Re: Some XEmacs issues and proposals
On 4290 September 1993, Ralf Angeli wrote:
> * Holger Schauer (2005-05-31) writes:
>> I wear glasses, but I'm not blind. Referring on a Debian list to a manual
>> installation of a package that Debian distributes as a Debian package is
>> a red herring.
> I don't consider it uncommon that people have software installed
> manually which is also available via the package system. Reasons
> could be special build options or the need for other versions of the
> software.
I agree, but if a manual installation of AUCTeX would be the only
example you have to offer, I still won't take that as a valid
example. E.g., I use beta/cvs versions of (X)Emacs myself and the
Debian packages were more than once a cause of conflicts just like in
your AUCTeX case. But that doesn't imply there's something wrong with
the Debian packages, it's the duplicate installation that's causing
the trouble. Exactly like in your AUCTeX example and I'm quite sorry
to say that I think if your begging for trouble (by a duplicate
installation of files) you're rightfully on your own in getting out of
it afterwards, IMHO.
Of course, that is not to say that it would be better if Elisp files
for different versions of Emacs would play along nicely.
Unfortunately, the world is not always as we would like to have
it. Breaking backwards compatibility to avoid a problem that has been
there and solved for years is just not my favourite way of dealing with
wishful thinking.
> Anyway, we are not talking about problems in XEmacs but about problems
> with its configuration imposed by the Debian Emacs policy. And it's
> not like only the AUCTeX maintainers have noticed problems with it.
True enough.
> Also XEmacs upstream is not very fond of the policy. At least Stephen
> expressed some of his views (also what he thinks about putting
> /usr/share/emacs/site-lisp into XEmacs' load-path) in
> <URL:http://article.gmane.org/gmane.emacs.xemacs.beta/19471>.
Thanks for the reference.
Holger
--
--- http://www.coling.uni-freiburg.de/~schauer/ ---
"You see, dear one-bit person, just because you have this magical
inference engine that is churning out all sorts of things that nobody
argues, does not mean that you arrive at conclusions that apply in
this universe." -- Erik Naggum <erik@naggum.net> in comp.lang.lisp
Reply to: