The following should be stabilized: x11-libs/gtk+-2.18.6 dev-libs/glib-2.22.4 Reproducible: Always
This is the C++ stuff that should also be stabilized: dev-cpp/gtkmm-2.18.2 dev-cpp/pangomm-2.26.0 dev-cpp/glibmm-2.22.1
x11-misc/shared-mime-info-0.70 should be stabilized at the same time with glib.
Also adding dev-libs/atk-1.28.0 as talked with leio and nirbheek on IRC
Arches, please test and mark stable the above list, if you find any problem please report
amd64 stable
x86 stable
OK to go with 2.18.7 instead since it fixes some bugs and the rest of gnome team agrees (sorry x86 team, I will take care of amd64)
Wow, this broke my whole gentoo (x86). First, I had to emerge dbus-glib, otherwise things like NetworkManager, gnome, pidgin/finch wouldn't start. I'm still investigating, what has to be repaired until my system becomes usuable again. This update is surely dangerous! Were you aware of that!?
I didn't suffer that problem at all on none of my amd64 and x86 systems
Maybe you are seeing bug 297483 (that should be fixed)
x11-libs/gtk+-2.18.7 causes problems to flash: sometimes, clicking on buttons does not perform any action. Tested on amd64 with firefox. Starting firefox with GDK_NATIVE_WINDOWS=1 firefox seems to solve this issue.
(In reply to comment #12) > x11-libs/gtk+-2.18.7 causes problems to flash: sometimes, clicking on buttons > does not perform any action. Tested on amd64 with firefox. Starting firefox > with > > GDK_NATIVE_WINDOWS=1 firefox > > seems to solve this issue. > Probably flash plugin (or nspluginwrapper... are you using it?) is still not "ported" to use newer gtk+, since, after googling a bit, I have verified that this affects to any distribution using gtk+-2.18
Frankly, we can't keep back GTK+-2.18 and subsequently GNOME-2.28 due to some binary proprietary browser plugin, much like xorg-server can't be held back to do proprietary drivers. That said, if an ~arch version of adobe-flash works better, we can work with its maintainers to get a newer version of that into stable tree too.
So x86 is out of here again. There is no newer adobe-flash available.
I have some problems with gtk-2.18 and as I see I am not alone with this. After a upgrade from 2.16 programs like firefox, eclipse behave strange even when I switch desktops apps do not refresh as smooth as with gtk 2.16. Now I know that the problem is the new gdk windows. I have to say that it is strange to introduce new (clearly not 100% compatible back) feature in this way. I manage ugly hack diff -Naur gtk+-2.18.6.orig/gdk/gdk.c gtk+-2.18.6/gdk/gdk.c --- gtk+-2.18.6.orig/gdk/gdk.c 2010-01-11 21:57:37.000000000 +0100 +++ gtk+-2.18.6/gdk/gdk.c 2010-03-19 22:52:20.000000000 +0100 @@ -211,7 +211,7 @@ _gdk_native_windows = TRUE; /* Ensure that this is not propagated to spawned applications */ - g_unsetenv ("GDK_NATIVE_WINDOWS"); + /*g_unsetenv ("GDK_NATIVE_WINDOWS");*/ } g_type_init (); after that echo "GDK_NATIVE_WINDOWS=1" >>/etc/env.d/50gtk2 Everything is back to normal. My point is that such feature should be introduce in a different way for example a new compatible applications should set something like GDK_NEW_WINDOWS! My 3 cents.
(In reply to comment #16) > I have some problems with gtk-2.18 and as I see I am not alone with this. > After a upgrade from 2.16 programs like firefox, eclipse behave strange > even when I switch desktops apps do not refresh as smooth as with gtk 2.16. > Now I know that the problem is the new gdk windows. I have to say that it > is strange to introduce new (clearly not 100% compatible back) feature > in this way. I manage ugly hack > Next time, please send separate bugs to let us organize this better. For eclipse please take a look on: http://bugs.gentoo.org/show_bug.cgi?id=310181 (I will take care of its stabilization on amd64 as soon as I am able to if nobody beats me ;-)) Regarding to "switch desktops" I don't see that regression :-/ and regarding firefox... maybe you should explain it a bit more on a new bug report Regards
> Next time, please send separate bugs to let us organize this better. For > eclipse please take a look on: > http://bugs.gentoo.org/show_bug.cgi?id=310181 (I will take care of its > stabilization on amd64 as soon as I am able to if nobody beats me ;-)) You right sorry about that but I thought that this issue could be addressed before stabilization. > Regarding to "switch desktops" I don't see that regression :-/ and regarding > firefox... maybe you should explain it a bit more on a new bug report Well it not really bug per se but annoying issue i.e. firefox scrolling is not as smooth, when a open new tab I see for a second some artifacts from the previous tab and so on. The same is when I switch a workspace. I hope that those issues disappear with gnome 2.28.
(In reply to comment #18) > > Next time, please send separate bugs to let us organize this better. For > > eclipse please take a look on: > > http://bugs.gentoo.org/show_bug.cgi?id=310181 (I will take care of its > > stabilization on amd64 as soon as I am able to if nobody beats me ;-)) > > You right sorry about that but I thought that this issue could be > addressed before stabilization. > Yes, I understand your point of view and, of course, we try to address that issues before stabilization, but, in this case, it wasn't possible since we (gnome team) were not aware (or forgot) about that problem. For improving that situation, we now use a tracker for controlling gtk+/glib regressions better if possible, see bug 309237
adding s390@ to CC, at least shared-mime-info is stable there.
alpha/arm/ia64/m68k/s390/sh/sparc stable
ppc stable
Stable for HPPA.
ppc64 done; closing as last arch