Subject: Re: merge-pathnames
From: Erik Naggum <>
Date: 1998/04/21
Newsgroups: comp.lang.lisp
Message-ID: <>

* Bruno Haible
| CLISP implements #P since version 1997-05-03.

  that's good.

| If you have problems building clisp on some particular platform, and want
| these problems resolved, a constructive approach would be to submit a
| report to <>.  This way the problems have a real
| chance to be fixed.

  support for sun-sparc-sunos4.1.3 has obviously been scuttled to support a
  myriad weird things in Linux.  attempts to get support have failed.  my
  last attempt (three months ago) was to build on both my SPARCstation and
  an Intel running Solaris 2.5.  both failed with an _astonishing_ amount
  of crud, most of it evidence of sloppy coding in that weird "D" language
  of yours, which, to my untrained eyes, only appear to be a nasty way of
  saying you don't like /* and */ in C.  I don't like gratuitous changes
  like that, so I am obviously just as free to ignore CLISP and warn people
  against it as you are against the Common Lisp and C that other people
  would like to use.  I'm sorry, but I don't find _your_ unconstructive
  approach to wasting my time worth it, especially when I can spend my time
  so much more constructively elsewhere.  maybe something to consider?

  (BTW, something of the same nature has happened to GCL.  I got my hands
  on KCL back in 1987, but the C code was obviously written by someone who
  had never seen anybody else's C code, just like yours, and although it
  has now become GCL (via AKCL), it is still _incredibly_ obtuse and just
  as buggy.  I guess the morale is: if you want people to be benevolent to
  you, at least don't start off by doing something boneheadedly different
  just for the heck of it.)

  Abort, Retry, or Upgrade?