Newsgroups: comp.lang.lisp
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!gatech!news.sprintlink.net!pipex!uknet!newsfeed.ed.ac.uk!edcogsci!colin
From: colin@cogsci.ed.ac.uk (Colin Matheson)
Subject: CMUCL Vs Allegro (was Re: cmucl Bug)
Message-ID: <D8s1LJ.4C6@cogsci.ed.ac.uk>
Organization: Centre for Cognitive Science, Edinburgh, UK
References: <3p25q8$d92@franklin.cc.utas.edu.au> <3p3s4l$3fj@tools.near.net>
Date: Thu, 18 May 1995 14:00:51 GMT
Lines: 21

I meant to mention this a while back when I came across it.  Allegro
(at least the Windows version) and CMUCL differ in the syntax of the
format command.  The code I was writing pads out numbers to be 3
characters - "001", "020", and so on.  Allegro wanted:

(format nil "~3'0D" NUMBER)

That makes CMUCL fall over here though - it wants:

(format nil "~3,'0D" NUMBER)

It looks to me, when I read the relevant section, like Steele is
interpretable in two ways.  Or is this a known bug in one or other
implementation?

Colin
-- 
Colin Matheson                    | Human Communication Research Centre
Phone: +44 131 650 4632           | University of Edinburgh
Fax:   +44 131 650 4587           | 2 Buccleuch Place
Email: Colin.Matheson@ed.ac.uk    | Edinburgh EH8 9LW Scotland
