On Fri, 26 Sep 2025 at 20:03:35 +0100, Simon McVittie wrote:
I think this sort-of-transition is at a point where we can go ahead
whenever the release team are ready for it.
Does the release team have an idea of when we will be able to do this
transition?
Unfortunately, some of the changes that were made in GLib-related
packages in preparation for this transition have caused regressions:
GioUnix objects exposed by pygobject do not accept the appropriate
arguments for their methods, causing a functional regression in at least
gnome-tweaks
(<https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1117444#17>). This
has caused several contributors to propose well-intentioned but wrong
patches for gnome-tweaks that would make it regress (again) after the
GLib 2.86 transition goes through, and I'm concerned that similar
patches might already have been applied in other packages without us
noticing.
One way to resolve this would be to do the GLib 2.86 transition, breaking
rebuilds of glib-d (which is depended on by one leaf package,
appstream-generator) but fixing the runtime regression in pygobject.
Another way to resolve this would be to revise the pygobject patches so
that their effect is only applied when GLib is >= 2.86, or to revert
them for now and then reapply them (with versioned Breaks) when we do
this transition. But it seems unnecessary to do that (waste of time +
risk of regressions) if this transition is imminent anyway...