Re: Death of gnome-common



Hi!

Dave and I are planning to kill off gnome-common. The macros from it
have been migrated to autoconf-archive[1], and gnome-autogen.sh is
mostly irrelevant boilerplate which should be deleted or open-coded in
people’s autogen.sh files in the first place.

Is it OK if we do this, and eliminate the module from the modulesets? We
were thinking about making a final release which contains
gnome-autogen.sh unmodified, plus all the m4 macros defined as
deprecated. After that, we would make a final commit which replaces
gnome-autogen.sh with an unconditional error message directing people
towards a wiki page for migration[2]. If they want to continue compiling
against an older version of gnome-common, that’s fine, but master would
be dead.

How does that sound?

I'm all for killing it but the timeframe is important; at the moment
most modules are still using gnome-autogen.sh, what's the plan to
update them? That's typically GnomeGoals territory, we can certainly
add a new one, or update the ModernAutotools one[1].

Once we reach an arbitrary threshold of conversion, it would then be
ok to replace gnome-autogen.sh by an error message.


        Fred

[1] https://wiki.gnome.org/Initiatives/GnomeGoals/ModernAutotools


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]