> > Writers of libraries that are not part of the Emacs distribution > > are often more concerned about compatibility for users of > > different releases and even different flavors of Emacs > > (e.g. XEmacs). > > They might be concerned, yes, but I don't think anyone can dictate > that a package author mustkeep their code backwards compatible -- even > when it _is_ possible.
I certainly didn't mean to dictate any such thing to anyone. > If the package author want to do so, it is ok, Thanks! > but for the specific > case of linum.el, alternative packages already exist which works ok > with older Emacsen, so why look back, rather than look forwards, and > provide some _new_ features for the latest and greatest Emacs? As you say, it's up to the package writer. I'd say that what Markus did in the latest version was what I would suggest, if a writer has the time and the will: make the library useful with various versions, but let it take advantage of recent features for recent versions. It takes a little more work (sometimes a lot more work) to do that, but I, for one, appreciate it. And, as I said before, for some libraries it makes little sense to try to be backward compatible. I think linum.el might become the general replacement for setnu.el and other packages, for old Emacs versions as well as for new ones. You are right that linum.el could take advantage of even more Emacs 22 features. And you (and Leo) are right that that would help test Emacs 22 and would help us discover new things we can do with the new features. That doesn't preclude making the basic functionality available also to users of older Emacs versions - provided that Markus has the time and will to do that. _______________________________________________ gnu-emacs-sources mailing list [email protected] http://lists.gnu.org/mailman/listinfo/gnu-emacs-sources
