Re: Release candidate for XEmacs 21.4.4 / my vacation

APA> I am dying to see this feature in 21.4.4, or at least 21.4.5. I was confused. It's not a feature, actually. This worked all along on UNIX, according to followups to a message of mine: http://list-archive.xemacs.org/xemacs-beta/200006/msg00136.html So, I would consider Ben's mouse-track-insert changes a bug fix for native Windows NT. Ben, what would you say? Stephen> Don't call it a "feature", then. Feature freeze was six Stephen> months ago.... ;-) Honestly, it's not fair to do this Stephen> just before handing off to Vin. Not to mention to all Stephen> the other people who want "just one safe little" feature. Stephen> It's not much fun saying "no" all the time, but we really Stephen> have to if "stable" is going to have any meaning. I understand your concerns. If this change turns out to be too risky, I can work with a locally modified 21.4.5. I would like to ask people to actually try it, on Windows and UNIX. APA> Can I do anything to get this feature in? Stephen> Ben sez, just add the Lisp code to init.el. Stephen> Alternatively, you could make a package, which people Stephen> could require in init.el. Then when we get it fixed, we Stephen> have core XEmacs provide the same symbol and the package Stephen> will just wither away. Stephen> With a little care, it could be a generic solution: the Stephen> package would be "lisp-updates" or some such. Each Stephen> separate fix would have its own feature. I suppose we'd Stephen> need to make the feature "list" into a hash-table because Stephen> these would probably accumulate pretty quickly. Stephen> Alternatively, we could obsolete the symbols with every Stephen> major release (so there would be "lisp-updates-21.1" and Stephen> so on). I'm not sure I like this. This would add yet another complication to debugging people's setups who turn to comp.emacs.xemacs, stating: "My XEmacs does not work." In addition to the usual stuff we would have to find whether they run with any or some "lisp-updates". Stephen> What do you think, Steve? This would really make life Stephen> nicer for the users and easier for the core maintainers, Stephen> but it definitely increases the burden on you---these Stephen> packages could end up being quite large. -- Adrian Aichner mailto:adrian＠xemacs.org http://www.xemacs.org/