Greg Stein wrote: > On Fri, Aug 30, 2002 at 11:00:49AM -0700, Brian Pane wrote: > >>Justin Erenkrantz wrote: >>... >> >>>I honestly don't care where this ends up. It just needs to get >>>in to our tree somewhere. The aaa code is broken. It needs to >>>be fixed (and I believe the patches we already have start the >>>process). 2.0, 2.1, 3.0, 1.4 - whatever. I'm perfectly happy >>>to break backwards-compatibility. -- justin >> >>I wouldn't object to breaking aaa backward compatibility in 2.0, >>if it eliminates (for now) the need to start 2.1. > > > Hey... I don't believe it is going to break backwards compat [in a big way]. > I'm not sure what the fuss is all about. Since it won't break compat, then > it doesn't need a 2.1 bump. > > I think in the end, there will be these minor compat problems: > > * some LoadModule lines will need to go away; some added > > [ and the failure mode here makes this painless; it will be very obvious > that you need to toss a LoadModule when Apache says it can't find it. > and that you need to load another when it doesn't understand a directive > that you had in the file. ] > > * an undocumented feature of AuthUserFile and AuthGroupFile will go away. it > is unknown how the hell they worked to begin with. > > > I believe that is it. Justin: what else do you think will change from a user > standpoint? > > And note that the API isn't changing here, either. Justin is adding APIs for > people, but no changes.
exactly, this talk of 2.1/branching etc is very premature. if you think your going to destablize the tree, then do your changes on a copy of the file.. and when your done just overwrite the old one. I'm thinking a in-progress directory should be created for these kind of things, simliar to the experimental mpm tree. what we need most is a stable tree for a couple of months not spliting out to a 2.1 tree > > Cheers, > -g >
