Marcus Harnisch writes:
Traditions never die. Try this for kicks:
Old code that works need not change. That's quite a different issue.
It's this business of using old fragile hacks in new code that makes
me puke when I look at GNU code.
Shouldn't we have all converted to (featurep 'xemacs) or
No. Why change what works? If somebody *wants* to do it, a patch
will probably be accepted.
> XEmacs has for several years had a (mostly) systematic way of
> with version numbers.
Sure. But Emacs doesn't, apart from these two:
So we have to wait for another decade because Emacs sucks even more
than XEmacs does? The ONLY way to get Emacs developers to adopt
modern practices is to get the code in front of their faces so that
they realize that it's better (although that's not necessarily going
to persuade them).
Before *rewriting* inversion.el for XEmacs, I thought that a less
invasive modification to support current XEmacs would be the better
Look, you're just going to end up with
etc etc (or the moral equivalents) anyway, right? How about wrapping
the current code with
(setq inversion-emacs-major-mode (when (boundp emacs-major-mode)
inversion-emacs-minor-mode (when (boundp emacs-minor-mode)
;; etc etc
;; if you want, we can even obsolete-alias the xemacs-* version
;; variables for future cross-emacs standardization but your code
;; will still need to deal with the old wacko xemacs-* versions.
(unless (and inversion-emacs-major-mode
;; etc etc
and then you can modify the old crufty way to snaggle out only stuff
that isn't available in native variables already, at some future date.
XEmacs-Beta mailing list