Re: [gdm-list] Status of GDM 2.22?

[Brian Cameron <Brian Cameron Sun COM>]

Jon:

> I completely understand your concern.  I'm sorry that I haven't had a
> chance to update the list or do much of anything else for well over a
> month.  Holidays didn't make it any better.  However, in that time,
> among other things, I've changed jobs.  This means that, at least for
> now, I'll be working full time on finishing GDM.  I also know that Ray
> will be able to devote quite a bit of time too.

That is great news to hear.  Congrats on the new job!

> Obviously, we are late in the GNOME 2.22 release cycle and some freeze
> dates have already passed.  However, I am confident that we can finish
> up at least the major items on our to-do list in a relatively short
> time.
>
> I am going to send the GNOME release-team a message to let them know
> where we stand and where we need to go.  I'll propose that we wait one
> week from now to see how much we can finish before we make any
> decisions.

That sounds like a good plan.  If we find that we can get the
functionality to a reasonable level quickly, then I would imagine
we could get permission to be a bit late on the deadlines.

That said, my main concern at the moment is the overall lack of ability
to configure the rewritten GDM, and the lack of support for the
previous configuration options (or auto-migration path to a new
configuration system).

In other words, it seems the new rewritten GDM is looking pretty good
for the normal desktop user who probably doesn't need to configure GDM
in the first place, but is badly lacking a lot of features that are
needed by power users, or users with novel configurations.  It would
be a shame to lose our userbase which depend on the ability to
configure GDM.

When we were first talking about making this rewrite the "next GDM
release" as opposed to a separate forked project, I felt I made these
concerns fairly clear.  My opinion is that it is probably okay to
expect users to deal with some pain (such as loss of gdmgreeter
themeability), but losing the ability to configure options that
affect security, how XDMCP operates, or needed daemon controls is
perhaps too much pain to expect users to bear.  At least, that is
my opinion.

In hindsight, it seems an oversight that we did not think to design
the old and new GDM to be more parallel installable.  This would
perhaps allow distros to ship both.  Allowing users to pick which
one they want to use for an interim period might be a good approach
to consider.

At any rate, I think these issues should be raised with the release
team for guidance about how to proceed.  Specifically, what should a
desktop module do when it previously declares its interfaces as Stable
and then breaks that promised stability.  If we do not have the desire
to really address these concerns, then perhaps an answer could be to
simply bump the release to 3.0 and put on our flameproof suits in
preparation for those expected unhappy users.  Or, as I suggested
in my last email, it might make sense for the classic and rewritten
GDM to be separate modules?  I am sure the GNOME release team can
help provide the right guidance here.

Would you mind raising these issues with the release team when you
send them the email.  Or should I, since this is perhaps more my
concern?

> Once I finish up adding reboot/shutdown support, I'll also update the
> wiki and send a status update message to this list.  Which hopefully
> will answer the questions you've raised here.

Great.  The GDM rewrite has come a long way in the past release cycle.
That said, I do not think we shold be upset if we find we need to skip a
release cycle to get everything done.  There is a lot of work to do.
I am looking forward to seeing the status update.

Brian