"Stephen J. Turnbull" <turnbull(a)sk.tsukuba.ac.jp> writes:
David Kastrup writes:
> "Stephen J. Turnbull" <stephen(a)xemacs.org> writes:
>
> > David Kastrup writes:
> >
> > > _If_ you are lucky, the results will be comparable, but not reached by
> > > the same means.
> >
> > Well, sure. But isn't that the kind of argument you've never
> > been willing to accept when I make it? You're the one who's been
> > insisting for years that all that matters is the installation
> > tarball unpacks in the right places.
>
> You really don't get the difference between two packages being
> identical and both having the full anatomy of an XEmacs package?
Of course I do. Since we don't have the new XEmacs package to
actually compare, who's to say that we won't get something identical
to what you've produced, or identical in function?
Probability. There is no point in testing something that has been
independently created unless the goal is to prove that it can
theoretically be made to work.
But that's not the main issue.
Presumably you have a test suite that can be run on installed AUCTeX.
Nope.
If the XEmacs-produced package passes that on all installations, I
don't see the problem. If you don't have such a test suite, then your
talk about your testing is all bluster anyway, because you have no way
to verify that AUCTeX is installed correctly on any sites except your
beta testers, and even there it's all about user habits; you can't
claim anything about coverage -- if someone tries something tomorrow
they didn't do yesterday, it might break.
Ok, so you are fine with the idea of our developers ceasing to test
XEmacs functionality even superficially, since it means nothing more
than an excuse for me to bluster.
We wouldn't want to have that, would we?
Sure, that "might break" will apply with more force to an
XEmacs-
produced package, but on the other hand it will get a lot more beta-
testing *on XEmacs* if we're distributing an up-to-date package, and
presumably will quickly converge to a level of quality commensurate
with the amount of testing done. That's something we can't know until
we try it.
So try it for once. Would be a welcome change to you explaining for six
years why it is an acrimony and frustration that we pitched in.
I am _perfectly_ willing by now to drop all work on supporting XEmacs
and leave the field to you so that you are not hampered in putting your
theories about how programs should port and quality control themselves
to work.
After all, you _are_ the appointed project leader of XEmacs. If it is
your explicit wish to get the AUCTeX developers to stop getting in your
hair by supporting XEmacs in all the wrong and acrimonious and
frustrating ways, there is little point in fighting you.
--
David Kastrup, Kriemhildstr. 15, 44793 Bochum
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta