Re: [Langband-devel] ACL 6 Problems (mk:defsystem, et.al)

[Marco A. <ma...@cs...>]

Hello

> Date: Sat, 4 Nov 2000 22:12:14 -0600 (CST)
> Cc: SDS Development <sds...@li...>,
>    Langband development <lan...@li...>,
>    ma...@us...
> From: da...@th...
> Organization:  Are you kidding?
> Content-Type: text/plain; charset=US-ASCII
> Sender: tho...@vi...
> Content-Length: 1136
> 
> > hi,
> > 
> > I'm updating the build-system for SDS and as the license for ACL 5
> > Trial has vanished I had to go with ACL 6.  ACL 6 comes out of the
> > box as case-sensitive which creates a lot of problems for
> > defsystem.lisp.
> 
> Oh, my!  That's evil, and really needs to be fixed.  Have you tried
> (setf (readtable-case *readtable*) :upcase)
> ?

It is *WORSE* than that.

   cl-user(1): (readtable-case *readtable*)
   :upcase

Franz uses a non ANSI machinery to make the CL case sensitive.

See the file 'case.htm'.




> That should make the Lisp reader do what it's bloody well supposed
> to do out of the box and smash all symbols to upper-case.

No it won't.  You need to generate a fresh ANSI image using
non-standard features.

All in all they do have a point.  I am convinced that the uppercasing
behavior is after all bogus.

However, they mede a decision to release something that is not
standard.

> 
> I think I'll drop plans for working with ACL.  When I get the Linux
> box working, I can at least use CLisp, even if I can't get CMUCL
> working on it.  (I've got a few problems; it's an old and quirky
> box.)
> 
>   I had to really mutilate defsystem.lisp to get it
> > working (the #+/#- reader macros in the start are awful) and I
> 
> If this is the case issue, I advise removing the changes and making
> the reader smash to uppercase.  That sounds like an ugly way to
> maintain something nonstandard.

The changes are not that bad.  They just look ugly and they force you
to waste symbols in DEFPACKAGE specs.

The problem is that they may break systems which relied on the
standard *readtable* case being :UPCASE.

E.g. ILISP.

They should at least set their variable regulating package name lookup
case to T.

Cheers.

-- 
Marco Antoniotti =============================================================
NYU Bioinformatics Group			 tel. +1 - 212 - 998 3488
719 Broadway 12th Floor                          fax  +1 - 212 - 995 4122
New York, NY 10003, USA				 http://galt.mrl.nyu.edu/valis
             Like DNA, such a language [Lisp] does not go out of style.
			      Paul Graham, ANSI Common Lisp