On Thu, Oct 23, 2008 at 09:30:57AM +0100, David Woodhouse wrote: > > Yeah, that sucks. We should do better than that. > > Although I've always been a little dubious about our 'Feature Process', > it does seem that it addresses this kind of problem. For the feature to > reach 100% completion, it should obviously involve fixes for the other > display managers. And there is a 'reversion plan' in case we don't > manage to complete the feature in time for the release. Here, simply keeping up using pam_console would have worked. But lead to much less testing of fast user switching. The whole story can be tracked down from this bug, from 2007-02-09: https://bugzilla.redhat.com/show_bug.cgi?id=228110 Later a pam module was added for login. And even later a solution has been found for dm, building on what was done for xinit, that plays well with wdm/xdm/slim, though it is still not applied: https://bugzilla.redhat.com/show_bug.cgi?id=452156 >From my point of view, this comment (certainly biased, it is a comment from me) summary the situation, and it is really a fedora issue, not only an upstream issue: https://bugzilla.redhat.com/show_bug.cgi?id=452156#c15 > From what you say above, it sounds like the ConsoleKit feature should be > declared incomplete, and we should be reverting it unless the feature > owners finish the job. If this is the case, then we would only have had ConsoleKit in F-11... At the time the issue happened, there was no feature process, but I can't see how it would have changed anything. A feature process cannot prevent lack of planning, focus on gnome, lack of integration with existing frameworks and lack of consideration for alternative setups, if this attitude is endorsed by the project on a whole, which is the case for fedora, see for example https://bugzilla.redhat.com/show_bug.cgi?id=228110#c19 and look at all the conversations on this mailing list or fesco decisions. A FESCo vote (at least of today FESCo) would certainly be to let the minor dm be broken (hopefully they will be fixed for the RHEL release), so even if a Feature was not considered finished I doubt the changes would have been reversed. Once again, I don't think that such an attitude is fundamentaly problematic, this allowed to have much more testing for gdm, kdm and fast user switching, which is good. But this is also, in my opinion, a good example of what fedora really is, and my personnal point of view is that it is becoming even more so. As a side note I thought that having more packages, and packages that are not mainstream, with the corresponding packagers being part of the community would lead to a push to another direction, but for good or bad, this didn't happened -- and I think that in the fedora community, those, like me, not in complete agreement with the 'mainstream' fedora (more testing than planning, innovation breaking old frameworks, focus on major desktop components) are, in my opinion, getting less and less consideration. The corresponding packages are simply broken intermitently, always on the struggle to catch up with the constant changes. -- Pat -- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
