had always done), but no longer allowed
> it to be "shared" (that is, given to their competitors by RMS).
Which is to say: the exact events which you claim never occured... Did in
fact occur.
> I don't know if there is any conspiracy or how grand it might be, but
> there has only really been one version of the history told, almost
> exclusively from the standpoint of one individual person, and it's been
> refined and retold and repeated all over the place. The people
> responsible for that version and who rabidly repeat it will admit right
> up front that it advances their adgenda (which they refer to as a
> "movement" or a "religion").
Please show me a free software advocate who describes free software as a
religion. It is my understanding that only opponents of free software
claim that free software is a religion.
> Meanwhile, I think it's time for another comp.lang.lisp vacation for me!
Hit and run? And you're calling *me* a troll?
Christopher C. Stacy wrote:
>
> 2. Why did you change your Usenet posting name?
"Mallor" is a posting nickname that I have used at various times in
various places. I used to use it when sowing the seeds of discord, as
in a quite unfinished mythology I've worked on for a number of years,
that's what Mallor does. But somewhere along the way when configuring
Google Groups I stuck it in there, and I have complete disinterest in
figuring out how to change it again. "Mallor" doesn't mean much of
anything for public posting purposes as of late, except that I think
it's a cooler name than my own. Which always appears anyways at the
end of my posts, so there's no danger of confusion. Actually, the lack
of a .sig feature on Google Groups is probably what prompts me to seek
compensatory coolness by some avenue.
Cheers,
Brandon Van Every
Mallor wrote:
> "Mallor" is a posting nickname that I have used at various times in
> various places. I used to use it when sowing the seeds of discord, as
> in a quite unfinished mythology I've worked on for a number of years,
> that's what Mallor does. But somewhere along the way when configuring
> Google Groups I stuck it in there, and I have complete disinterest in
> figuring out how to change it again. "Mallor" doesn't mean much of
> anything for public posting purposes as of late, except that I think
> it's a cooler name than my own. Which always appears anyways at the
> end of my posts, so there's no danger of confusion. Actually, the lack
> of a .sig feature on Google Groups is probably what prompts me to seek
> compensatory coolness by some avenue.
A better nickname would be "Marvin" (you know, Douglas Adams...) :-)
--
Frank Buss, ··@frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de
Frank Buss wrote:
> Mallor wrote:
>
> > "Mallor" is a posting nickname that I have used at various times in
> > various places. I used to use it when sowing the seeds of discord, as
> > in a quite unfinished mythology I've worked on for a number of years,
> > that's what Mallor does. But somewhere along the way when configuring
> > Google Groups I stuck it in there, and I have complete disinterest in
> > figuring out how to change it again. "Mallor" doesn't mean much of
> > anything for public posting purposes as of late, except that I think
> > it's a cooler name than my own. Which always appears anyways at the
> > end of my posts, so there's no danger of confusion. Actually, the lack
> > of a .sig feature on Google Groups is probably what prompts me to seek
> > compensatory coolness by some avenue.
>
> A better nickname would be "Marvin" (you know, Douglas Adams...) :-)
I'm neither paranoid nor despondent, so I'm afraid I don't understand
the joke. It's been awhile since I've read THHGTTG, so if there's some
subtlety of Marvin The Paranoid Android's behavior that I'm missing,
feel free to explain.
Cheers,
Brandon Van Every
Darren New wrote:
> ··············@gmail.com wrote:
>
>> The only thing vendors have to fear from OSS is that the OSS
>> implementations will match all the features of the commercial product.
>
>
> Not really. If you had the choice of flying across-country free in coach
> class, or pay for a first class flight, which would you do?
>
> A commercial product has to not only have more features, but enough
> *extra* features to make the price of developing *all* the features
> attactive to buyers.
And apparently our FSF fans are willing to flap all the way across the
country and the check themselves into intensive care on the other side
as long as it is "free".
kt
ps. I said "country" and South Africa is a country, cue Spitzer on the
primitives vs. the Europeans. k
--
Cells: http://common-lisp.net/project/cells/
"I'll say I'm losing my grip, and it feels terrific."
-- Smiling husband to scowling wife, New Yorker cartoon
From: Ken Tilton
Subject: Re: Any good free cross Plattform lisps ?
Date:
Message-ID: <CLyAg.2577$Lz7.21@fe09.lga>
··············@gmail.com wrote:
> Ken Tilton wrote:
>
>>[snip]
>>I should add that my first win32 Lisp was a 1998 ACL "Personal Edition"
>>(not to be confused with later Trial/Express editions) $495 or so back
>>then. That is not $50-100, but then you have the other problem driving
>>prices up: a miniscule market.
>>
>>[snip]
>>
>>Thanks, Stallman!
>>
>
>
> I don't want to start a religious/political debate here, but with a
> miniscule market like Lisp I would think that Open Source Software
> would actually help create a larger market,...
I suggest you ping Franz, Digitool, and Lispwroks for the precise
numbers on how moch it has helped them to have those yobbo idiots
slaving away at CMUCL instead of <gasp> simply buying a license to a
real environment.
kt
--
Cells: http://common-lisp.net/project/cells/
"I'll say I'm losing my grip, and it feels terrific."
-- Smiling husband to scowling wife, New Yorker cartoon
Ken Tilton wrote:
> I suggest you ping Franz, Digitool, and Lispwroks for the precise
> numbers on how moch it has helped them to have those yobbo idiots
> slaving away at CMUCL instead of <gasp> simply buying a license to a
> real environment.
>
I suggest you ping those "yobbo idiots" and ask how many of them are
slaving away at CMUCL simply because they could never afford to buy a
license from those commercial vendors in the first place.
What about SBCL? Are they "yobbo idiots" too?
··············@gmail.com wrote:
> Ken Tilton wrote:
> > I suggest you ping Franz, Digitool, and Lispwroks for the precise
> > numbers on how moch it has helped them to have those yobbo idiots
> > slaving away at CMUCL instead of <gasp> simply buying a license to a
> > real environment.
> >
>
> I suggest you ping those "yobbo idiots" and ask how many of them are
> slaving away at CMUCL simply because they could never afford to buy a
> license from those commercial vendors in the first place.
The problem is that these LISP businesses don't appear to be interested
in spreading LISP unless they can make a buck and while there's nothing
wrong with that, it's led to less LISP usage I think.
From: Ken Tilton
Subject: Re: Any good free cross Plattform lisps ?
Date:
Message-ID: <Gq4Bg.78$qy1.21@fe11.lga>
OMouse wrote:
> What about SBCL? Are they "yobbo idiots" too?
>
> ··············@gmail.com wrote:
>
>>Ken Tilton wrote:
>>
>>>I suggest you ping Franz, Digitool, and Lispwroks for the precise
>>>numbers on how moch it has helped them to have those yobbo idiots
>>>slaving away at CMUCL instead of <gasp> simply buying a license to a
>>>real environment.
>>>
>>
>>I suggest you ping those "yobbo idiots" and ask how many of them are
>>slaving away at CMUCL simply because they could never afford to buy a
>>license from those commercial vendors in the first place.
>
>
> The problem is that these LISP businesses don't appear to be interested
> in spreading LISP unless they can make a buck and while there's nothing
> wrong with that, it's led to less LISP usage I think.
Profiteering has certainly held down Windows, VC++ and Microsoft Office.
kenny
--
Cells: http://common-lisp.net/project/cells/
"I'll say I'm losing my grip, and it feels terrific."
-- Smiling husband to scowling wife, New Yorker cartoon
On Sat, 5 Aug 2006, Ken Tilton wrote:
>
> OMouse wrote:
>> What about SBCL? Are they "yobbo idiots" too?
>>
>> ··············@gmail.com wrote:
>>
>> > Ken Tilton wrote:
>> >
>> > > I suggest you ping Franz, Digitool, and Lispwroks for the precise
>> > > numbers on how moch it has helped them to have those yobbo idiots
>> > > slaving away at CMUCL instead of <gasp> simply buying a license to a
>> > > real environment.
>> > >
>> >
>> > I suggest you ping those "yobbo idiots" and ask how many of them are
>> > slaving away at CMUCL simply because they could never afford to buy a
>> > license from those commercial vendors in the first place.
>>
>>
>> The problem is that these LISP businesses don't appear to be interested
>> in spreading LISP unless they can make a buck and while there's nothing
>> wrong with that, it's led to less LISP usage I think.
>
> Profiteering has certainly held down Windows, VC++ and Microsoft Office.
But those products are not widely used because of Microsoft's
profiteering, but *despite* it.
It has everything to do with marketing and distribution: Windows comes as
standard; Office is *the* standard.
I don't think Microsoft use VC++ for profiteering -- if they did, they'd
probably try to pin some obnoxious royalty scheme on developers. They seem
to understand that their business in Windows and Office depends upon them
being nice to developers.
GENERIC-COMMERCIAL-LISP-VENDOR has no particular incentive to ensure wide
usage of Lisp, because GENERIC-COMMERCIAL-LISP-VENDOR doesn't make a
widely used operating system whose continued wide usage depends upon wide
usage of Lisp: if Microsoft make it very expensive to develop and
distribute Windows programs, they might lose developers to other
platforms, which in turn could lead to them losing users to other
platforms.
For Franz, LispWorks and Digitool, the small market for Lisp is a
non-issue. They're making money, and (one presumes) don't believe they
could make more money by expanding their market -- a larger market brings
with it greater competition. How would commercial vendors fair against a
possibly inferior but aggresively marketed Microsoft Common Lisp? (Or
perhaps it would be Microsoft L#? Microsoft Lisp.NET?)
> kenny
>
> --
> Cells: http://common-lisp.net/project/cells/
>
> "I'll say I'm losing my grip, and it feels terrific."
> -- Smiling husband to scowling wife, New Yorker cartoon
Nathan Baum wrote:
> On Sat, 5 Aug 2006, Ken Tilton wrote:
>>
>> OMouse wrote:
>>> What about SBCL? Are they "yobbo idiots" too?
>>>
>>> ··············@gmail.com wrote:
>>>
>>> > Ken Tilton wrote:
>>> > > > I suggest you ping Franz, Digitool, and Lispwroks for the precise
>>> > > numbers on how moch it has helped them to have those yobbo idiots
>>> > > slaving away at CMUCL instead of <gasp> simply buying a license to a
>>> > > real environment.
>>> > > > > I suggest you ping those "yobbo idiots" and ask how many of
>>> them are
>>> > slaving away at CMUCL simply because they could never afford to buy a
>>> > license from those commercial vendors in the first place.
>>>
>>>
>>> The problem is that these LISP businesses don't appear to be interested
>>> in spreading LISP unless they can make a buck and while there's nothing
>>> wrong with that, it's led to less LISP usage I think.
>>
>> Profiteering has certainly held down Windows, VC++ and Microsoft Office.
>
> But those products are not widely used because of Microsoft's
> profiteering, but *despite* it.
>
> It has everything to do with marketing and distribution: Windows comes
> as standard; Office is *the* standard.
>
> I don't think Microsoft use VC++ for profiteering -- if they did, they'd
> probably try to pin some obnoxious royalty scheme on developers. They
> seem to understand that their business in Windows and Office depends
> upon them being nice to developers.
Exactly. That was one of the reasons IBM OS/2 eventually failed. Not
only did they not know how to market it, they were not nearly so
friendly to developers. IIRC, IBM was trying to make a profit off
selling their OS/2 device driver development kits (I can't remember the
price they were asking but it was significant) at a time when Microsoft
was practically throwing theirs at any developer that expressed any
interest. End result of that and other factors -- lots of third-party
drivers for Windows, few for OS/2. And OS/2 was a lot better OS in many
ways than Windows.
--Larry
I just looked at the price of the Lispworks 5.0, it's even more
expensive than the 4.x version. I am not sure why the price is getting
higher and higher.
·············@gmail.com" <············@gmail.com> writes:
> I just looked at the price of the Lispworks 5.0, it's even more
> expensive than the 4.x version. I am not sure why the price is getting
> higher and higher.
Perhaps the product is getting better and better. Perhaps they are
now well enough known, and have a good enough reputation that they
think the additional $$$ will not be a problem for their potential
customers. Perhaps they thought they suffered from the effect where
products priced too cheaply are frowned upon (there _must_ be something
wrong with it if it's so cheap).
Perhaps a millions things. But in the end, the only answer which is
obviously true is: because lispworks thinks (rightly or wrongly) that
that is the price at which they will maximise their long term profits.
Isn't that obvious?
--ap
Alain Picard wrote:
> ·············@gmail.com" <············@gmail.com> writes:
>
> > I just looked at the price of the Lispworks 5.0, it's even more
> > expensive than the 4.x version. I am not sure why the price is getting
> > higher and higher.
>
> Perhaps the product is getting better and better. Perhaps they are
> now well enough known, and have a good enough reputation that they
> think the additional $$$ will not be a problem for their potential
> customers. Perhaps they thought they suffered from the effect where
> products priced too cheaply are frowned upon
Perhaps they think that anyone who've used their product
is effectively locked in[1]?
> (there _must_ be something
> wrong with it if it's so cheap).
>
> Perhaps a millions things.
perhaps, perhaps, perhaps ...
? But in the end, the only answer which is
> obviously true is: because lispworks thinks (rightly or wrongly) that
> that is the price at which they will maximise their long term profits.
>
> Isn't that obvious?
>
[1] How hard is it to port a non-trivial application
written in a commmercial lisp to another commercial
lisp?
goose,
"goose" <····@webmail.co.za> writes:
> [1] How hard is it to port a non-trivial application
> written in a commmercial lisp to another commercial
> lisp?
How long is a piece of string?
The answer to your question is of course "It depends". It can be
trivial, if you application is written entirely inside Common Lisp.
It can be completely infeasible, if your application makes heavy use
of features of one implementation not available on the other.
It can be somewhere in between these extremes if the application is
written with portability in mind, or if the use of vendor-specific
features is limited, or if the vendor-specific features used have
relatively simple mappings to the vendor-specific features of the
other implementation (sockets and mp usually fall within this
category).
I would have thought this obvious?
...Peder...
--
I wish a new life awaited _me_ in some off-world colony.
Nathan Baum <···········@btinternet.com> writes:
> For Franz, LispWorks and Digitool, the small market for Lisp is a
> non-issue. They're making money, and (one presumes) don't believe they
> could make more money by expanding their market -- a larger market brings
> with it greater competition. How would commercial vendors fair against a
> possibly inferior but aggresively marketed Microsoft Common Lisp? (Or
> perhaps it would be Microsoft L#? Microsoft Lisp.NET?)
Put the spark plug down and step away from the crack.
Thomas F. Burdick wrote:
> Nathan Baum <···········@btinternet.com> writes:
>
> > For Franz, LispWorks and Digitool, the small market for Lisp is a
> > non-issue. They're making money, and (one presumes) don't believe they
> > could make more money by expanding their market -- a larger market brings
> > with it greater competition. How would commercial vendors fair against a
> > possibly inferior but aggresively marketed Microsoft Common Lisp? (Or
> > perhaps it would be Microsoft L#? Microsoft Lisp.NET?)
>
> Put the spark plug down and step away from the crack.
His fantasy would work better if it was Apple marketing Apple Lisp or
something. Or why not have Paul Graham fund a Lisp start-up? I bet that
would scare a few people.
Ken Tilton wrote:
> OMouse wrote:
> > What about SBCL? Are they "yobbo idiots" too?
> >
> > ··············@gmail.com wrote:
> >
> >>Ken Tilton wrote:
> >>
> >>>I suggest you ping Franz, Digitool, and Lispwroks for the precise
> >>>numbers on how moch it has helped them to have those yobbo idiots
> >>>slaving away at CMUCL instead of <gasp> simply buying a license to a
> >>>real environment.
> >>>
> >>
> >>I suggest you ping those "yobbo idiots" and ask how many of them are
> >>slaving away at CMUCL simply because they could never afford to buy a
> >>license from those commercial vendors in the first place.
> >
> >
> > The problem is that these LISP businesses don't appear to be interested
> > in spreading LISP unless they can make a buck and while there's nothing
> > wrong with that, it's led to less LISP usage I think.
>
> Profiteering has certainly held down Windows, VC++ and Microsoft Office.
There is an important distinction to be made between profiteering and
commodification. Microsoft commodified the GUI-driven OS. Franz and
Lispworks are doing absolutely nothing for Lisp's commodification. In
other industries, those who resist commodification eventually die.
Such happened to SGI in the 3D hardware realm, for instance. Many Unix
vendors have gone by the wayside in the face of the Windows NT PC, and
also Linux, because the PC HW became much cheaper than vendor-specific
Unix HW. Although possibly, the sphere of Lisp's value propositions is
not easily commodified, so it gets more leeway. Still I wonder how
many advantages Lisp can offer the world vs. outsourcing typical
problems to legions of cheap programmers in India et al.
Cheers,
Brandon Van Every
"Mallor" <···········@gmail.com> writes:
> There is an important distinction to be made between profiteering and
> commodification. Microsoft commodified the GUI-driven OS. Franz and
> Lispworks are doing absolutely nothing for Lisp's commodification. In
> other industries, those who resist commodification eventually die.
> Such happened to SGI in the 3D hardware realm, for instance. Many Unix
> vendors have gone by the wayside in the face of the Windows NT PC, and
> also Linux, because the PC HW became much cheaper than vendor-specific
> Unix HW. Although possibly, the sphere of Lisp's value propositions is
> not easily commodified, so it gets more leeway. Still I wonder how
> many advantages Lisp can offer the world vs. outsourcing typical
> problems to legions of cheap programmers in India et al.
Yes, lisp systems don't come with an integrated AI doing the
programming for you, while java systems, in India, do. ;-)
--
__Pascal Bourguignon__ http://www.informatimago.com/
"Do not adjust your mind, there is a fault in reality"
-- on a wall many years ago in Oxford.
Pascal Bourguignon wrote:
> "Mallor" <···········@gmail.com> writes:
> > There is an important distinction to be made between profiteering and
> > commodification. Microsoft commodified the GUI-driven OS. Franz and
> > Lispworks are doing absolutely nothing for Lisp's commodification. In
> > other industries, those who resist commodification eventually die.
> > Such happened to SGI in the 3D hardware realm, for instance. Many Unix
> > vendors have gone by the wayside in the face of the Windows NT PC, and
> > also Linux, because the PC HW became much cheaper than vendor-specific
> > Unix HW. Although possibly, the sphere of Lisp's value propositions is
> > not easily commodified, so it gets more leeway. Still I wonder how
> > many advantages Lisp can offer the world vs. outsourcing typical
> > problems to legions of cheap programmers in India et al.
>
> Yes, lisp systems don't come with an integrated AI doing the
> programming for you, while java systems, in India, do. ;-)
Nobody's using Lisp in the game industry for their "AI". The
oft-quoted Naughty Dog is a one hit wonder, and it's been awhile since
they were doing that.
As for "real" AI, I don't know what the market size is, or what
offshore people can provide. I was alarmed to see that Indians could
design semiconductors for you. Maybe I overestimate the difficulty of
that task, but it says to me that very complicated problems can be
offshored. Or portions of them, if they're sufficiently canned /
commodifiable.
My perusals of comp.ai.* have led me to believe that Lisp is just one
language chosen out of many, and that it is not synonymous with AI
anymore. In other words, a person in the AI market sells his AI
expertise, not his Lisp expertise.
Cheers,
Brandon Van Every
"Mallor" <···········@gmail.com> writes:
> The oft-quoted Naughty Dog is a one hit wonder, and it's been awhile
> since they were doing that.
Actually, I do believe they used GOAL for several successful games,
and seem to recall that they also licensed their engines to another
studio/title/series which also did quite well (Insomniac Games,
Ratchet & Clank 1/2/3). Then they were acquired by Sony and lost their
soul, more or less.
(If memory serves, the problem was that the technology, GOAL, couldn't
be taught to C++ programmers, so they had to standardize on
C++. Sounds familiar enough to be plausible, at least, but corrections
are welcome.)
Best,
Thomas
--
Thomas Lindgren
Thomas Lindgren <···········@*****.***> writes:
> "Mallor" <···········@gmail.com> writes:
>
>> The oft-quoted Naughty Dog is a one hit wonder, and it's been awhile
>> since they were doing that.
>
> Actually, I do believe they used GOAL for several successful games,
> and seem to recall that they also licensed their engines to another
> studio/title/series which also did quite well (Insomniac Games,
> Ratchet & Clank 1/2/3). Then they were acquired by Sony and lost their
> soul, more or less.
>
> (If memory serves, the problem was that the technology, GOAL, couldn't
> be taught to C++ programmers, so they had to standardize on
> C++. Sounds familiar enough to be plausible, at least, but corrections
> are welcome.)
At least, it seems that using lisp doesn't prevent startups to be sold
and their owners to cash in.
--
__Pascal Bourguignon__ http://www.informatimago.com/
"By filing this bug report you have challenged the honor of my
family. Prepare to die!"
>
> At least, it seems that using lisp doesn't prevent startups to be sold
> and their owners to cash in.
>
Since when are company takeovers based on technology. The people
responsible for this kind of deals are all Pointy-Haired Bosses
llothar wrote:
> >
> > At least, it seems that using lisp doesn't prevent startups to be sold
> > and their owners to cash in.
> >
>
> Since when are company takeovers based on technology. The people
> responsible for this kind of deals are all Pointy-Haired Bosses
It also indicates that Lisp cannot survive in large corporations, which
speaks ill of the language's future in terms of widespread job
availability. Examples like Naughty Dog can be viewed as long term
negatives. The GameDeveloper postmortems were a two-edged sword: GOAL
was listed both in "What Went Right" and "What Went Wrong." "Goal
Rules!" "Goal sucks!" Most people in the game industry looking for
advice and history on what works and what doesn't, won't find the
Naughty Dog story to be compelling. Lispers may say it sounds great,
but are they *game developers* ? That's the opinion that counts, from
a movers / shakers / conversion standpoint.
I don't think Lisp will get traction in any market, until it
demonstrates it can do something that other languages can't do. Big
automation / genesis problems. I think about such problems from a game
development standpoint. I am hoping that Will Wright's "Spore" will
popularize procedural content in the game industry, and that Chicken
Scheme will then prove to be the right tool for the job that everyone
was looking for.
As a Grand Strategy, "an efficiency advantage for smart programmers
only" will never win.
Cheers,
Brandon Van Every
"Mallor" <···········@gmail.com> writes:
> Examples like Naughty Dog can be viewed as long term
> negatives. The GameDeveloper postmortems were a two-edged sword: GOAL
> was listed both in "What Went Right" and "What Went Wrong." "Goal
> Rules!" "Goal sucks!" Most people in the game industry looking for
> advice and history on what works and what doesn't, won't find the
> Naughty Dog story to be compelling.
As I recall, the "GOAL sucks!" issues were due to immature technology.
Basically, this is (similar to) an early adopter problem. I'd thus
leave the fence sitters for later and go for those who are willing to
focus on the competitive advantages.
By the way, I agree with the gist of what you write further on. In
short, there has to be a clear, compelling advantage to switching.
Best,
Thomas
--
Thomas Lindgren
··············@gmail.com wrote:
> Ken Tilton wrote:
> > I suggest you ping Franz, Digitool, and Lispwroks for the precise
> > numbers on how moch it has helped them to have those yobbo idiots
> > slaving away at CMUCL instead of <gasp> simply buying a license to a
> > real environment.
> >
>
> I suggest you ping those "yobbo idiots" and ask how many of them are
> slaving away at CMUCL simply because they could never afford to buy a
> license from those commercial vendors in the first place.
Licenses are not *that* expensive. You can get a royalty-free
Professional edition from LispWorks for $1100. Franz is a different
story, but $1100 is not so bad when you consider what a commercial IDE
can offer. If you're a hobbyist, then sure, this is a big expense and
maybe using CMUCL (or CLISP or SBCL) makes sense, especially if you're
interested in developing the environment itself.
But if you're a business, it's a no-brainer. As nice as SLIME is, the
LispWorks IDE is even nicer. A thousand dollars up front for a serious
competitive advantage is well worth it. If you're in business and
can't afford the cost of a commercial Lisp, you have bigger problems
than you think.
Also, your claim that the authors of CMUCL "could never afford to buy a
license from those commercial vendors" (probably not true, btw) doesn't
detract from Kenny's point - that getting by with a free Lisp when
there are more productive commercial environments out there is a very
silly thing to do, if the only reason for taking that path is saving
money (again, if you're a hobbyist or you want to work on Lisp itself,
then of course that's a different story).
--
The journey of 10,000 miles begins with a single clichè.
> But if you're a business, it's a no-brainer. As nice as SLIME is, the
> LispWorks IDE is even nicer. A thousand dollars up front for a serious
> competitive advantage is well worth it. If you're in business and
> can't afford the cost of a commercial Lisp, you have bigger problems
> than you think.
I have a small shareware company and make just a little bit money
(enough for paying my bills, my house and having a beer or two).
But at the beginning the price seemed like serious money. And while i
would have considered a purchase even for the 1200 Euro, i didn't make
it. The reason is that i want to support multiple pattforms (mostly for
marketing reasons) and this includes Solaris (Sun/Intel) and FreeBSD
and MacOSX and of course Linux/Windows. This sums up for no additional
value because i don't need an IDE on more then one system but i have to
pay the whole package for each system.
If they would offer a naked compiler/interpreter for say 300 Euro it
would be indeed a no brainer (all other successfull languages have this
option). But 6800 Euro and 2000 Euro support each year is far beyond
the number on my bank account.
This price policy really sucks (it's by the way the same for Eiffel but
there the whole package only takes 500 Euro per system).
llothar wrote:
> [snip]
> But 6800 Euro and 2000 Euro support each year is far beyond
> the number on my bank account.
>
Case in point. Here we have a small business developer who has already
been an asset to the Ruby community, and who could potentially be an
asset to the Lisp community as well. But it sure doesn't seem like the
"big guys" are interested in a small business developer like him with
prices like that, does it? And unfortunately, the free implementations
usually don't have the features a small business developer really needs
(as already shown by the OP), so they are forced to look elsewhere and
find another programming language altogether.
And the Lisp community has lost another potential member...
··············@gmail.com wrote:
> llothar wrote:
> > [snip]
> > But 6800 Euro and 2000 Euro support each year is far beyond
> > the number on my bank account.
> >
>
> Case in point. Here we have a small business developer who has already
> been an asset to the Ruby community, and who could potentially be an
> asset to the Lisp community as well. But it sure doesn't seem like the
> "big guys" are interested in a small business developer like him with
> prices like that, does it? And unfortunately, the free implementations
> usually don't have the features a small business developer really needs
> (as already shown by the OP), so they are forced to look elsewhere and
> find another programming language altogether.
>
> And the Lisp community has lost another potential member...
Lothar is in a different situation than I am. Since I write and deploy
on Linux only, a single Professional license is enough. He's right
that price is a more significant hurdle when you're developing for
Windows, Mac, Linux, and FreeBSD ( 4 x 1200 = 4800 + whatever the
LispWorks for generic UNIX costs). But there are a couple of choices
here. If you believe Lisp (+ a full Lisp IDE) will make you insanely
more productive than other languages, then you pay the license costs
up-front and make better software faster than your competitors can and
you win. If you believe in the power of Lisp, but really can't afford
it, then you wrestle with a free Lisp a little. It's definitely
workable; you just distribute CLISP or SBCL and toss a couple of FASL's
in (you do miss out on the advanced IDE's, though). Or you work to
improve the free tools out there. If you really don't think Lisp is
worth the trouble, you use another language. That's all.
If Lothar would rather not have to buy an IDE for platforms he's just
deploying to, maybe the vendor can work out a deal with him (give it a
shot, Lothar). But things certainly aren't going to change by
discussing it on comp.lang.lisp. The Lisp vendors haven't just picked
their prices willy-nilly - and whatever prices they've picked are
obviously reasonable enough that they're running pretty good
businesses. I don't understand why this thread comes up so much.
There are basically two choices: 1) buy a Lisp, or 2) use a free Lisp.
How would you like this situation to change? The vendors are not going
to change their prices just because you'd prefer to pay less; free
software can only get better if interested people work on it. Am I
missing something?
Bill Atkins wrote:
> If you believe Lisp (+ a full Lisp IDE) will make you insanely
> more productive than other languages, then you pay the license costs
> up-front and make better software faster than your competitors can and
> you win.
I see no reason to believe. Granted, I've worked on getting Chicken
Scheme to basically work on Windows for 10 months now, and that has
colored my perception of the benefit and cost of open source. I got to
do all the grungy stuff that in a commercial context, people get paid
$$$$$ to bother with. It's all very laborious and time consuming.
Now, perhaps I could have just started with something that works and
didn't need basic maintenance. Will Lisp or Scheme or any other exotic
HLL perform miracles for me? I really don't believe it. Too much of
computerdom is tedious and requires abundant RTFM for me to put
credence in it. I persist with HLLs because I believe they make me
more productive than LLLs, not because I believe there's really some
quantum leap in productivity to be gained from it.
Cheers,
Brandon Van Every
> didn't need basic maintenance. Will Lisp or Scheme or any other exotic
> HLL perform miracles for me? I really don't believe it. Too much of
> computerdom is tedious and requires abundant RTFM for me to put
> credence in it. I persist with HLLs because I believe they make me
> more productive than LLLs, not because I believe there's really some
> quantum leap in productivity to be gained from it.
I agree here.
The old saying "There is nothing like a silver bullet" includes
language selection. I hear many times that people are getting 10x more
productive with Smalltalk, Python or Ruby but i don't think this scales
to application level design.
I say the biggest advantage is the increased productivity coming from
satisfaction by the user. A bad language, a bad environment, a bad
library can make programmers depressive and kill productivity.
So i think many reasons to use X instead of Y are first human reasons
and only second technology based.
llothar wrote:
>
> I say the biggest advantage is the increased productivity coming from
> satisfaction by the user. A bad language, a bad environment, a bad
> library can make programmers depressive and kill productivity.
>
> So i think many reasons to use X instead of Y are first human reasons
> and only second technology based.
Case in point: 2 weeks ago, I thought all the Chicken Scheme CMake
build issues might be "done." Maybe I'd licked the last major problem
and I could move on to the real mission, OpenGL and game development.
So I started to do so, and, um, ahh... it wasn't the same drill as I'd
been doing for the past 9 months. I really had no idea what I intended
to write. I spun wheels for about a week trying to make the transition
and really didn't get anything done. Then I figured out the basic
OpenGL support didn't actually work, and I was back in business fixing
things. I was happier having a task that I knew what I was doing from
moment to moment and could see progress towards a concrete goal.
Cheers,
Brandon Van Every
Bill Atkins <·········@gmail.com> wrote:
+---------------
| ··············@gmail.com wrote:
| > Ken Tilton wrote:
| > > I suggest you ping Franz, Digitool, and Lispwroks for the precise
| > > numbers on how moch it has helped them to have those yobbo idiots
| > > slaving away at CMUCL instead of <gasp> simply buying a license to a
| > > real environment.
| >
| > I suggest you ping those "yobbo idiots" and ask how many of them are
| > slaving away at CMUCL simply because they could never afford to buy a
| > license from those commercial vendors in the first place.
...
| Also, your claim that the authors of CMUCL "could never afford to buy a
| license from those commercial vendors" (probably not true, btw) ...
+---------------
Not even close to true. CMUCL was developed at -- wait for it --
Carnegie Mellon University (CMU) [duh!], funded with grants from
the good ol' U.S. Department of Defense (specifically, DARPA),
under a contract which specifically placed the results into the
public domain. "Your tax dollars at work!!" Enjoy, folks!
Yes, when its DARPA funding ran out in 1994, support/development
was picked up by community volunteers, which continues even today.
See <http://www.cons.org/cmucl/credits.html> for more details...
-Rob
p.s. CMUCL's CLOS, LOOP, & CLX are derived from Xerox PARC's "PCL,
Symbolics LOOP [from code written at MIT], and IT's "CLX", respectively.
The copyrights on all of those pieces are "MIT-like", which is to say
about as open as you can get short of public-domain.
-----
Rob Warnock <····@rpw3.org>
627 26th Avenue <URL:http://rpw3.org/>
San Mateo, CA 94403 (650)572-2607
Rob Warnock wrote:
> Bill Atkins <·········@gmail.com> wrote:
> +---------------
> ...
> | Also, your claim that the authors of CMUCL "could never afford to buy a
> | license from those commercial vendors" (probably not true, btw) ...
> +---------------
>
> Not even close to true. CMUCL was developed at -- wait for it --
> Carnegie Mellon University (CMU) [duh!], funded with grants from
> the good ol' U.S. Department of Defense (specifically, DARPA),
> under a contract which specifically placed the results into the
> public domain. "Your tax dollars at work!!" Enjoy, folks!
Looks like I should point out that I was speaking about users, not
developers, of CMUCL. Poor choice of preposition on my part, I guess.
From: Ken Tilton
Subject: Re: Any good free cross Plattform lisps ?
Date:
Message-ID: <1T0Bg.262$Fs3.143@fe09.lga>
··············@gmail.com wrote:
> Ken Tilton wrote:
>
>>I suggest you ping Franz, Digitool, and Lispwroks for the precise
>>numbers on how moch it has helped them to have those yobbo idiots
>>slaving away at CMUCL instead of <gasp> simply buying a license to a
>>real environment.
>>
>
>
> I suggest you ping those "yobbo idiots" and ask how many of them are
> slaving away at CMUCL simply because they could never afford to buy a
> license from those commercial vendors in the first place.
>
Bah. A few weeks down in the coal mines will get them a shiny new LW,
and they could use the exercise.
kt
--
Cells: http://common-lisp.net/project/cells/
"I'll say I'm losing my grip, and it feels terrific."
-- Smiling husband to scowling wife, New Yorker cartoon
Ken Tilton <·········@gmail.com> writes:
> ··············@gmail.com wrote:
>> Ken Tilton wrote:
>>
>>>I suggest you ping Franz, Digitool, and Lispwroks for the precise
>>>numbers on how moch it has helped them to have those yobbo idiots
>>>slaving away at CMUCL instead of <gasp> simply buying a license to a
>>>real environment.
>>>
>> I suggest you ping those "yobbo idiots" and ask how many of them are
>> slaving away at CMUCL simply because they could never afford to buy a
>> license from those commercial vendors in the first place.
>>
>
> Bah. A few weeks down in the coal mines will get them a shiny new LW,
> and they could use the exercise.
What coal mine?
All the remaining coal mines in my country have been closed in 2005.
And the rest of coal mines in my continent are way on the same way.
--
__Pascal Bourguignon__ http://www.informatimago.com/
NOTE: The most fundamental particles in this product are held
together by a "gluing" force about which little is currently known
and whose adhesive power can therefore not be permanently
guaranteed.
llothar wrote:
>
> OSS everywhere is not a real advantage for the Software World, it
> results in much worse but cost-free software. It is not good - but
> people don't see it at the moment.
Expensive proprietary software results in slavery to empires like
Microsoft. Bad for independent minded developers like myself, who
really don't care to live some kind of "normal" lifestyle as a
corporate worker bee. I am left to conclude that everything about
computers takes way too much work no matter what you do. OSS is just
one style of how "too much work" is going to pan out. Being slave to
your paycheck is another.
Cheers,
Brandon Van Every
> Seems that there is only the expensive way of LispWorks or ACL (summing
> up to 5000 US$). Free Lisp systems still are do not meet my goals.
Uuups, just found out that even for this money i'm not getting a system
that uses multi CPU's/Cores, LispWorks is doing the same as python,
using a global interpreter lock. Bad.
So this is not an option for me.
llothar wrote:
> > Seems that there is only the expensive way of LispWorks or ACL (summing
> > up to 5000 US$). Free Lisp systems still are do not meet my goals.
>
> Uuups, just found out that even for this money i'm not getting a system
> that uses multi CPU's/Cores, LispWorks is doing the same as python,
> using a global interpreter lock. Bad.
>
> So this is not an option for me.
They haven't killed off this middle-ground competitor: Corman Lisp
($250 for a full license) http://www.cormanlisp.com
Here's their specs say about threading support:
* Multi-threading. The compiler has been designed to support multiple
threads of execution, with special variable bindings on a per thread
basis. These threads are managed by the operating system, affording you
the full advantage of the features provided by the operating system.
It's Windows only, so it may not fill your previously stated
multi-platform requirements, but for the Windows platform it seems to
have what you need for a reasonable price.
llothar wrote:
> Uuups, just found out that even for this money i'm not getting a system
> that uses multi CPU's/Cores, LispWorks is doing the same as python,
> using a global interpreter lock. Bad.
Where did you see this?
Bill Atkins wrote:
> llothar wrote:
> > Uuups, just found out that even for this money i'm not getting a system
> > that uses multi CPU's/Cores, LispWorks is doing the same as python,
> > using a global interpreter lock. Bad.
>
> Where did you see this?
LispWorks User Guide Page 140:
Each Lisp mp:process has a separate native thread, but Lisp code can
only run
in one thread at a time and a lock is used to enforce this.
llothar wrote:
> > Seems that there is only the expensive way of LispWorks or ACL (summing
> > up to 5000 US$). Free Lisp systems still are do not meet my goals.
>
> Uuups, just found out that even for this money i'm not getting a system
> that uses multi CPU's/Cores, LispWorks is doing the same as python,
> using a global interpreter lock. Bad.
>
> So this is not an option for me.
You might want to write to LispWorks Ltd. and state your interest to
buy a license if
it would support multi-cpu systems. They recently released LispWorks 5,
which
improved a lot of the LispWorks internals (for example they now have
64bit
versions). I agree, supporting multi-cpu systems could be a nice
feature for
a LispWorks 6.
Currently OpenMCL and Scieneer CL are providing this feature. I'm not
sure about Corman CL (Windows), but it could be.
From: Christophe Rhodes
Subject: Re: Any good free cross Plattform lisps ?
Date:
Message-ID: <sqac6hc42v.fsf@cam.ac.uk>
·······@corporate-world.lisp.de" <······@corporate-world.lisp.de> writes:
> llothar wrote:
>> Uuups, just found out that even for this money i'm not getting a system
>> that uses multi CPU's/Cores, LispWorks is doing the same as python,
>> using a global interpreter lock. Bad.
>>
>> So this is not an option for me.
>
> You might want to write to LispWorks Ltd. and state your interest to
> buy a license if it would support multi-cpu systems. They recently
> released LispWorks 5, which improved a lot of the LispWorks
> internals (for example they now have 64bit versions). I agree,
> supporting multi-cpu systems could be a nice feature for a LispWorks
> 6.
>
> Currently OpenMCL and Scieneer CL
... and SBCL, and ABCL, and (I believe) ECL ...
> are providing this feature. I'm not sure about Corman CL (Windows),
> but it could be.
Christophe
Christophe Rhodes wrote:
> ·······@corporate-world.lisp.de" <······@corporate-world.lisp.de> writes:
>
> > llothar wrote:
> >> Uuups, just found out that even for this money i'm not getting a system
> >> that uses multi CPU's/Cores, LispWorks is doing the same as python,
> >> using a global interpreter lock. Bad.
> >>
> >> So this is not an option for me.
> >
> > You might want to write to LispWorks Ltd. and state your interest to
> > buy a license if it would support multi-cpu systems. They recently
> > released LispWorks 5, which improved a lot of the LispWorks
> > internals (for example they now have 64bit versions). I agree,
> > supporting multi-cpu systems could be a nice feature for a LispWorks
> > 6.
> >
> > Currently OpenMCL and Scieneer CL
>
> ... and SBCL, and ABCL, and (I believe) ECL ...
Interesting. Are we really talking about the same thing?
* running native threads in parallel on multi-processor machines?
For example LispWorks has some kind of native threads on several
platforms, but on multi-cpu/core machines, only one thread
will execute at a time. Thus only one cpu/core can be used at a time.
On which platforms would SBCL/ABCL/ECL support threads running
in parallel on multi-cpu/cores?
>
> > are providing this feature. I'm not sure about Corman CL (Windows),
> > but it could be.
>
> Christophe
From: Christophe Rhodes
Subject: Re: Any good free cross Plattform lisps ?
Date:
Message-ID: <sq64h5c0s9.fsf@cam.ac.uk>
·······@corporate-world.lisp.de" <······@corporate-world.lisp.de> writes:
> Christophe Rhodes wrote:
>> ·······@corporate-world.lisp.de" <······@corporate-world.lisp.de> writes:
>>
>> > Currently OpenMCL and Scieneer CL
>>
>> ... and SBCL, and ABCL, and (I believe) ECL ...
>
> Interesting. Are we really talking about the same thing?
>
> * running native threads in parallel on multi-processor machines?
Yes.
> For example LispWorks has some kind of native threads on several
> platforms, but on multi-cpu/core machines, only one thread
> will execute at a time. Thus only one cpu/core can be used at a time.
>
> On which platforms would SBCL/ABCL/ECL support threads running
> in parallel on multi-cpu/cores?
SBCL: x86 Linux, Mac OS and Solaris; x86-64 Linux
ABCL: anywhere, where the underlying JVM's threads being used are
native. (I believe all the current JVMs support this, at least as an
option, on most platforms).
ECL: not sure. The multithreading support still seems experimental
(or at least underdocumented :-) so I suggest contacting the
developers for more information, if you are interested.
Christophe
Christophe Rhodes wrote:
> ·······@corporate-world.lisp.de" <······@corporate-world.lisp.de> writes:
>
> > Christophe Rhodes wrote:
> >> ·······@corporate-world.lisp.de" <······@corporate-world.lisp.de> writes:
> >>
> >> > Currently OpenMCL and Scieneer CL
> >>
> >> ... and SBCL, and ABCL, and (I believe) ECL ...
> >
> > Interesting. Are we really talking about the same thing?
> >
> > * running native threads in parallel on multi-processor machines?
>
> Yes.
>
> > For example LispWorks has some kind of native threads on several
> > platforms, but on multi-cpu/core machines, only one thread
> > will execute at a time. Thus only one cpu/core can be used at a time.
> >
> > On which platforms would SBCL/ABCL/ECL support threads running
> > in parallel on multi-cpu/cores?
>
> SBCL: x86 Linux, Mac OS and Solaris; x86-64 Linux
Thanks for mentioning it.
I just checked it out and SBCL on my MacBook under Mac OS X
indeed uses both cores. That's mighty cool.
Though under Parallels/Linux on the MacBook I can only use one core.
>
> ABCL: anywhere, where the underlying JVM's threads being used are
> native. (I believe all the current JVMs support this, at least as an
> option, on most platforms).
>
> ECL: not sure. The multithreading support still seems experimental
> (or at least underdocumented :-) so I suggest contacting the
> developers for more information, if you are interested.
>
> Christophe
·······@corporate-world.lisp.de" <······@corporate-world.lisp.de> writes:
> I just checked it out and SBCL on my MacBook under Mac OS X
> indeed uses both cores. That's mighty cool.
>
> Though under Parallels/Linux on the MacBook I can only use one core.
Since Parallels only makes one core available to the guest OS, that's
not surprising :-)
llothar wrote:
> ··············@gmail.com wrote:
>
> > It's Scheme, not Common Lisp, but has support for lots of systems and
> > includes threads. It compiles to native code, the speed of which it's
> > admirers boast somewhat obnoxiously. :)
>
> No it does not have native threads, just unuseable cooperate threads.
My recollection of Bigloo is that it has a cooperative threading system
as part of the Bigloo implementation, and this is supported the best,
but that it also has some kind of native system threads, which don't
receive as much TLC. So you might want to inspect what Bigloo actually
has.
> Seems that there is only the expensive way of LispWorks or ACL (summing
> up to 5000 US$). Free Lisp systems still are do not meet my goals.
That's why I went for Chicken Scheme. There really isn't a good open
source compiled Common Lisp on Windows at present.
http://www.call-with-current-continuation.org
Cheers,
Brandon Van Every
On 2006-08-02, llothar <·······@web.de> wrote:
> Hello,
>
> is there any good modern Lisp that works on Unix (MacOSX, Solaris,
> Linux) and Windows 2000/XP ? I would also need real operating level
> concurrent threading and a good integrated compiler.
Seems like a long list of things for one implementation to do.
If there is one implementation of any language that can do it
is probably best to stick with that language.
> > is there any good modern Lisp that works on Unix (MacOSX, Solaris,
> > Linux) and Windows 2000/XP ? I would also need real operating level
> > concurrent threading and a good integrated compiler.
>
> Seems like a long list of things for one implementation to do.
>
> If there is one implementation of any language that can do it
> is probably best to stick with that language.
At the moment i maintain my own branch of the GNU (Smart)Eiffel
compiler.
So i can tell you, no it is not really a long list.
But if the existing systems are all designed 10 years ago and never
refactored
then indeed it is a problem. But the OS core of all the systems
mentioned is not that different any more.
Have you considered an alternate approach, such as multiple processes
using shared memory or some other form of shared communication? It's a
bit heavier and not as elegant, but it could be more portable across
the different Lisp implementations.