2012-10-08 18:49, Jonas Platte skrev:
Am 08.10.2012 09:55, schrieb Kjell
Ahlstedt:
Can you please file a bug in Bugzilla! This bug seems to have
gone unnoticed by the gtkmm developers for a long time. (https://bugs.launchpad.net/ubuntu/+source/gtkmm2.4/+bug/30345
from 2006-02-02) That's probably what to expect when bug reports
have been filed only in launchpad. If you don't have a Bugzilla
account, you can get one very easily. Just go to https://bugzilla.gnome.org,
click "Open a New Account" and follow the instructions.
Making Gtk::Bin::remove() virtual will break ABI, and that's not
allowed now. Probably it will not be allowed before gtk+ 4 and
gtkmm 4 are released. I don't know when that will be. Adding a
virtual function will add an entry to the vfunc table.
You can add a non-virtual Gtk::ScrolledWindow::remove() without
breaking ABI, but I suppose that's not as good.
Kjell
Okay I filed a bug on GNOME bugzilla now. It's available here: https://bugzilla.gnome.org/show_bug.cgi?id=685739
Who decides now how to deal with this?
If you can suggest a solution, you should add it to your bug report,
preferably as an attached git patch. If the best solution you can
find breaks ABI and/or API, it would still be fine to have it in the
bug. It can't be implemented now, but hopefully later.
I suppose you want a fix in gtkmm 3. Then we need a solution that
does not break ABI or API, even if it's not perfect.
If you don't suggest a solution yourself, you just have to wait, and
hope that someone else will.
The question "who decides" has no easy answer. Very often it's
Murray Cumming that decides if a patch shall be pushed to the git
repository. If it's a fairly simple fix, with obviously no bad side
effects, someone else can decide.
Kjell
|