Does anyone know if there is still any work goin on in the LispOS project?
If not, why not?
I've been learning as much as I can to help, but I'm afraid
I'm pretty young in LISP. I've started writing a Persistant Object Storage
system for use with LispOS, but I'm afraid without any knowledge of the rest
of the project it will be pretty useless (it might be anyhow, but the
probability increases enourmously without knowledge. :)
Any word would be greatly apreciated. I'm sick of fighting my OS
for power of my computer :)
Gavin E. Gleason
In article <··············@mail.unm.edu>, ········@tvi.cc.nm.us (Gavin E. Gleason) writes:
>>
>> Does anyone know if there is still any work goin on in the LispOS project?
>> If not, why not?
>>
>> I've been learning as much as I can to help, but I'm afraid
>> I'm pretty young in LISP. I've started writing a Persistant Object Storage
>> system for use with LispOS, but I'm afraid without any knowledge of the rest
>> of the project it will be pretty useless (it might be anyhow, but the
>> probability increases enourmously without knowledge. :)
>> Any word would be greatly apreciated. I'm sick of fighting my OS
>> for power of my computer :)
>>
I have continued my own efforts in this direction.
The "LispOS project" as was discussed a while ago on the LispOS mailing list
seems quite dead. IMNSHO, it seemed most everyone wanted to work on
the very low-level implementation details (kernel hacking, the
persistent object system (witness yourself!)), few of which had any
experience doing so. While the low-levels implementation is important,
it is COMPLETELY USELESS without all the higher level,
utilities and APPLICATIONS which are ultimately what
makes an OS useful and worth something.
And such utilites can be created WITHOUT a LispOS,
so there is no "excuse" not to create them FIRST.
After all, this is exactly how Linux become an OS.
All the years of work by the GNU people BEFORE
an actual kernel was ever worked on.
I recall few, if anyone, was willing to volunteer for this work.
-Kelly Murray ···@franz.com
From: Matthias Hoelzl (tc)
Subject: Re: What ever happened to LispOS?
Date:
Message-ID: <87iutyvv0i.fsf@gauss.muc.de>
···@franz.com (Kelly Murray) writes:
> The "LispOS project" as was discussed a while ago on the LispOS mailing list
> seems quite dead. IMNSHO, it seemed most everyone wanted to work on
> the very low-level implementation details (kernel hacking, the
> persistent object system (witness yourself!)), few of which had any
> experience doing so. While the low-levels implementation is important,
> it is COMPLETELY USELESS without all the higher level,
> utilities and APPLICATIONS which are ultimately what
> makes an OS useful and worth something.
Very true. I think another problem was that many people had rather
strong opinions about some technical matters (use Common Lisp / Scheme
/ design a new Lisp dialect, and so on) and debates about them were
more intense than discussions about what kind of system should be
built.
> And such utilites can be created WITHOUT a LispOS,
> so there is no "excuse" not to create them FIRST.
> After all, this is exactly how Linux become an OS.
> All the years of work by the GNU people BEFORE
> an actual kernel was ever worked on.
> I recall few, if anyone, was willing to volunteer for this work.
I think we would need some free "high profile" applications written in
Lisp before the idea of a LispOS would gather enough followers to
become viable. (And they would benefit everybody, even if no LispOS
would ever be built.) However the problem with building these
applications is that there are no free Lisp development environments
available that you could use for building the kind of program that the
avarage user would find useful. Of course the commercial Lisps are
excellent environments, but I guess that even many developers using
them at work cannot use them for spare time projects.
So, in my opinion we would have to improve one of the freely available
Lisp environments before we can expect free Lisp based applications to
appear. If enough people are interested we could perhaps recycle the
LispOS mailing list to discuss the development of a free Lisp
development environment.
Matthias
Matthias Hoelzl (tc) wrote in message <··············@gauss.muc.de>...
>However the problem with building these
>applications is that there are no free Lisp development environments
>available that you could use for building the kind of program that the
>avarage user would find useful. Of course the commercial Lisps are
>excellent environments, but I guess that even many developers using
>them at work cannot use them for spare time projects.
If you are willing to run a Lisp OS on top of Linux, then Franz's Allegro
Common Lisp for Linux is *free* for non-commercial software.
As long as all code used to build the system and it's utilities were
distributed as such (i.e., as common lisp source code) all a user would need
to do is
1. Acquire Linux (I believe Franz reccommends Redhat)
2. Acquire ACL for Linux (also free)
3. Compile the lisp source files
and you would have a Lisp OS.
Am I missing something?
Raf
Raf Cavallaro wrote:
> If you are willing to run a Lisp OS on top of Linux, then Franz's Allegro
> Common Lisp for Linux is *free* for non-commercial software.
>
> As long as all code used to build the system and it's utilities were
> distributed as such (i.e., as common lisp source code) all a user would need
> to do is
>
> 1. Acquire Linux (I believe Franz reccommends Redhat)
>
> 2. Acquire ACL for Linux (also free)
>
> 3. Compile the lisp source files
>
> and you would have a Lisp OS.
>
> Am I missing something?
Two things.
Firstly, ACL for Linux is free only in a fairly weak sense. The
licence expires in less than two months, and in any case it doesn't
come with source code. This need not be a fatal objection, but
I think most of the people who were interested in the possibility
of a Lisp OS wanted a genuinely free and open system.
Secondly, it's not clear to what extent one could get a decent
*OS* by doing that. Either you'd just be having a Lisp interface
to Linux semantics (which would rather miss the point), or you'd
be emulating a whole bunch of OS services on top of Linux
semantics (which might be rather inefficient). In any case
you wouldn't get the extreme reflectivity on environment that
was one of the major concerns of most of the Lisp OS people.
(Random example: would it be possible, with this approach, to
sit down in your shell/listener/whateveryoucallit, write a
couple of functions, and thereby make and install a simple
web server?)
--
Gareth McCaughan Dept. of Pure Mathematics & Mathematical Statistics,
·····@dpmms.cam.ac.uk Cambridge University, England.
In <··············@g.pet.cam.ac.uk> Gareth McCaughan <·····@dpmms.cam.ac.uk> writes:
>Raf Cavallaro wrote:
>
>> If you are willing to run a Lisp OS on top of Linux, then Franz's Allegro
>> Common Lisp for Linux is *free* for non-commercial software.
>>
>> Am I missing something?
>
>Two things.
>
>Firstly, ACL for Linux is free only in a fairly weak sense. The
>licence expires in less than two months, and in any case it doesn't
>come with source code. This need not be a fatal objection, but
>I think most of the people who were interested in the possibility
>of a Lisp OS wanted a genuinely free and open system.
If there are people with some spare time left, it would be better to
start implementing a public domain version of CLIM. One of the
nicest things of CL is easy portability of source code among
platforms (actually, usually I have only to change one single
line of code), and the same could be achieved for user interface
using CLIM. However, CLIM is currently only sold by vendors, there
exists no PD implementation of it. The same might be true for CLOS
(I know that most PD CL include PCL, but that's not entirely
the same). Note that ACL does not include CLIM either, although
it is likely that the programmers of Franz have it running on Linux.
So, the major effort should be to develop a public domain ANSI Lisp system,
based on CMU Lisp, CLISP or GCLISP, that offers CLIM as an extension. This
seems to be a project that can be done in a reasonable amount of time.
Peter
--
Peter Lucas
Dept. of Computer Science, Utrecht University
Padualaan 14, 3584 CH Utrecht, The Netherlands
Tel: + 31 30 2534094; E-mail: ·····@cs.ruu.nl
·····@cs.ruu.nl (Peter Lucas) writes:
[...]
>If there are people with some spare time left, it would be better to
>start implementing a public domain version of CLIM.
This already happend, but is sleeping. See
http://www.cons.org/free-clim/
>However, CLIM is currently only sold by vendors, there
>exists no PD implementation of it. The same might be true for CLOS
>(I know that most PD CL include PCL, but that's not entirely
>the same).
Hm, I'd say that PCL isn't much more different from avarage CLOS that
any other vendor implementation.
>Note that ACL does not include CLIM either, although
>it is likely that the programmers of Franz have it running on Linux.
>So, the major effort should be to develop a public domain ANSI Lisp system,
>based on CMU Lisp, CLISP or GCLISP, that offers CLIM as an extension. This
>seems to be a project that can be done in a reasonable amount of time.
But it also isn't what many people on the LispOS want, and that's the
problem.
Martin
--
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
···············@wavehh.hanse.de http://www.cons.org/cracauer/
BSD User Group Hamburg/Germany http://www.bsdhh.org/
········@wavehh.hanse.de (Martin Cracauer) writes:
> >So, the major effort should be to develop a public domain ANSI Lisp system,
> >based on CMU Lisp, CLISP or GCLISP, that offers CLIM as an extension. This
> >seems to be a project that can be done in a reasonable amount of time.
>
> But it also isn't what many people on the LispOS want, and that's the
> problem.
Is it that many people don't want this, or that the most vocal people
don't want it?
On another note, I'm distressed that people started waiting for Flux.
My general rule of thumb is to never wait for software - use only
what's available (fixing, enhancing, adding to it, etc.) Replace it
when the new stuff arrives. I'm afraid that breaking this rule caused
many months of delay on the project. Worse is that we lost steam.
--
Harvey J. Stein
Berger Financial Research
·······@bfr.co.il
In article <··············@gauss.muc.de>,
··@gauss.muc.de (Matthias Hoelzl (tc)) writes:
> So, in my opinion we would have to improve one of the freely available
> Lisp environments before we can expect free Lisp based applications to
> appear. If enough people are interested we could perhaps recycle the
> LispOS mailing list to discuss the development of a free Lisp
> development environment.
>
> Matthias
Some people have been experimenting with threads in the Linux version of
CMUCL. When that gets released, then we might have a lisp implementation to
base some interesting apps/OS work on.
Mike McDonald
·······@sgi.com
·······@SGI.com (Mike McDonald) writes:
>In article <··············@gauss.muc.de>,
> ··@gauss.muc.de (Matthias Hoelzl (tc)) writes:
>> So, in my opinion we would have to improve one of the freely available
>> Lisp environments before we can expect free Lisp based applications to
>> appear. If enough people are interested we could perhaps recycle the
>> LispOS mailing list to discuss the development of a free Lisp
>> development environment.
>>
>> Matthias
> Some people have been experimenting with threads in the Linux version of
>CMUCL. When that gets released, then we might have a lisp implementation to
>base some interesting apps/OS work on.
It's more or less official now. Douglas Crosher implemented thread
support (and better GC and more debugger fixes) for CMUCL/x86. It
isn't in 18a, but you can download it from
ftp://ftp2.cons.org/pub/languages/lisp/cmucl/experimental/
The work is done on FreeBSD, but Linux support files are
available. Let us know how well this works for you.
Martin
--
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
···············@wavehh.hanse.de http://www.cons.org/cracauer/
BSD User Group Hamburg/Germany http://www.bsdhh.org/
········@wavehh.hanse.de (Martin Cracauer) writes:
> It's more or less official now. Douglas Crosher implemented thread
> support (and better GC and more debugger fixes) for CMUCL/x86. It
> isn't in 18a, but you can download it from
> ftp://ftp2.cons.org/pub/languages/lisp/cmucl/experimental/
>
> The work is done on FreeBSD, but Linux support files are
> available. Let us know how well this works for you.
Will that become a part of the mainstream release?
--
Hrvoje Niksic <·······@srce.hr> | Student at FER Zagreb, Croatia
--------------------------------+--------------------------------
Oh lord won't you buy me a color TV...
Hrvoje Niksic <·······@srce.hr> writes:
>········@wavehh.hanse.de (Martin Cracauer) writes:
>> It's more or less official now. Douglas Crosher implemented thread
>> support (and better GC and more debugger fixes) for CMUCL/x86. It
>> isn't in 18a, but you can download it from
>> ftp://ftp2.cons.org/pub/languages/lisp/cmucl/experimental/
>>
>> The work is done on FreeBSD, but Linux support files are
>> available. Let us know how well this works for you.
>Will that become a part of the mainstream release?
Is is part of the development branch now, that means, the next release
after 18a (which isn't official announced for now, either) will have
it.
If people would beat on it (the devlopment branch) and send problem
reports, it will help a lot.
Martin
--
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
···············@wavehh.hanse.de http://www.cons.org/cracauer/
BSD User Group Hamburg/Germany http://www.bsdhh.org/
········@wavehh.hanse.de (Martin Cracauer) writes:
> Is is part of the development branch now, that means, the next release
> after 18a (which isn't official announced for now, either) will have
> it.
>
> If people would beat on it (the devlopment branch) and send problem
> reports, it will help a lot.
Are the extensions portable? CMUCL works on many other OSes (Solaris
and HPUX, most notably).
I'd appreciate if you answer by email (or forward the answer), since
I'm on and off USENET nowadays. :)
--
I refuse to use .sig
········@tvi.cc.nm.us (Gavin E. Gleason) writes:
>Does anyone know if there is still any work goin on in the LispOS project?
>If not, why not?
Too much subprojects to choose from and people don't picking those
required to get the project off the ground, then rejecting to do
anything unless the required parts are done.
Martin
--
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
···············@wavehh.hanse.de http://www.cons.org/cracauer/
BSD User Group Hamburg/Germany http://www.bsdhh.org/
>>>>> "MC" == Martin Cracauer <········@wavehh.hanse.de> writes:
GG> Does anyone know if there is still any work goin on in the LispOS
GG> project? If not, why not?
MC> Too much subprojects to choose from and people don't picking those
MC> required to get the project off the ground, then rejecting to do
MC> anything unless the required parts are done.
I think that's a little harsh. On the mailing lists I subscribe to, I
often see several names I didn't know before partcipating in the
LispOS discussion list. These individuals are doing things like
improving the free Lisp implementations and experimenting with new
operating system technology. For example, I think that a CMU Common
Lisp with threads is a significant step forward.
···@wijiji.santafe.edu (Marcus G. Daniels) writes:
>>>>>> "MC" == Martin Cracauer <········@wavehh.hanse.de> writes:
>GG> Does anyone know if there is still any work goin on in the LispOS
>GG> project? If not, why not?
>MC> Too much subprojects to choose from and people don't picking those
>MC> required to get the project off the ground, then rejecting to do
>MC> anything unless the required parts are done.
>I think that's a little harsh. On the mailing lists I subscribe to, I
>often see several names I didn't know before partcipating in the
>LispOS discussion list. These individuals are doing things like
>improving the free Lisp implementations and experimenting with new
>operating system technology. For example, I think that a CMU Common
>Lisp with threads is a significant step forward.
Yes, sure. I didn't want to imply the LispOS people don't want or are
not able to get something out of the ground, just that they currently
don't do it.
A certain amount of decent tools from outside is needed for a
"Linux-effect". Linus wouldn't have written Linux if he had to rework
gcc in advance.
Martin
--
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
···············@wavehh.hanse.de http://www.cons.org/cracauer/
BSD User Group Hamburg/Germany http://www.bsdhh.org/
In article <·····················@wavehh.hanse.de>,
········@wavehh.hanse.de (Martin Cracauer) writes:
> ········@tvi.cc.nm.us (Gavin E. Gleason) writes:
>
>>Does anyone know if there is still any work goin on in the LispOS project?
>>If not, why not?
>
> Too much subprojects to choose from and people don't picking those
> required to get the project off the ground, then rejecting to do
> anything unless the required parts are done.
>
> Martin
A lot of people who didn't know lisp well nor who knew OS kernels wanted to
invent everything from scratch. That's fine if you want to do it as a
learning exercise but to get something as big as a complete OS off the
ground, you need someone who knows both well to step forward and build the
framework. No one did that ie we had no Linus.
Mike McDonald
·······@sgi.com