John Stoffel wrote: > So I'm running debian 12 and the current version of emacs is 28, so I > do think we should try to support back one more version of Debian if > it's not going to be a big problem.
I wouldn't MIND back support. What I worry about is the amount of extra effort needed to do the testing on different versions. If we had automated tests doing it for us it might not be an issue, but we don't. > … doesn't Fedora update much more frequently that > Debian? That is certainly true. Actually, isn't it true about almost all other distributions? :-) Maybe just my prejudices showing here. :-) > I just don't want us to leave too many people in the dust, Me neither, but I tidn't think I was suggesting that. My thinking is that you would get the same effect you would have if VM was one of all the packages included with emacs itself. People using emacs 28 don't have access to the new features of the C++ mode with support cof C++20 coming with emacs 29. That doesn't mean they are "left in the dust" to me. A new feature of VM would be the same. Mark Diekhans wrote: > Features get add, but they are very good about not breaking > compatibility. Deprecation period is long. That is mostly true. But as a counterexample I did get bitten by the name change from native-comp-deferred-compilation-deny-list to native-comp-jit-compilation-deny-list recently. Julian Bradfield wrote: > Now you > could argue that anybody using VM is probably a developer No, I'm not trying to argue that. I don't have knowledge about the user base. > You could also argue that people like that aren't going > to want to upgrade VM, and that may be true. That is closer to what I'm saying. But I would rather frame it as people on old releases will not EXPECT to get the newest version of any package, VM included. What they WANT could be something completely different. See my reasoning around other emacs packages above.
pgp9xfsmfM71n.pgp
Description: PGP signature
