[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Gcl-devel] Re: gcl compilation report (redhat 7.3, gcc 3.2, gcl 2.4.3)

From: Camm Maguire
Subject: [Gcl-devel] Re: gcl compilation report (redhat 7.3, gcc 3.2, gcl 2.4.3)
Date: 10 Nov 2002 23:50:34 -0500

Greetings, and thanks for your interest in GCL.

Dan Hitt <address@hidden> writes:

> Hi Camm,
> I'm writing to you to give you a report on my experience compiling
> version 2.4.3 of gcl on a redhat 7.3 (GNU/Linux) machine, using
> gcc 3.2.
> If this should properly go elsewhere, please either forward it,
> or drop me a line and i'll send it to the right address (but
> i got your address from http://www.gnu.org/directory/GCL.html).
> (1) In order to compile, i had to edit h/object.h to change
>     the typedef specifying bool.  I actually changed it to
>     a #define of bool to int, but there is probably some better
>     solution.  As it stands, the definition generates
>     a fatal conflict over the type of a couple of variables,
>     because gcc has its own idea of what `bool' is (and it
>     lets you know by warning about a `useless keyword or
>     typename in an empty declaration').  This is true for both
>     gcc-3.2 and gcc-2.96.
>     Also, as a little experiment, you can typedef float to int
>     (as an analogy to typedef'ing bool to int), and it is
>     quite ineffective.  So the `typedef int bool' is trouble,
>     or rather, does not save trouble, at least for my system.

OK, this is fixed in CVS  I believe, where we basically did an #undef
bool in object.h.  If 2.5.0 takes too long, I'll try to backport this
into a 2.4.4.

> (2) configure must be done in the source directory (as opposed to
>     a separate build directory)

I don't understand this.

> (3) The --prefix option to configure indeed makes the binaries and
>     lib go to the right place, but the info and emacs files do
>     not follow.  I hand-edited makedefs and makedefc to force
>     the info to go in a subdirectory of the prefix directory.

Can you provide some specifics as to where configure decided to place
these on your box, and where you would like them to go?  Also what
machine you are using?  The Linux file system standard specifies that
non-machine code like these files must be located separately from the
binaries, so that /usr/share can be NFS mounted.

> (4) The readme looks a little dated.  In particular, the Wikipedia
>     reveals the unhappy news that Professor Schelter died at the end
>     of July, 2001 (http://www.wikipedia.org/wiki/Bill_Schelter).

Alas, this is true!  The current team has been working on GCL for
almost one year, shortly after Dr. Schelter's unfortunate passing. 
the 2.4.x series is designed to do nothing more than stabilize,
i.e. bugfix, Dr. Schelter's last work.  CVS 2.5.0 has quite a bit more
development behind it already.  When we release it, we will of course
update the docs.

> All of this is meant for your data-gathering purposes, not to complain,
> because i really appreciate the fact that there is a free common
> lisp i can mess around with.  Schelter and you other workers have
> done a real service.
> Thanks for your efforts.

And thank you for your note.  Please feel free to send any feedback to

Take care,

> dan
> address@hidden

Camm Maguire                                            address@hidden
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah

reply via email to

[Prev in Thread] Current Thread [Next in Thread]