From hoehle@zeus.gmd.de Wed Jul 1 02:28:14 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 CAA26485 for ; Wed, 1 Jul 1998 02:28:09 -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 LAA32584 for ; Wed, 1 Jul 1998 11:37:07 +0200 (MET DST) Received: by zeus.gmd.de id AA14812 (5.67b8/IDA-1.5 for clisp-list@clisp.cons.org); Wed, 1 Jul 1998 11:36:34 +0200 Date: Wed, 1 Jul 1998 11:36:34 +0200 Message-Id: <199807010936.AA14812@zeus.gmd.de> From: Joerg.Hoehle@gmd.de (Joerg Hoehle) To: clisp-list@clisp.cons.org Subject: *error-output* and *debug-io* Content-Type: text Hi, I was just reading Peter Norvig and Kent Pitman's Tutorial on good Lisp programming style from http://www.norwig.com/. It says p.36: -------- Use *error-output* for warnings and error messages that are not accompanied by any user interaction. Use *debug-io* for interactive warnings, error-messages, and other interactions not related to the normal function of a program. In particular, do not first print a message on *error-output* and then do a debugging session on *debug-io*, expecting those to be the same stream. Instead, do each interaction consistently on one stream. -------- I didn't check the latest release but I think CLISP's error routines still print the error message to *error-output* and then proceed to the debugger in *debug-io* (user1.lsp). I think the advocated behaviour is better than current CLISP's. The Amiga makes it trivial to attach different console windows to each of *...-i/o* and doing so I've already been disturbed by the uncohesive (mathematically) output to two distinct streams/windows. Ok, I'll provide a patch sometime. Regards, Jo"rg Ho"hle. Joerg.Hoehle@gmd.de http://zeus.gmd.de/~hoehle/amiga-clisp.html From haible@ilog.fr Wed Jul 15 10:02:18 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 KAA28466 for ; Wed, 15 Jul 1998 10:02:11 -0700 (PDT) Received: from ilog.ilog.fr ([172.17.4.22]) by sceaux.ilog.fr (8.8.8/8.8.7) with ESMTP id TAA00716 for ; Wed, 15 Jul 1998 19:01:56 +0200 (MET DST) Received: from halles.ilog.fr (halles.ilog.fr [172.16.1.96]) by ilog.ilog.fr (8.8.8/8.7.3) with ESMTP id TAA19262; Wed, 15 Jul 1998 19:03:33 +0200 (MET DST) From: Bruno Haible Received: (from haible@localhost) by halles.ilog.fr (8.8.5/8.8.5) id TAA04229; Wed, 15 Jul 1998 19:02:52 +0200 (MET DST) Date: Wed, 15 Jul 1998 19:02:52 +0200 (MET DST) Message-Id: <199807151702.TAA04229@halles.ilog.fr> To: clisp-list@clisp.cons.org Subject: Re: *error-output* and *debug-io* In-Reply-To: <199807010936.AA14812@zeus.gmd.de> References: <199807010936.AA14812@zeus.gmd.de> Joerg Hoehle wrote in July: > I was just reading Peter Norvig and Kent Pitman's Tutorial on good > Lisp programming style from http://www.norwig.com/. > > It says p.36: > -------- > Use *error-output* for warnings and error messages that are not > accompanied by any user interaction. > > Use *debug-io* for interactive warnings, error-messages, and other > interactions not related to the normal function of a program. > > In particular, do not first print a message on *error-output* and then > do a debugging session on *debug-io*, expecting those to be the same > stream. Instead, do each interaction consistently on one stream. > -------- This advice is not coherent with the ANSI CL specification, in the case of interactive warnings (when *break-on-signals* is true): The value of *error-output*, called error output, is a stream to which warnings and non-interactive error messages should be sent. > I didn't check the latest release but I think CLISP's error routines > still print the error message to *error-output* and then proceed to > the debugger in *debug-io* (user1.lsp). Yes, this is how clisp does it (see conditio.lsp as well). > I think the advocated behaviour is better than current CLISP's. The > Amiga makes it trivial to attach different console windows to each of > *...-i/o* and doing so I've already been disturbed by the uncohesive > (mathematically) output to two distinct streams/windows. I think the spec and Norvig's advice are both totally messy. In their model, the semantics of *error-output* depend on both *break-on-signals* and on the nature of *terminal-io*. And they are unsatisfactory, because *error-output* is typically redirected to a log file, and if some error messages don't show up in the log file, the log file is useless. (Redirecting *debug-io* to the same or another log file doesn't make sense.) CLISP's behaviour is to always output a non-handled error to *error-output*. To improve the *debug-io* output, it could be modified to send the error message to *debug-io* as well, if *debug-io* differs from *error-output*. Bruno From sds@goems.com Thu Jul 16 15:26:12 1998 Received: from darius.concentric.net (darius.concentric.net [207.155.184.79]) by seagull.cdrom.com (8.8.8/8.6.6) with ESMTP id PAA05692 for ; Thu, 16 Jul 1998 15:26:12 -0700 (PDT) Received: from newman.concentric.net (newman [207.155.184.71]) by darius.concentric.net (8.8.8/(98/04/23 5.10)) id SAA29172; Thu, 16 Jul 1998 18:27:52 -0400 (EDT) [1-800-745-2747 The Concentric Network] Errors-To: Received: from mute.eaglets.com (ts007d01.phe-pa.concentric.net [209.31.155.61]) by newman.concentric.net (8.8.8) id SAA17848; Thu, 16 Jul 1998 18:27:46 -0400 (EDT) Received: (from sds@localhost) by mute.eaglets.com (8.9.1/8.9.1) id SAA29905; Thu, 16 Jul 1998 18:06:25 -0400 To: clisp-devel@seagull.cons.org, clisp-list@clisp.cons.org Subject: places.lsp: multiple values 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 From: Sam Steingold Date: 16 Jul 1998 18:06:22 -0400 Message-ID: Lines: 17 X-Mailer: Gnus v5.5/Emacs 20.2 I started to convert the setf stuff to multiple-value-compliance. The main file under surgery (or is it vivisection?) is places.lsp. It looks like the functions which will have to be modified will have to be replaced outright (the modifications are too large). Already `diff -cwb` is almost half the size of the file, and getting worse. The problem I am having at the time is with `defsetf': it stubbornly insists on a single store value. It looks like much of the processing there happens somewhere else, where the variables SYSTEM::%LET-LIST et al are handled. What do I do? Where are these monsters treated? (ANALYZE1 doesn't seem to be a lisp function...) -- Sam Steingold, running RedHat5.1 GNU/Linux (http://www.linux.org) Micros**t is not the answer. Micros**t is a question, and the answer is Linux, the choice of the GNU (http://www.gnu.org) generation. MS: Brain off-line, please wait. From gko@alcatel.com.tw Wed Jul 22 20:42:09 1998 Received: from alcatel.com.tw (pobox.alcatel.com.tw [202.39.115.1]) by seagull.cdrom.com (8.8.8/8.6.6) with ESMTP id UAA17749 for ; Wed, 22 Jul 1998 20:42:03 -0700 (PDT) Received: from capricorn.taisel.alcatel.com.tw ([192.72.20.9]) by ns.alcatel.com.tw with SMTP id <26519-1>; Thu, 23 Jul 1998 19:41:11 +0800 Received: from GKO ([202.39.116.241]) by capricorn.taisel.alcatel.com.tw (4.1/SMI-4.1) id AA07270; Thu, 23 Jul 98 11:44:53 CST To: clisp-list@clisp.cons.org Subject: clisp be used as a library for extension language of C programs Organization: Alcatel Telecom Taiwan Reply-To: gko@gko.net Sender: Original-Sender: gko@alcatel.com.tw Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit From: Georges KO Date: Thu, 23 Jul 1998 12:01:16 +0800 Message-Id: Lines: 10 X-Mailer: Gnus v5.3/Emacs 19.34 Hi, Has someone tried to embed clisp to some C programs, like SIOD for Scheme? Thanks. -- Georges KO, Taipei Alcatel Telecom Taiwan gko@alcatel.com.tw Republic of China, Year 87, Thu Jul 23