From sds@goems.com Wed Aug 12 16:03:56 1998 Received: from smtp0-alterdial.uu.net (smtp0-alterdial.UU.NET [192.48.96.28]) by seagull.cdrom.com (8.8.8/8.6.6) with ESMTP id QAA03602 for ; Wed, 12 Aug 1998 16:03:54 -0700 (PDT) Received: from mute.eaglets.com by smtp0-alterdial.uu.net with ESMTP (peer crosschecked as: [208.235.77.225]) id QQfcee04988; Wed, 12 Aug 1998 23:09:02 GMT Received: (from sds@localhost) by mute.eaglets.com (8.9.1/8.9.1) id TAA09099; Wed, 12 Aug 1998 19:09:01 -0400 To: Bruno Haible Subject: clisp, ffi and blas/lapack Return-Receipt-To: sds@goems.com Reply-To: sds@goems.com X-Disclaimer: You should not expect anyone to agree with me. X-Attribution: Sam X-No-Archive: Yes Mail-Copies-To: never From: Sam Steingold Date: 12 Aug 1998 19:09:01 -0400 Message-ID: Lines: 52 Is it possible to call blas/lapack procedures from CLISP? (I seem to vaguely remember that ffi is limited to C). If yes, has anyone done it? # rpm -qilp blas-2.0-10.i386.rpm lapack-2.0-10.i386.rpm Name : blas Distribution: (none) Version : 2.0 Vendor: (none) Release : 10 Build Date: Mon Aug 10 11:37:13 1998 Install date: (not installed) Build Host: s9412a.steinan.ntnu.no Group : Development/Libraries Source RPM: lapack-2.0-10.src.rpm Size : 1029839 Packager : Trond Eivind Glomsrød (teg@pvv.ntnu.no) URL : http://www.netlib.org/lapack/ Summary : The Basic Linear Algebra libraries for Linux Description : BLAS (Basic Linear Algebra Subprograms) is a standard lib for numerical algebra. It provides a number of basic algorithms for linear algebra, and is fast and well-tested. It is written in FORTRAN 77, and built with egcs. /usr/lib/libblas.a /usr/lib/libblas.so /usr/lib/libblas.so.2 /usr/lib/libblas.so.2.0 /usr/lib/libblas.so.2.0.1 Name : lapack Distribution: (none) Version : 2.0 Vendor: (none) Release : 10 Build Date: Mon Aug 10 11:37:13 1998 Install date: (not installed) Build Host: s9412a.steinan.ntnu.no Group : Development/Libraries Source RPM: lapack-2.0-10.src.rpm Size : 7430256 Packager : Trond Eivind Glomsrød (teg@pvv.ntnu.no) URL : http://www.netlib.org/lapack/ Summary : The LAPACK libraries for Linux. Description : LAPACK (Linear Algebra Package) is a standard lib for numerical linear algebra. It is built on the BLAS, and provides many advanced and well-tested routines. It is coded in FORTRAN 77, and built with egcs. /usr/doc/lapack-2.0 /usr/doc/lapack-2.0/README /usr/lib/liblapack.a /usr/lib/liblapack.so /usr/lib/liblapack.so.2 /usr/lib/liblapack.so.2.0 /usr/lib/liblapack.so.2.0.1 -- Sam Steingold (http://www.goems.com/~sds) running RedHat5.1 GNU/Linux Micros**t is not the answer. Micros**t is a question, and the answer is Linux, (http://www.linux.org) the choice of the GNU (http://www.gnu.org) generation. Why use Windows, when there are Doors? From toy@rtp.ericsson.se Tue Aug 18 05:07:34 1998 Received: from gwa.ericsson.com (gwa.ericsson.com [198.215.127.2]) by seagull.cdrom.com (8.8.8/8.6.6) with ESMTP id FAA10487 for ; Tue, 18 Aug 1998 05:07:33 -0700 (PDT) Received: from mr4.exu.ericsson.se ([138.85.11.56]) by gwa.ericsson.com (8.8.8/8.8.8) with ESMTP id HAA14329 for ; Tue, 18 Aug 1998 07:12:50 -0500 (CDT) Received: from netmanager7.rtp.ericsson.se (netmanager7.rtp.ericsson.se [147.117.132.245]) by mr4.exu.ericsson.se (8.8.8/8.8.8) with SMTP id HAA03466 for ; Tue, 18 Aug 1998 07:12:49 -0500 (CDT) Received: from rcur (wcsdsp4 [147.117.132.216]) by netmanager7.rtp.ericsson.se (8.6.12/8.6.4) with ESMTP id IAA22469 for ; Tue, 18 Aug 1998 08:12:37 -0400 To: CLISP Mailing List Subject: Re: clisp, ffi and blas/lapack Mime-Version: 1.0 (generated by tm-edit 7.108) Content-Type: text/plain; charset=US-ASCII Date: Tue, 18 Aug 1998 08:12:47 -0400 Message-ID: <2778.903442367@rtp.ericsson.se> From: Raymond Toy >>>>> "Sam" == Sam Steingold writes: Sam> Is it possible to call blas/lapack procedures from CLISP? Sam> (I seem to vaguely remember that ffi is limited to C). Sam> If yes, has anyone done it? This doesn't really help you, but if the ffi handles C, you can definitely handle Fortran (on Linux), since C can easily call Fortran. You need to remember these things: 1. All parameters in Fortran are pointers. 2. Fortran names have an underscore appended to them. Thus, if the Fortran function is "daxpy", the C name is "daxpy_". (Or was that prepended?) This should be enough to get blas/lapack hooked up to CLISP. I haven't done this for CLISP, but I've done it for CMUCL. I've done most of BLAS1, some parts of BLAS2, BLAS3, and a very small part of LAPACK. Basically just the parts I needed. Ray From toy@rtp.ericsson.se Tue Aug 18 05:07:55 1998 Received: from gwa.ericsson.com (gwa.ericsson.com [198.215.127.2]) by seagull.cdrom.com (8.8.8/8.6.6) with ESMTP id FAA10504 for ; Tue, 18 Aug 1998 05:07:55 -0700 (PDT) Received: from mr4.exu.ericsson.se ([138.85.11.56]) by gwa.ericsson.com (8.8.8/8.8.8) with ESMTP id HAA14361 for ; Tue, 18 Aug 1998 07:13:18 -0500 (CDT) Received: from netmanager7.rtp.ericsson.se (netmanager7.rtp.ericsson.se [147.117.132.245]) by mr4.exu.ericsson.se (8.8.8/8.8.8) with SMTP id HAA03504 for ; Tue, 18 Aug 1998 07:13:17 -0500 (CDT) Received: from rcur (wcsdsp4 [147.117.132.216]) by netmanager7.rtp.ericsson.se (8.6.12/8.6.4) with ESMTP id IAA22503 for ; Tue, 18 Aug 1998 08:13:05 -0400 To: CLISP Mailing List Subject: Re: compiling on a linux system Mime-Version: 1.0 (generated by tm-edit 7.108) Content-Type: text/plain; charset=US-ASCII Date: Tue, 18 Aug 1998 08:13:16 -0400 Message-ID: <2785.903442396@rtp.ericsson.se> From: Raymond Toy >>>>> "Sam" == Sam Steingold writes: Sam> cvs update Sam> ./configure --build Sam> Any suggestiond what this might mean? Sam> gcc -O -W -Wswitch -Wcomment -Wpointer-arith -Wimplicit -Wreturn-type -fomit-frame-pointer -O2 -fexpensive-optimizations -DDYNAMIC_FFI -x none spvw.o spvwtabf.o spvwtabs.o spvwtabo.o eval.o control.o pathname.o stream.o socket.o io.o array.o hashtabl.o list.o package.o record.o sequence.o charstrg.o debug.o error.o misc.o time.o predtype.o symbol.o lisparit.o graph.o foreign.o unixaux.o ari80386.o modules.o libreadline.a libavcall.a libcallback.a -ldl -o lisp.run Sam> spvw.o: In function `init_language': Sam> spvw.o(.text+0x45a3): the `getwd' function is dangerous and should not be used. I think this means getwd assumes a fixed size buffer. If it's not, it may scribble past the end. We may want to use getcwd instead since it takes a length parameter. Sam> spvw.o: In function `update_linelength': Sam> spvw.o(.text+0x481f): undefined reference to `tgetent' Sam> spvw.o(.text+0x4831): undefined reference to `tgetnum' I think the rest of the errors mean you forgot to link with libtermcap or libcurses. Ray From jay@ruby.kaist.ac.kr Tue Aug 18 08:58:28 1998 Received: from ruby.kaist.ac.kr (ruby.kaist.ac.kr [143.248.190.10]) by seagull.cdrom.com (8.8.8/8.6.6) with ESMTP id IAA11858 for ; Tue, 18 Aug 1998 08:58:27 -0700 (PDT) Received: (from jay@localhost) by ruby.kaist.ac.kr (8.8.8H1/8.8.8) id BAA03256; Wed, 19 Aug 1998 01:03:33 +0900 (KST) To: clisp-list@clisp.cons.org Subject: clisp question X-Url: http://pllab.kaist.ac.kr/~jay Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Reply-to: jay@compiler.kaist.ac.kr From: Jae-Youn Chung Date: 19 Aug 1998 01:03:33 +0900 Message-ID: Lines: 31 Hello clisp users, About a weeks ago, I downloaded the clisp and became the happy user of clisp. But I've got a few problems: in cmulisp: * #2a((1) (2) (3)) #2A((1) (2) (3)) this kind of sharp quote notation works fine. in clisp: > #2a((1) (2) (3)) *** - EVAL: illegal form #2A((1) (2) (3)) 1. Break> Why? And one more question. I can't make regexp library to work with current clisp. Has anybody compiled the clispsrc-egexp.tar.gz? Thanks in advance. -jay -- Chung jay youn mailto:jay@pllab.kaist.ac.kr http://pllab.kaist.ac.kr/~jay From sds@goems.com Tue Aug 18 10:49:54 1998 Received: from smtp0-alterdial.uu.net (smtp0-alterdial.UU.NET [192.48.96.28]) by seagull.cdrom.com (8.8.8/8.6.6) with ESMTP id KAA12440 for ; Tue, 18 Aug 1998 10:49:41 -0700 (PDT) Received: from mute.eaglets.com by smtp0-alterdial.uu.net with ESMTP (peer crosschecked as: [208.235.77.225]) id QQfczn10301; Tue, 18 Aug 1998 17:55:16 GMT Received: (from sds@localhost) by mute.eaglets.com (8.9.1/8.9.1) id NAA03442; Tue, 18 Aug 1998 13:55:13 -0400 To: clisp-list@clisp.cons.org Subject: Re: clisp question References: Return-Receipt-To: sds@goems.com Reply-To: sds@goems.com X-Disclaimer: You should not expect anyone to agree with me. X-Attribution: Sam X-No-Archive: Yes Mail-Copies-To: never From: Sam Steingold In-Reply-To: Jae-Youn Chung's message of "Tue, 18 Aug 1998 08:59:04 -0700 (PDT)" Date: 18 Aug 1998 13:55:12 -0400 Message-ID: Lines: 51 X-Mailer: Gnus v5.5/Emacs 20.2 >>>> In message >>>> Sent on Tue, 18 Aug 1998 08:59:04 -0700 (PDT) >>>> Honorable Jae-Youn Chung writes >>>> on the subject of "clisp question": >> Hello clisp users, >> About a weeks ago, I downloaded the clisp and became the happy user of >> clisp. But I've got a few problems: >> >> in cmulisp: >> * #2a((1) (2) (3)) >> >> #2A((1) (2) (3)) >> >> this kind of sharp quote notation works fine. >> >> in clisp: >> > #2a((1) (2) (3)) >> >> *** - EVAL: illegal form #2A((1) (2) (3)) >> 1. Break> >> >> Why? CLISP starts in USER package. you have to quote #() there: > #2a((1) (2) (3)) *** - EVAL: illegal form #2A((1) (2) (3)) 1. Break> > '#2a((1) (2) (3)) #2A((1) (2) (3)) > (in-package :cl-user) # [COMMON-LISP-USER]> #2a((1) (2) (3)) #2A((1) (2) (3)) [COMMON-LISP-USER]> you can make CLISP start in CL-USER: $ clisp -p CL-USER > *package* # > #2a((1) (2) (3)) #2A((1) (2) (3)) -- Sam Steingold (http://www.goems.com/~sds) running RedHat5.1 GNU/Linux Micros**t is not the answer. Micros**t is a question, and the answer is Linux, (http://www.linux.org) the choice of the GNU (http://www.gnu.org) generation. If brute force does not work, you are not using enough. From lsarasua@epo.e-mail.com Tue Aug 18 23:44:38 1998 Received: from E-MAIL.COM (e-mail.com [204.146.168.195]) by seagull.cdrom.com (8.8.8/8.6.6) with SMTP id XAA15936 for ; Tue, 18 Aug 1998 23:44:37 -0700 (PDT) From: lsarasua@epo.e-mail.com Message-Id: <199808190644.XAA15936@seagull.cdrom.com> Received: by E-MAIL.COM (IBM VM SMTP V2R4a) via spool with SMTP id 1726 ; Wed, 19 Aug 1998 02:50:25 EDT Date: Wed, 19 Aug 1998 02:50:24 EDT To: clisp-list@clisp.cons.org X-Sender-Info: Leo Sarasua Tel. (31) 70 - 340 31 77 The Hague, NL lsarasua@epo.e-mail.com MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Subject: $ sign in path not recognized I am using the DOS version of CLISP. When I try to load a file which is in a s bdirectory whose name contains the $ sign, it returns a syntax error: (load "c:\\dir$\\myfile") --> SYNTAX ERROR Is this normal? I see that for other systems the $ sign is used to indicate th root directory, but for DOS it's a perfectly valid character in a path. Any ideas? Thanks, Leo From bwmott@eos.ncsu.edu Tue Aug 25 07:03:11 1998 Received: from zermelo.csc.ncsu.edu (zermelo.csc.ncsu.edu [152.1.213.201]) by seagull.cdrom.com (8.8.8/8.6.6) with ESMTP id HAA26968 for ; Tue, 25 Aug 1998 07:03:11 -0700 (PDT) From: bwmott@eos.ncsu.edu Received: (from bwmott@localhost) by zermelo.csc.ncsu.edu (8.8.4/UC02Jan97) id KAA21432 for clisp-list@clisp.cons.org; Tue, 25 Aug 1998 10:09:57 -0400 (EDT) Message-Id: <199808251409.KAA21432@zermelo.csc.ncsu.edu> Subject: Stack overflow problem with CLISP To: clisp-list@clisp.cons.org Date: Tue, 25 Aug 1998 10:09:57 -0400 (EDT) X-Mailer: ELM [version 2.4 PL24/POP] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Hi! I'm trying to run a DFS routine using CLISP, however, I'm getting the following error when it tries to print the solution of the DFS: The solution is: *** - Lisp stack overflow. RESET If I compile the code then it will work, however, is there another way to solve this problem? (i.e. can I increase the stack size somehow). Thanks, Brad P.S. This problem occurs on the Linux and the Solaris versions of CLISP -------------------------------------------------------------------------- Bradford W. Mott (bwmott@acm.org) Computer Science Department http://www4.ncsu.edu/~bwmott/www North Carolina State University -------------------------------------------------------------------------- From bernardp@cli.di.unipi.it Tue Aug 25 07:34:31 1998 Received: from mailserver.cli.di.unipi.it (alice.cli.di.unipi.it [131.114.11.37]) by seagull.cdrom.com (8.8.8/8.6.6) with ESMTP id HAA27372 for ; Tue, 25 Aug 1998 07:34:28 -0700 (PDT) Organization: Centro di Calcolo - Dipartimento di Informatica di Pisa - Italy Received: from fire-cdc.cli.di.unipi.it (IDENT:qmailr@fire-ext.cli.di.unipi.it [131.114.11.52]) by mailserver.cli.di.unipi.it (8.9.1a/8.9.1) with SMTP id QAA06866 for ; Tue, 25 Aug 1998 16:45:03 +0200 (MET DST) Received: (qmail 24202 invoked by uid 7794); 25 Aug 1998 14:41:04 -0000 Received: from carlotta.cli.di.unipi.it(131.114.11.15), claiming to be "carlotta" via SMTP by 2000, id smtpda23931; Tue Aug 25 16:39:52 1998 Received: from localhost (bernardp@localhost) by carlotta (SMI-8.6/8.6.12) with SMTP id QAA05194 for ; Tue, 25 Aug 1998 16:43:54 +0200 X-Authentication-Warning: carlotta.cli.di.unipi.it: bernardp owned process doing -bs Date: Tue, 25 Aug 1998 16:43:50 +0200 (MET DST) From: Pierpaolo Bernardi To: Multiple recipients of list Subject: Re: Stack overflow problem with CLISP In-Reply-To: <199808251409.KAA21432@zermelo.csc.ncsu.edu> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On Tue, 25 Aug 1998 bwmott@eos.ncsu.edu wrote: > > Hi! > > I'm trying to run a DFS routine using CLISP, however, I'm getting the > following error when it tries to print the solution of the DFS: > > The solution is: > > *** - Lisp stack overflow. RESET > > If I compile the code then it will work, however, is there another way > to solve this problem? (i.e. can I increase the stack size somehow). No, you cannot increase the stack size (without modifying the source of Clisp and recompiling from scratch, that is). Why you don't want to compile the code? Pierpaolo. From bwmott@eos.ncsu.edu Wed Aug 26 05:51:33 1998 Received: from zermelo.csc.ncsu.edu (zermelo.csc.ncsu.edu [152.1.213.201]) by seagull.cdrom.com (8.8.8/8.6.6) with ESMTP id FAA03946 for ; Wed, 26 Aug 1998 05:51:32 -0700 (PDT) From: bwmott@eos.ncsu.edu Received: (from bwmott@localhost) by zermelo.csc.ncsu.edu (8.8.4/UC02Jan97) id IAA26046 for clisp-list@clisp.cons.org; Wed, 26 Aug 1998 08:58:22 -0400 (EDT) Message-Id: <199808261258.IAA26046@zermelo.csc.ncsu.edu> Subject: Re: Stack overflow problem with CLISP To: clisp-list@clisp.cons.org Date: Wed, 26 Aug 1998 08:58:22 -0400 (EDT) In-Reply-To: from "Pierpaolo Bernardi" at Aug 25, 98 07:35:03 am X-Mailer: ELM [version 2.4 PL24/POP] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit > > I'm trying to run a DFS routine using CLISP, however, I'm getting the > > following error when it tries to print the solution of the DFS: > > > > The solution is: > > > > *** - Lisp stack overflow. RESET > > > > If I compile the code then it will work, however, is there another way > > to solve this problem? (i.e. can I increase the stack size somehow). > > No, you cannot increase the stack size (without modifying the source of > Clisp and recompiling from scratch, that is). Recompiling isn't a problem. What needs modifying? > Why you don't want to compile the code? I'm setting CLISP up for a class and I'd like the stack to be a little larger so students don't run into the stack overflow problem if they forget to compile their code :-) Thanks, Brad -------------------------------------------------------------------------- Bradford W. Mott (bwmott@acm.org) Computer Science Department http://www4.ncsu.edu/~bwmott/www North Carolina State University -------------------------------------------------------------------------- From hoehle@zeus.gmd.de Thu Aug 27 15:12:31 1998 Received: from mail.gmd.de (mail.gmd.de [129.26.8.90]) by seagull.cdrom.com (8.8.8/8.6.6) with ESMTP id PAA15016 for ; Thu, 27 Aug 1998 15:12:18 -0700 (PDT) Received: from zeus.gmd.de (zeus.gmd.de [129.26.145.1]) by mail.gmd.de (8.8.8/8.8.8) with SMTP id AAA04556; Fri, 28 Aug 1998 00:18:56 +0200 (MET DST) Received: by zeus.gmd.de id AA03273 (5.67b8/IDA-1.5); Fri, 28 Aug 1998 00:17:59 +0200 Date: Fri, 28 Aug 1998 00:17:59 +0200 From: Joerg Hoehle Message-Id: <199808272217.AA03273@zeus.gmd.de> To: meekerdb@rain.org Cc: clisp-list@clisp.cons.org Subject: Re: CLisp compiler Brent Meeker wrote: > Are there any settings for the CLisp compiler on the Amiga - speed, > safety, etc? I'm asking because I asked Franz Inc. why their commercial > Allegro Common Lisp 5.0b is only one tenth as fast as CLisp and they > asked me about settings on the compiler. The compiler of CLISP ignores speed, safety and debug settings and type declarations. Types are only checked by the interpreter. We should probably add something about this somewhere in the docs. But I think you're comparing apples and oranges as there's no Allegro Common Lisp 5.0b from Franz on the Amiga (inimaginable :-) (Macintosh 680x0 emulator?) So how did you measure these speed differences? Regards, Jorg Hohle http://zeus.gmd.de/~hoehle/amiga-clisp.html From bernardp@CLI.DI.Unipi.IT Fri Aug 28 06:47:32 1998 Received: from mailserver.cli.di.unipi.it (alice.cli.di.unipi.it [131.114.11.37]) by seagull.cdrom.com (8.8.8/8.6.6) with ESMTP id GAA19572 for ; Fri, 28 Aug 1998 06:47:27 -0700 (PDT) Organization: Centro di Calcolo - Dipartimento di Informatica di Pisa - Italy Received: from fire-cdc.cli.di.unipi.it (IDENT:qmailr@fire-ext.cli.di.unipi.it [131.114.11.52]) by mailserver.cli.di.unipi.it (8.9.1a/8.9.1) with SMTP id PAA11863 for ; Fri, 28 Aug 1998 15:51:18 +0200 (MET DST) From: Pierpaolo Bernardi Received: (qmail 28453 invoked by uid 7794); 28 Aug 1998 13:54:06 -0000 Received: from carlotta.cli.di.unipi.it(131.114.11.15), claiming to be "carlotta" via SMTP by 2000, id smtpda28183; Fri Aug 28 15:53:31 1998 Received: (bernardp@localhost) by carlotta (SMI-8.6/8.6.12) id PAA04168; Fri, 28 Aug 1998 15:57:34 +0200 Date: Fri, 28 Aug 1998 15:57:34 +0200 Message-Id: <199808281357.PAA04168@carlotta> To: clisp-list@clisp.cons.org In-reply-to: <199808261258.IAA26046@zermelo.csc.ncsu.edu> (bwmott@eos.ncsu.edu) Subject: Re: Stack overflow problem with CLISP From: bwmott@eos.ncsu.edu > No, you cannot increase the stack size (without modifying the source of > Clisp and recompiling from scratch, that is). Recompiling isn't a problem. What needs modifying? On a second thought, maybe increasing the lisp stack size is not a good idea. Clisp uses a `lisp stack', and the regular OS stack; with the current sizes, the lisp stack (usually) overflows before the OS stack does. If the size of the lisp stack is increased, then there's the possibility of the OS stack to overflow first. An OS stack overflow causes an uncatchable segfault (i.e. the lisp session is aborted). Increasing the OS stack size is OS-dependent, and usually is forbidden to non-root users. All this may or may not be correct. For an exact analysis, we must wait for Bruno to come back from its vacation. 8-) Hope this helps. Pierpaolo From amoroso@mclink.it Sun Aug 30 05:00:41 1998 Received: from ammi.mclink.it (ammi.mclink.it [195.110.128.1]) by seagull.cdrom.com (8.8.8/8.6.6) with ESMTP id FAA02740 for ; Sun, 30 Aug 1998 05:00:40 -0700 (PDT) Received: from net145-046.mclink.it (net145-046.mclink.it [195.110.145.46]) by ammi.mclink.it (8.8.8/8.6.12) with SMTP id OAA17923 for ; Sun, 30 Aug 1998 14:07:57 +0200 (CEST) From: amoroso@mclink.it (Paolo Amoroso) To: clisp-list@clisp.cons.org Subject: Problem installing CLISP 1997-12-06 binaries under Linux (kernel 2.0.0) Date: Sun, 30 Aug 1998 12:08:02 GMT Organization: Paolo Amoroso - Milan, ITALY Message-ID: <35e85076.340810@mail.mclink.it> X-Mailer: Forte Agent 1.5/32.451 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I tried to install the binary distribution of CLISP 1997-12-06 available at: ftp://ftp2.cons.org/pub/lisp/clisp/binaries/i386-linux-elf/clisp.tar.gz on my Slackware Linux 3.1.0 system (kernel 2.0.0). Since I am interested in the full version, I edited the fromdir macro in the Makefile as follows: fromdir = full When I run make as suggested in the README file as the first step of the installation procedure, it stopped with the following error: cc -O full/modules.o full/clx.o -lXpm -lXext full/linux.o -lm full/regexp.o full/regexi.o full/regex.o full/wildcard.o full/fnmatch.o full/lisp.a full/libintl.a full/libreadline.a full/libavcall.a full/libvacall.a full/libtrampoline.a -lncurses -o full/lisp.run /usr/i486-linux/bin/ld: cannot open -lXext: No such file or directory make: *** [full/lisp.run] Error 1 On my system libXext is apparently called liblbxXext.a and lives in /usr/X11/lib. So I changed the references in the lines for targets full/lisp.run and full/wlisp.run from -lXext to -llbxXext. After restarting make I got the following errors: cc -O full/modules.o full/clx.o -lXpm -llbxXext full/linux.o -lm full/regexp.o full/regexi.o full/regex.o full/wildcard.o full/fnmatch.o full/lisp.a full/libin tl.a full/libreadline.a full/libavcall.a full/libvacall.a full/libtrampoline.a - lncurses -o full/lisp.run full/clx.o: In function `get_font_info_and_display': full/clx.o(.text+0xd95): undefined reference to `XQueryFont' full/clx.o: In function `get_font': full/clx.o(.text+0xf37): undefined reference to `XLoadFont' ... [More messages about undefined references from clx.o] ... /usr/i486-linux/lib/libXpm.a(CrPFrI.o): In function `xpmCreatePixmapFromImage': CrPFrI.o(.text+0x29): undefined reference to `XCreatePixmap' CrPFrI.o(.text+0x3c): undefined reference to `XCreateGC' CrPFrI.o(.text+0x61): undefined reference to `XPutImage' CrPFrI.o(.text+0x6e): undefined reference to `XFreeGC' /usr/i486-linux/lib/libXpm.a(create.o): In function `SetCloseColor': create.o(.text+0x2bc): undefined reference to `XAllocColor' create.o(.text+0x2cf): undefined reference to `XUngrabServer' create.o(.text+0x37b): undefined reference to `XUngrabServer' create.o(.text+0x3a9): undefined reference to `XAllocColor' create.o(.text+0x3c3): undefined reference to `XGrabServer' create.o(.text+0x3db): undefined reference to `XQueryColors' ... [More messages about undefined references from create.o] ... /usr/i486-linux/lib/libXpm.a(create.o): In function `CreateXImage': create.o(.text+0x1147): undefined reference to `XCreateImage' make: *** [full/lisp.run] Error 1 It seems that many of the referenced symbols refer to Xlib (typically libX11.a) functions. I tried to include it in the Makefile, but I found no trace on my system of a static version of Xlib: there is only the shared version of the library. I also checked the Slackware CD. Any comments or suggestions? Is the installation from a source distribution likely to solve the problem? Paolo -- Paolo Amoroso From haible@ilog.fr Mon Aug 31 12:54:43 1998 Received: from sceaux.ilog.fr (sceaux.ilog.fr [193.55.64.10]) by seagull.cdrom.com (8.8.8/8.6.6) with ESMTP id MAA12376 for ; Mon, 31 Aug 1998 12:54:40 -0700 (PDT) Received: from ilog.ilog.fr ([172.17.4.22]) by sceaux.ilog.fr (8.8.8/8.8.7) with ESMTP id WAA21260 for ; Mon, 31 Aug 1998 22:02:12 +0200 (MET DST) Received: from jaures.ilog.fr (jaures.ilog.fr [172.16.1.81]) by ilog.ilog.fr (8.8.8/8.7.3) with ESMTP id WAA04826; Mon, 31 Aug 1998 22:02:11 +0200 (MET DST) From: Bruno Haible Received: (from haible@localhost) by jaures.ilog.fr (8.9.0/8.9.0) id WAA26945; Mon, 31 Aug 1998 22:02:11 +0200 (MET DST) Date: Mon, 31 Aug 1998 22:02:11 +0200 (MET DST) Message-Id: <199808312002.WAA26945@jaures.ilog.fr> To: clisp-list@clisp.cons.org Subject: Re: $ sign in path not recognized In-Reply-To: <199808190644.XAA15936@seagull.cdrom.com> References: <199808190644.XAA15936@seagull.cdrom.com> Leo Sarasua writes: > I am using the DOS version of CLISP. When I try to load a file which is > in a subdirectory whose name contains the $ sign, it returns a syntax error: > (load "c:\\dir$\\myfile") --> SYNTAX ERROR > > Is this normal? Yes, clisp assumes that DOS filenames contain only alphanumeric characters, dashes and underscores. That may be a little over-restrictive, but it helps discovering typos. > I see that for other systems the $ sign is used to indicate th > root directory, but for DOS it's a perfectly valid character in a path. On other systems $ introduces an environment variable, just like % on DOS. What would you expect clisp to do if you try to create a file whose name is "%PATH%" ? Expand the environment variable or not? Bruno From lsarasua@epo.e-mail.com Mon Aug 31 23:47:22 1998 Received: from E-MAIL.COM (e-mail.com [204.146.168.195]) by seagull.cdrom.com (8.8.8/8.6.6) with SMTP id XAA17875 for ; Mon, 31 Aug 1998 23:47:21 -0700 (PDT) From: lsarasua@epo.e-mail.com Message-Id: <199809010647.XAA17875@seagull.cdrom.com> Received: by E-MAIL.COM (IBM VM SMTP V2R4a) via spool with SMTP id 7733 ; Tue, 01 Sep 1998 02:54:48 EDT Date: Tue, 01 Sep 1998 02:54:48 EDT To: clisp-list@clisp.cons.org In-Reply-To: note of 31/08/98 21:04 X-Sender-Info: Leo Sarasua Tel. (31) 70 - 340 31 77 The Hague, NL lsarasua@epo.e-mail.com MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Subject: Re: $ sign in path not recognized Hello Bruno, thanks for replying to my message, >> I am using the DOS version of CLISP. When I try to load a file which is >> in a subdirectory whose name contains the $ sign, it returns a syntax error >> (load "c:\\dir$\\myfile") --> SYNTAX ERROR >> >> Is this normal? >Yes, clisp assumes that DOS filenames contain only alphanumeric characters, >dashes and underscores. That may be a little over-restrictive, but it helps >discovering typos. >> I see that for other systems the $ sign is used to indicate th >> root directory, but for DOS it's a perfectly valid character in a path. >On other systems $ introduces an environment variable, just like % on DOS. >What would you expect clisp to do if you try to create a file whose name >is "%PATH%" ? Expand the environment variable or not? No, I think that's beyond the scope of CLISP to get into OS functions. However, I think it would be a good idea to allow in a pathname all characters which a given OS allows as normal names. So, for the DOS version, it should allow the $ sign but not the %. For Linux and others, the $ sign should not be accepted. I think it's not that difficult to implement it and it's worth doing it. The problem is that, with the situation as it is now, some directories remain u nreachable for CLISP, unless you modify your directory names, which is not alwa ys possible. Merci et a bien tot, Leo