gcl-devel
[Top][All Lists]
Advanced

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

Re: [Gcl-devel] 2.6.2 errata


From: Camm Maguire
Subject: Re: [Gcl-devel] 2.6.2 errata
Date: 29 Jun 2004 16:49:04 -0400
User-agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2

Greetings!

Michael Koehne <address@hidden> writes:

> Moin Camm Maguire,
> 
> > Right now I'm inclined to just place this in debian/rules as a diff
> > against the upstream 2.6.2.  Does anyone use arm with any other OS?
>   
>   its something, that should go on the website - but its not a bug
>   that breaks compiling itself - only building by the debian robots.
>   Think ARM users, who compile their own GCL, are so seldom, that
>   they'll find their way to this mailinglist, to receive the clue
>   to call autoconf on their own.
> 

OK, website should be updated within 6 hrs.

> > Going this route, we'd post an errata note to the website.  Problem is
> > I'm not sure anyone looks at it.
> > Alternatively, we could post another source tarball, but this seems
> > rather unpleasant.
> 
>   lets collect those in debian patches first, to make the build robot
>   happy - and perhaps release a 2.6.3 in a few month including backports
>   of the most important bugfixes from 2.7.0, Debian and other places.
> 

OK

> > Thoughts?
> 
>   for 2.7.0, GCL has the luck, that makefile is not generated by makefile.in.
> 
>   delete ./configure from CVS and improve `make configure` :
> 
> cvsclean : clean
>       find . ! -type d | xargs chmod u=rw,og=r
>       find . -type d | xargs chmod u=rwx,og=rx
>       find . -type d | sort > MANINEW
>       diff MANICVS MANINEW
>       rm MANINEW 
> 
>   The CVS containes a lot of executeable files, because of M$Windows upload.
>   The `diff` will generate an error every time CVS MANIFEST has changed.
>   This is a savetybelt, to give some time to add an entry to Changelog and
>   mv MANINEW MANICVS, if you are really sure. If first requires some
>   cleanup, to delete those files that dont belong into CVS, of course.
> 
> clean: gclclean
>       -(cd ./gmp3 && $(MAKE) distclean)
>       rm -rf ./gmp3/.deps ./gmp3/libgmp.a
>       -(cd binutils/bfd && $(MAKE) distclean)
>       -(cd binutils/libiberty && $(MAKE) distclean)
> 
>   the current clean is using $(GMPDIR) resulting a `make distclean`
>   in my $HOME !
> 
> configure: configure.in cvsclean
>         autoconf
>       chmod a+rx configure
>       ( cd ./binutils/; autoconf )
>       ( cd ./gmp3/; autoconf )
>       find . -type d | sort > MANIFEST
>       diff MANIFEST MANINEW
>       rm MANINEW 
> 
>   configure should call autoconf without arguments and descend into
>   subdirectories. The MANIFEST is the normal MANIFEST containing
>   some files in addition.
> 
> tar: configure
>       rm -f gcl-`cat majvers`.`cat minvers`
>       find MANIFEST | cpio -pdumv gcl-`cat majvers`.`cat minvers`-beta-`date 
> +%y%j`
>       (cd ..; tar cvfz gcl-`cat majvers`.`cat minvers`-beta-`date +%y%j`.tgz 
> gcl-`cat majvers`.`cat minvers`-beta-`date +%y%j`)
>       rm -rf gcl-`cat majvers`.`cat minvers`-beta-`date +%y%j` 
> 
>   the tar is dependend on configure, as normal users dont have autoconf,
>   the side effect of maintaining a MANIFEST, is that one could remove
>   xbin/distribute both from CVS and from MANIFEST.


These are all good ideas, but I'm wondering if they should be achieved
by writing a proper Makefile.am and bootstrapping like maxima.  Lets
bring this up in the roadmap discussion.

Take care,

> 
> Bye Michael
> -- 
>   mailto:address@hidden             UNA:+.? 'CED+2+:::Linux:2.4.22'UNZ+1'
>   http://www.xml-edifact.org/           CETERUM CENSEO WINDOWS ESSE DELENDAM
> 
> 
> _______________________________________________
> Gcl-devel mailing list
> address@hidden
> http://lists.gnu.org/mailman/listinfo/gcl-devel
> 
> 

-- 
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]