Now that 2.4.* is a source only build with no feature work or security patches, should we consider removing it from the supported list of Python's. also note that 2.4.6 was last touched in December 2008 so it has missed a number of security fixes in the past 2 years See http://www.python.org/download/releases/2.4.6/ for details
Several of the Linux distros we still support (RHEL variants) still ship with 2.4 as their default or perhaps only version of python. Until a compelling reason to drop support comes along I really don't want to do it just because.
It wasn't a complete "just because" (tho yes, it's close) - I was thinking more of the coming soon battle between Python 2.6 (and 2.7) syntax differences. Should I close as INVALID or WONTFIX?
No, but you could point out compelling reasons why 2.5 is a better baseline than 2.4. :)
Created attachment 466695 [details] [diff] [review] Patch to fix I need python 2.5 over in bug 546857 to be able to automatically generate the permissions.sqlite database.
Comment on attachment 466695 [details] [diff] [review] Patch to fix Yeah, let's do it.
Requiring python 2.6 would be even nicer, see e.g. bug 450645
bug 460052, rather
I'd love to go straight to 2.6 to be able to use relpath.
I'll let Ted make the call. I'm happy to update the patch if everyone agrees it's the right thing to do.
Let's take that to a different bug, this immediately blocks sicking's work and 2.6 is a different argument.
This blocks a blocker so marking approved
http://hg.mozilla.org/mozilla-central/rev/bb200e1f52b4 Checked in. Thanks for the quick review!
(In reply to comment #4) > I need python 2.5 over in bug 546857 to be able to automatically generate the > permissions.sqlite database. Actually, this is totally untrue, since you can generate hostperm.1 instead.
What is hostperm.1? And is that beneficial to permissions.sqlite?