thinlisp-develop — All things ThinLisp discussed here.

[ThinLisp-develop] Declaration error while compiling

[Anil G. <ani...@gm...>]

Hello,

I'm trying to run thinlisp but I'm facing an error while compiling the
system. The part that defines type declarations in decls.lisp gives
errors on SBCL, CCL and CLISP, all to the effect of "Cannot define
declaration ARRAY because it is the name of a type". The error seems
to be stemming from the (define-declaration array ...) call, which
calls (declaim (declaration array)) in its expansion, causing an error
since array defines a type already.

Is this expected or am I doing the setup wrong? I'd appreciate any
guidance on fixing this: I'm assuming that this is done to translate
type declarations present in the input code, would it be fine to
disable these declarations and use the other ones defined in the tl
package instead (eg. tl:array)?

Best Regards,
Anıl

[ThinLisp-develop] test

[Ben H. <bh...@po...>]

this is a test

Re: [ThinLisp-develop] How thin is ThinLisp?

[Jim A. <ja...@al...>]

Hello,

To answer your first question, yes.  Thinlisp is still alive, though not by
much.  I am still working on lexical closures in my spare time, but as you
can see I haven't been devoting much spare time to it recently.

To answer your second question, no.  Thinlisp is not small enough to run on
your Z80.  The sample application to merely echo command line arguments to
standard output compiles down to 73 Kbytes of compiled code on an Intel 386
Linux box.  By default Thinlisp comes up consuming about 450 Kbytes of RAM
just to initialize all the symbols and packages.  So, even if the more
compact instruction set of the Z80 reduced code size enough, the RAM needs
are insurmountable.

You might be able to use Thinlisp to translate isolated functions from Lisp
and then hand-tweak the resulting C code to remove dependencies on
Thinlisp's support libraries, but at that point I'd start asking if this was
worth the trouble.

Sorry, but a micro like the Z80 is below the size we were targeting by
several orders of magnitude.

Thanks for your interest in Thinlisp, and best of luck to you.

Cheers,

Jim

/---------------------------------------------------------
Jim Allard | ja...@al... | http://www.jallard.org
 
-----Original Message-----
Sent: Monday, June 19, 2006 6:38 AM
To: thi...@li...
Subject: [ThinLisp-develop] Is it still alive?

I see a CVS commit in 2005, the last release is from 2001.
Is ThinLisp still alive?

Philipp

-----Original Message-----
Sent: Monday, June 19, 2006 7:06 AM
To: thi...@li...
Subject: [ThinLisp-develop] How thin is ThinLisp?

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

My target platform is a Z80 at 3.58 Mhz with 1KB of RAM and 32 KB of ROM.
Is Thin Lisp small enough for this?

Philipp
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFEloT6btUV+xsoLpoRAqF2AJ9d4/yP+Uc7ktvwuZ2xcu82P8osxQCg3E6O
zhMZc3qnLQ8P8Oleg5YnYdA=
=0fl/
-----END PGP SIGNATURE-----


_______________________________________________
ThinLisp-develop mailing list
Thi...@li...
https://lists.sourceforge.net/lists/listinfo/thinlisp-develop

[ThinLisp-develop] How thin is ThinLisp?

[Philipp K. K. <pk...@sp...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

My target platform is a Z80 at 3.58 Mhz with 1KB of RAM and 32 KB of ROM.
Is Thin Lisp small enough for this?

Philipp
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFEloT6btUV+xsoLpoRAqF2AJ9d4/yP+Uc7ktvwuZ2xcu82P8osxQCg3E6O
zhMZc3qnLQ8P8Oleg5YnYdA=
=0fl/
-----END PGP SIGNATURE-----

[ThinLisp-develop] Is it still alive?

[Philipp K. K. <pk...@sp...>]

I see a CVS commit in 2005, the last release is from 2001.
Is ThinLisp still alive?

Philipp

[ThinLisp-develop] FASD project: Online survey launched

[Benno L. <ben...@id...>]

Dear Open Source developer

I am doing a research project on "Fun and Software Development" in which I
kindly invite you to participate.
You will find the online survey under http://fasd.ethz.ch/qsf/. The
questionnaire consists of 53 questions and you will need about 15 minutes to
complete it.

With the FASD project (Fun and Software Development) we want to define the
motivational significance of fun when software developers decide to engage
in Open Source projects. What is special about our research project is that
a similar survey is planned with software developers in commercial firms.
This procedure allows the immediate comparison between the involved
individuals and the conditions of production of these two development
models. Thus we hope to obtain substantial new insights to the phenomenon of
Open Source Development.


With many thanks for your participation,
Benno Luthiger


PS:
The results of the survey will be published under
http://www.isu.unizh.ch/fuehrung/blprojects/FASD/.
We have set up the mailing list fa...@we... for this study. Please
see http://fasd.ethz.ch/qsf/mailinglist_de.html for registration to this
mailing list.

_______________________________________________________________________

Benno Luthiger
Swiss Federal Institute of Technology Zurich
8092 Zurich

Mail: benno.luthiger(at)id.ethz.ch
_______________________________________________________________________

Re: [ThinLisp-develop] Problem with CMU CL, current in Debian/unstable

[Jim A. <ja...@ir...>]

Thanks for the report.  I'll take a look at your patch this weekend.

I'd love to hear about your experiences with ThinLisp.

Jim

At 09:27 AM 2/27/2003 +0100, Martin ``rydis'' Rydstr|m wrote:
>Hi,
>
>I was having some trouble with ThinLisp in the CMU CL currently in
>Debian's unstable version, a somewhat patched 18d.
>
>To get stuff to build, I applied the following patch.
>
>It seems to work, at least to build the lecho and pthread system.
>
>---
>Index: src/tlt/lisp/boot.lisp
>===================================================================
>RCS file: /cvsroot/thinlisp/thinlisp-1.0/src/tlt/lisp/boot.lisp,v
>retrieving revision 1.27
>diff -r1.27 boot.lisp
>266c266
><         #+lucid
>---
> >         #+(or lucid cmu)
>271c271
><         #-lucid
>---
> >         #-(or lucid cmu)
>Index: src/tlt/lisp/system.lisp
>===================================================================
>RCS file: /cvsroot/thinlisp/thinlisp-1.0/src/tlt/lisp/system.lisp,v
>retrieving revision 1.19
>diff -r1.19 system.lisp
>355c355
><   #+lucid
>---
> >   #+(or lucid cmu)
>357c357
><   #+lucid
>---
> >   #+(or lucid cmu)
>366c366
><   #-lucid
>---
> >   #-(or lucid cmu)
>---
>
>Regards,
>
>'mr
>
>--
>[Emacs] is written in Lisp, which is the only computer language that is
>beautiful.  -- Neal Stephenson, _In the Beginning was the Command Line_
>
>
>-------------------------------------------------------
>This SF.net email is sponsored by: Scholarships for Techies!
>Can't afford IT training? All 2003 ictp students receive scholarships.
>Get hands-on training in Microsoft, Cisco, Sun, Linux/UNIX, and more.
>www.ictp.com/training/sourceforge.asp
>_______________________________________________
>ThinLisp-develop mailing list
>Thi...@li...
>https://lists.sourceforge.net/lists/listinfo/thinlisp-develop


/------------------------------------------------------------------\
Jim Allard                                        ja...@ir...
iRobot Corporation                                 Industrial Robots
63 South Avenue                                http://www.irobot.com
Burlington, MA  01803          (781) 345-0200x250, FAX (781)345-0201

[ThinLisp-develop] Problem with CMU CL, current in Debian/unstable

[Martin ``rydis'' Rydstr|m <ry...@cd...>]

Hi,

I was having some trouble with ThinLisp in the CMU CL currently in
Debian's unstable version, a somewhat patched 18d.

To get stuff to build, I applied the following patch.

It seems to work, at least to build the lecho and pthread system.

---
Index: src/tlt/lisp/boot.lisp
===================================================================
RCS file: /cvsroot/thinlisp/thinlisp-1.0/src/tlt/lisp/boot.lisp,v
retrieving revision 1.27
diff -r1.27 boot.lisp
266c266
<         #+lucid
---
>         #+(or lucid cmu)
271c271
<         #-lucid
---
>         #-(or lucid cmu)
Index: src/tlt/lisp/system.lisp
===================================================================
RCS file: /cvsroot/thinlisp/thinlisp-1.0/src/tlt/lisp/system.lisp,v
retrieving revision 1.19
diff -r1.19 system.lisp
355c355
<   #+lucid
---
>   #+(or lucid cmu)
357c357
<   #+lucid
---
>   #+(or lucid cmu)
366c366
<   #-lucid
---
>   #-(or lucid cmu)
---

Regards,

'mr

-- 
[Emacs] is written in Lisp, which is the only computer language that is
beautiful.  -- Neal Stephenson, _In the Beginning was the Command Line_

Re: [ThinLisp-develop] Re: Autoconf Work

[Ben H. <bh...@po...>]

Jim Allard wrote:
 > I noticed that even when you give the "OPT=1" argument to make, the 
 > autoconf makefile includes "-g" in the CFLAGS. ...  I don't know if this is 
 > configurable in autoconf.

Nope, or so it seems.  I invoke AC_PROG_CC to get @CC@ set up.
It falls thru into a tangle that decides if -g is approprate,
in particular AC_PROG_CC_G (which was renamed _AC_PROG_CC_G in
later versions of autoconf making it hard to hack around).

The extent of it's definition of '-g is approprate' is that
the compiler is sufficently like a gcc.  This is all in
/sw/share/autoconf/acspecific.m4 on my powerbook.

I'm confused that they are so casual and insistent about
pushing it into CFLAGS.

[ThinLisp-develop] Re: Autoconf Work

[Jim A. <ja...@al...>]

I gave it a try.  During the C link it couldn't find the symbols floor, 
ceil, fmod, pow, and log10.  So, I presume that we need to tell autoconf 
that we want the floating point functions so it includes the math 
library.  Hopefully you've got the information about how to do that in your 
hip pocket.

I noticed that even when you give the "OPT=1" argument to make, the 
autoconf makefile includes "-g" in the CFLAGS.  While I know that some 
people insist that a debuggable optimized compile is OK in gcc, I've found 
some performance and size penalties that seemed troubling.  Also, gdb lies 
about source code locations in optimized debuggable code, since code motion 
done in optimization makes the source line locations somewhat 
meaningless.  I guess I've just not found that there's a reasonable 
compromise implemented in the combination of debuggable and optimized, so 
I've generally avoided the attempt to mix them.  I don't know if this is 
configurable in autoconf.

I've been reconsidering the whole C build directory structure, and changing 
that during a move towards autoconf might be reasonable.  The idea of build 
directories that are parallel to the C directories instead of 
subdirectories of C is kind of strange in retrospect.

The port that still is lacking testing is the makefile support for NMAKE in 
Windows.  I don't know if autoconf can help there.

Anyhow, it's a nice start on autoconf.

Jim

P.S.  Sorry it took so long to turn around testing results on this.  Sleep 
is overrated. :-)


At 12:28 AM 1/2/2002 -0500, Ben Hyde wrote:

>Ben Hyde wrote:
>  > Log Message:
>  > Autoconf style makefiles work on at least mac os x, and probably most
>  > other things too.
>
>Well that's a fib.  If these were done right the
>configure.in file (i.e. the thing that autoconf
>reads to generate that fine ./configure script)
>would be rolled up from knowledge gleaned from the
>declare-system forms.  Instead I just generate
>a default configure.in file that happens to work
>with the makefile.in that is generated for the
>pseudo port-name "autoconf".
>
>This is close enought that it is tempting to
>eliminate all the makefile-<port-name> makefiles,
>but that would require getting the opt/bin etc.
>stuff figured out.
>
>For the time being to use this you do:
>
>   cd lecho/c
>   autoconf
>   ./configure
>   cd ../opt
>   make -f ../makefile
>
>that should work on most platforms, maybe...
>
>I'm going to bed.
>
>_______________________________________________
>ThinLisp-develop mailing list
>Thi...@li...
>https://lists.sourceforge.net/lists/listinfo/thinlisp-develop

/--------------------------------------------------------------\
Jim Allard, Newton, MA                      ja...@al...

[ThinLisp-develop] Re: [ThinLisp-commits] CVS: thinlisp-1.0/src/tl/c makefile-cygnus,1.8,1.9 makefile-freebsd,1.8,1.9 makefile-linux,1.8,1.9 makefile-macosx,1.1,1.2

[Ben H. <bh...@po...>]

Ben Hyde wrote:
 > Log Message:
 > Copyright date changes, ought to get all these...

Bleck...  This is forced me to commit a change
to the c-names I'd hoped to avoid committing; and
then i didn't finish getting the copyright change
spread all over before I called it quit for the 
night.

[ThinLisp-develop] Re: [ThinLisp-commits] CVS: thinlisp-1.0/src/tlt/lisp system.lisp,1.18,1.19 makefiles.lisp,1.16,1.17

[Ben H. <bh...@po...>]

Ben Hyde wrote:
 > Log Message:
 > Autoconf style makefiles work on at least mac os x, and probably most
 > other things too.

Well that's a fib.  If these were done right the
configure.in file (i.e. the thing that autoconf
reads to generate that fine ./configure script)
would be rolled up from knowledge gleaned from the
declare-system forms.  Instead I just generate
a default configure.in file that happens to work
with the makefile.in that is generated for the
pseudo port-name "autoconf".

This is close enought that it is tempting to
eliminate all the makefile-<port-name> makefiles,
but that would require getting the opt/bin etc.
stuff figured out.

For the time being to use this you do:

  cd lecho/c
  autoconf
  ./configure
  cd ../opt
  make -f ../makefile

that should work on most platforms, maybe...

I'm going to bed.