I must say that I'm not particularly fond of Windows, but as a
programmer I must admit that given that it is on 95% of the
worlds computers ( sans embedded stuff ) and must therefore
be "respected" as a development platform. ( Besides I must admit that
the graphics look better. I presume this is because Windows
must look OK on 800X680 on 14". So it looks even better at higher
resolutions on my 19" screen. Whereas the opposite is true of X
which was designed to look good on a screen at least 21". )
I hear many people complain that there are no free Lisp
implementations that run on both Linux and Windows.
Looking around there is really only one free implementation that runs
on Windows and that's CLisp. UNIX/LINUX has a wealth of free
implementations.
So given the size of the market, why do Lisp implementors "dis"
Windows?
"Thaddeus L Olczyk" <······@interaccess.com> wrote in message
·······················@nntp.interaccess.com...
> So given the size of the market, why do Lisp implementors "dis"
> Windows?
The simple answer is that (hopefully) they do not just because its Windows.
At least, I do not. Much of the complaining about Windows is its
unreliability and crashes that freeze the entire machine. This is the only
reason I would have for not using Windows. My Windows machine does not
freeze (much) but it still does (Windows 98). If you need something to run
for long periods of time this is definitely a problem. One has to put in
prespective though, Windows runs on the most types of hardware and there are
literally thousands of subtypes of PC. With so much to support it is no
wonder than some device drivers may be flaky, or that the hardware is
marginal or that 10% of the apps are buggy. Even the commercial Lisp
implementations have bugs.
Getting a reliable machine/OS combination is not easy. Microsoft is trying
but it has many things working against stability, the first is flaky
hardware. PC hardware is getting better, but simple things like boards or
memory physically coming loose, RAM manufactuers not rejecting marginal
memory all adds up to grief. Pile on that programs with memory leaks and a
large group of hackers who are constantly trying to break your machines,
well.... DUH!
I have come to the conclusion that a computer are best as single application
machines. You do not have to worry about other applications hogging
resources.
These problems are not limited to MS/Windows but other systems as well. To
get the really reliable systems, it takes (or used to take) $$$$$$, and
maybe settling for older hardware that is tried and tested.
Wade
······@interaccess.com (Thaddeus L Olczyk) writes:
>
> I hear many people complain that there are no free Lisp
> implementations that run on both Linux and Windows.
> Looking around there is really only one free implementation that runs
> on Windows and that's CLisp. UNIX/LINUX has a wealth of free
> implementations.
where free = open source, non-commercial
Both Xanalys and ACL have free versions for both Linux and Windows.
> So given the size of the market, why do Lisp implementors "dis"
> Windows?
My guess, and this is just speculation, is age. Most of the "free"
Lisps were written, or are based on code that was written, when home
PCs were a marginal market and all the interesting stuff was happening
on Unix. (To be fair, interesting stuff was happening on other OSes
that have since faded into the sunset.) Add to this the Linux
phenomena: a large number of bright college students and
experimenters, the people who are generally the most likely to
implement free software, are now using Linux.
So, Windows's window of opportunity for a free Lisp was probably from
about 1990 (or so) to 1997, fitting in between the "death of Unix" (or
more realistically, the ascention of the home PC) and the "birth of
Linux" (when the world really started to hear about it,
anyway). Since, in my experience, most Windows users/programmers
during that period had no real concept of free software, I'm not
surprised that it didn't happen.
Thom
--
(let ((e-mail-address ····@IXG.IUS")) (loop with new-string =
(make-string (length e-mail-address)) for count from 0 to (1- (length
e-mail-address)) for char-code = (char-code (aref e-mail-address
count)) for new-char-code = (if (and (> char-code 64) (< char-code
123)) (+ (mod (+ 13 char-code) 52) 65) char-code) do (setf (aref
new-string count) (code-char new-char-code)) finally (return
new-string)))
Thom Goodsell <·········@DESPAM.cra.com> writes:
> So, Windows's window of opportunity for a free Lisp was probably from
> about 1990 (or so) to 1997, fitting in between the "death of Unix" (or
> more realistically, the ascention of the home PC) and the "birth of
> Linux" (when the world really started to hear about it,
> anyway). Since, in my experience, most Windows users/programmers
> during that period had no real concept of free software, I'm not
> surprised that it didn't happen.
1990? Windows 3 and DOS? I wouldn't imagine people to take PC's
seriously until at least 1993 and the first release of NT.
--
/|_ .-----------------------.
,' .\ / | No to Imperialist war |
,--' _,' | Wage class war! |
/ / `-----------------------'
( -. |
| ) |
(`-. '--.)
`. )----'
Thaddeus L Olczyk wrote:
>
> I must say that I'm not particularly fond of Windows, but as a
> programmer I must admit that given that it is on 95% of the
> worlds computers ( sans embedded stuff ) and must therefore
> be "respected" as a development platform.
As a /deployment/ platform, perhaps, but for development
it's a disaster, IMHO.
> ( Besides I must admit that
> the graphics look better. I presume this is because Windows
> must look OK on 800X680 on 14". So it looks even better at higher
> resolutions on my 19" screen. Whereas the opposite is true of X
> which was designed to look good on a screen at least 21". )
Dunno about that - I've two identical machines here, Linux and NT,
and the graphics look much the same to me.
>
> I hear many people complain that there are no free Lisp
> implementations that run on both Linux and Windows.
> Looking around there is really only one free implementation that runs
> on Windows and that's CLisp. UNIX/LINUX has a wealth of free
> implementations.
>
> So given the size of the market, why do Lisp implementors "dis"
> Windows?
I can't answer for anyone else, but if /I/ were developing a Lisp
system, and intending to give it away for free, I would /certainly/
not target an operating system that made my life harder. To put
up with that sort of pain, someone needs to offer me money.
······@interaccess.com (Thaddeus L Olczyk) writes:
> I must say that I'm not particularly fond of Windows, but as a
> programmer I must admit that given that it is on 95% of the
> worlds computers ( sans embedded stuff ) and must therefore
> be "respected" as a development platform. ( Besides I must admit that
> the graphics look better. I presume this is because Windows
> must look OK on 800X680 on 14". So it looks even better at higher
> resolutions on my 19" screen. Whereas the opposite is true of X
> which was designed to look good on a screen at least 21". )
>
> I hear many people complain that there are no free Lisp
> implementations that run on both Linux and Windows.
> Looking around there is really only one free implementation that runs
> on Windows and that's CLisp. UNIX/LINUX has a wealth of free
> implementations.
>
> So given the size of the market, why do Lisp implementors "dis"
> Windows?
How can there be a market for something that is free?
Some of the commercial Lisp implementors have products for the
Windows market, and they seem to be as serious about that platform as
the Unix platforms.
One reason that few of the "free" Lisp implementations support
the PC platform is that it's only recently that the PC platform has
become viable - a Lisp implementation needs more support than a
working printf().
I suspect that most of the people involved with opensource
projects *in general* have a greater affinity for free platforms
(which is mainly Unixalikes). Since they are not charging for what
they produce, you do not have *any* right to expect them to provide
something that runs on the platform *you* prefer.
On the other hand, you are entirely free to port one of the
opensource Lisp implementations to whatever platform you want.
--
Raymond Wiker Mail: ·············@fast.no
Senior Software Engineer Web: http://www.fast.no/
Fast Search & Transfer ASA Phone: +47 23 01 11 60
P.O. Box 1677 Vika Fax: +47 35 54 87 99
NO-0120 Oslo, NORWAY Mob: +47 48 01 11 60
Try FAST Search: http://alltheweb.com/
······@interaccess.com (Thaddeus L Olczyk) writes:
> I must say that I'm not particularly fond of Windows, but as a
> programmer I must admit that given that it is on 95% of the
> worlds computers ( sans embedded stuff ) and must therefore
> be "respected" as a development platform. ( Besides I must admit that
While Windows runs on most desktop machines, the ratio is quite a bit
different on servers. Given the continuing trend of putting most
application intelligence onto servers, that's where lots of
complicated programming tasks are to be found, and hence CL shines[1].
Hence the focus of many _developers_ on non-Windows platforms, and
this of course influences CL implementors.
That said, both Franz and Xanalys provide full-fledged CL
implementations for Windows, and they also provide their free
offerings on those platforms.
> I hear many people complain that there are no free Lisp
> implementations that run on both Linux and Windows.
Depends on how you quantify "many". There is occasional interest in
e.g. CMU CL for Windows, but it is hardly a constant theme. People
who are interested in free stuff are often also interested in free
OSs... [ for whatever definition of free you like ].
> Looking around there is really only one free implementation that runs
> on Windows and that's CLisp. UNIX/LINUX has a wealth of free
> implementations.
>
> So given the size of the market, why do Lisp implementors "dis"
> Windows?
As someone else has said, from the POV of the implementor, there is no
"market" for free stuff, unless he also offers something else (a
service, a non-free implementation), that people will then upgrade to
that will bring in revenues.
As to CMU CL: None of the core developers and contributors seem overly
interested in CMU CL on Windows, and no one else has yet stepped up to
the plate to do the port. Doing the port is not going to be trivial
(though it is likely to be much less effort than porting CMU CL to a new
processor), and it likely isn't going to be interesting (who wants to
spend lots of hours trying to find out whether a given function really
conforms to POSIX, or silently follows Microsoft's own interpretation),
so no one is going to do it just for fun. Maybe if one of our customers
forced[2] us to support Windows as a server platform, I'd do the port,
but then again we might deliver on ACL for Windows instead.
As it stands, I work on CMU CL because it is fun[3], and because as
the current platform of choice of our business, we profit from a
lively CMU CL user and developer community. As part of this I'm
prepared to do the occasional port to systems I don't personally use
(much), like e.g. the OpenBSD/NetBSD mini-ports, if it's not too much
work. But Windows is just too much hassle for me to invest lots of
time without pay.
Of course, if a user was really seriously interested in doing the port
himself, I'm sure people would support him in this endeavour. It's
just that noone has yet showed that kind of interest.
Regs, Pierre.
Footnotes:
[1] That's not to say that there aren't really interesting
client-side problems where CL can also shine. But for lots of
client-side stuff, things like Delphi are quite capable enough.
[2] Although I can't think of us agreeing to that sort of deal, since
we are responsible for the maintenance and uptime of the given
servers, and I don't want that kind of responsibility for Windows
servers.
[3] Working on CMU CL is a nice contrast to working on simulation
software, because it has a much more direct feedback loop, and
you don't have to work with hundreds of megabytes of data, and
probably for lots of other reasons I'll have to explore more
deeply. It's just a very different.
--
Pierre R. Mai <····@acm.org> http://www.pmsf.de/pmai/
The most likely way for the world to be destroyed, most experts agree,
is by accident. That's where we come in; we're computer professionals.
We cause accidents. -- Nathaniel Borenstein
On 22 Jan 2002 15:21:22 +0100, "Pierre R. Mai" <····@acm.org> wrote:
>While Windows runs on most desktop machines, the ratio is quite a bit
>different on servers. Given the continuing trend of putting most
>application intelligence onto servers, that's where lots of
>complicated programming tasks are to be found, and hence CL shines[1].
The trend for servers however is to do more stuff using Java, JSP and
Perl ( and some Ruby )
CL does not even rate a mention.
------------------------------------------------------------
Add this to your .sig. Highly effective at reducing spam !
Please direct all spam to:
·········@whitehouse.gov ··············@whitehouse.gov
·····@cia.gov ·····@fbi.gov ·······@spamcop.net ,
······@liberal.org.au ,···@chq.liberal.org.au,
·······@onenation.com.au
On 22 Jan 2002 15:21:22 +0100, "Pierre R. Mai" <····@acm.org> wrote:
>While Windows runs on most desktop machines, the ratio is quite a bit
>different on servers. Given the continuing trend of putting most
>application intelligence onto servers, that's where lots of
>complicated programming tasks are to be found, and hence CL shines[1].
The trend for servers however is to do more stuff using Java, JSP and
Perl ( and some Ruby )
CL does not even rate a mention.
------------------------------------------------------------
Add this to your .sig. Highly effective at reducing spam !
Please direct all spam to:
·········@whitehouse.gov ··············@whitehouse.gov
·····@cia.gov ·····@fbi.gov ·······@spamcop.net ,
······@liberal.org.au ,···@chq.liberal.org.au,
·······@onenation.com.au
Thaddeus L Olczyk wrote:
> I must say that I'm not particularly fond of Windows, but as a
> programmer I must admit that given that it is on 95% of the
> worlds computers ( sans embedded stuff ) and must therefore
> be "respected" as a development platform.
That is true, as far as it goes. However, which Windows are you
talking about. MS have a policy of breaking OS compatability
every could of years.
Then on what OS does much of the worlds *interesting* computation
happen? By this I really want to count everything that is not
writing email, documents and the ilk of trivial desktop stuff. I
know that the percentage machines doing interesting stuff on
windows is no where near as high as 95%.
[...elided some stuff about graphics...]
Why this obsession with graphics? My first computer work was for
code that ran on an RS2000 on vt100 terminals. The development
environment was at vt320 (24x80 lines) and vi. This was a bit
painful but you do not have to have pretty pictures to get stuff
done.
Hmmm. Maybe I have just answered your question. For me, I dis
"windows" because I see no absolute need for Windows.
:)w
In article <··················@nntp.interaccess.com>, Thaddeus L Olczyk wrote:
>I must say that I'm not particularly fond of Windows, but as a
>programmer I must admit that given that it is on 95% of the
>worlds computers ( sans embedded stuff ) and must therefore
>be "respected" as a development platform.
Dust mites are in about 100% of the world's used mattresses.
Windows is a steaming pile of excrement. Everyone with a clue knows this,
not just Lisp implementors or users.
>I hear many people complain that there are no free Lisp
>implementations that run on both Linux and Windows.
You are forgetting that the huge Windows user base (``95% of
the word's computers'' are not looking for a free Lisp, or a lisp
of any kind. The majority of those of them who speak English think that
lisp is a speech impediment.
>Looking around there is really only one free implementation that runs
>on Windows and that's CLisp. UNIX/LINUX has a wealth of free
>implementations.
>
>So given the size of the market, why do Lisp implementors "dis"
>Windows?
You are talking about free implementations, and then about market in
the next breath. Are you dense or what? Free software is written by
people who choose what they think are the most pleasant, convenient
and reliable tools and platforms to work with.
Windows is something one is paid to do. Implementors of *proprietary*
Lisps aren't shying away from Windows! Off the top of my head, you have
Xanalys's LispWorks, you have Franz's Allegro, you have Corman.
Most reasonable people don't voluntarily subject themselves to
torture. Those who do are called masochists, but even masochists can only
take so much before they invoke the protocol for stopping the torture.
Usually when people have to do something painful or unpleasant, they want
compensation.
Mind you, there is some free software for Windows, like Cygwin, WinCVS,
Vim, Emacs, and others. There is a pattern: most of these exist for the
sake of providing a reasonable environment within an intellect-hostile
platform. Developres and porters suffered so that others would suffer
less.
Another consideration is that Windows does not implement any
industry-standard programming interfaces out of the box. If you write
POSIX code, you can target a whole bunch of operating systems, but not
Windows. At least not without installing a whole POSIX environment.
Microsoft's choice to go with entirely proprietary interfaces all the
way through the system makes it a huge chore to port software to Windows.
That these interfaces are broken or badly designed, and that they
are moving targets doesn't help either.
So, in summary, Windows doesn't attract free software development because
it is of low quality, awkwardly designed, non-free, and nonstandard.
Why do you continue bitching about this anyway? If you don't mind buying
and using a proprietary operating system, why don't you also buy and
use a proprietary Lisp to go with it?