Forgot the upstream commit link:
[1]
https://github.com/mate-desktop/mate-settings-daemon/commit/474cc44934eb915f97dd8ba876e2256c40b4ea9a
Hi,
In 1.24 org.mate.SettingsDaemon.DateTimeMechanism.conf has been moved from /etc
to /usr/share [1] (if using the default configure setting). I guess you’ll have
to add debian/maintscript to actually remove it from /etc on the next package
upgrade...
Hi,
Indeed, in 1.24 org.mate.CPUFreqSelector.conf has been moved from /etc to
/usr/share [1] (if using the default configure setting). As a consequence, it’s
no longer in mate-applets but in mate-applets-common instead. I guess you’ll
have to add debian/maintscript to actually remove it from
Hi,
So, snapd sets $PATH and $XDG_DATA_DIRS in /etc/profile.d/apps-bin-path.sh, but
/etc/profile
and /etc/profile.d aren't used in the graphical logins in Debian. [1]
Older versions of snapd also had set $XDG_DATA_DIRS in
/etc/X11/Xsession.d/65snappy, which
was used in the X sessions. But
Hi,
This extension seems to work with Python 3 after one import change:
-from MediaInfoDLL import *
+from MediaInfoDLL3 import *
After that, it also needs a dependency change from python-mediainfodll
to python3-mediainfodll.
Hi,
Upstream migrated to Python 3 in version 19.7.9, so it can be uploaded to fix
this.
Note: version 19.7.15 depends on development version of python-caja (1.23), so
it's
too new for now (until MATE 1.24 is released in the beginning of 2020).
Changelog:
Control: fixed -1 0.57.0-1
Control: found -1 0.48.0-2+deb9u2
Hi,
Upstream should be now compatible with Python 3:
https://github.com/infirit/caja-admin/commit/0af109ac79df76fbcc406d577d79f1c64d1a1fca
There was no new upstream release though, so that commit should be used as a
patch.
Hi,
Martin updated packaging for Python 3 before 1.22.1 was uploaded to Unstable:
https://salsa.debian.org/debian-mate-team/caja-dropbox/commit/8694f61e85248792de2bbacd11c4999860125c17
Hi,
Starting from version 1.22.0, python-caja is compatible with both Python 2 and
3.
So the maintainers can pass PYTHON=python3 environment variable to ./configure
to build it for Python 3.
We might also drop Python 2 support upstream in version 1.24. It will be
probably
released in the
Hi,
The migration to Python 3 and GI had been done in caja-dropbox 1.22.0.
I guess we can close this?
Hi,
Just in case it will ever be updated to 1.2.0... people should know that 1.2.0
ignores any backups made with 1.1.x,
so the first backup made with 1.2.0 will be FULL one. Not a nice surprise,
especially if your backup drive doesn't
have enough space to hold one more full backup.
Source: mate-dock-applet
Version: 0.88-1
Severity: normal
Tags: buster bullseye sid
Hi,
Currently mate-dock-applet depends on a transitional package gir1.2-mate-panel,
which in its turn
depends on a real GI package gir1.2-matepanelapplet-4.0. It's better to make
mate-dock-applet
depend
Source: mate-applets
Version: 1.20.3-2
Severity: normal
Tags: buster sid
Hi,
As I've mentioned in [1], mate-applets 1.20.x doesn't have any Python applets
inside.
So, build dependency on python-gi-dev package and runtime dependency on python
and python-gi packages can be dropped.
In addition,
Hi,
Upstream has a workaround for this crash at [1], you might want to use it as a
patch.
Also this report can be probably merged with [2].
[1]
https://github.com/lanoxx/tilda/commit/daa64f579b4d544cb7da614612f1d35c461b43be
[2] https://bugs.debian.org/904309
Hi,
Crash under Wayland is also confirmed in [1].
Maybe the two reports should be merged, with setting severity to Grave for this
one?
[1] https://bugs.debian.org/925483
Source: mate-menu
Version: 18.04.3-3
Severity: normal
Tags: buster sid
Hi,
Currently mate-menu depends on a transitional package gir1.2-mate-panel, which
in its turn
depends on a real GI package gir1.2-matepanelapplet-4.0. It's better to make
mate-menu
depend directly on
Source: python-caja
Version: 1.20.2-1
Severity: normal
Tags: buster sid
Hi,
Currently python-caja depends on a transitional package gir1.2-caja, which in
its turn
depends on a real GI package gir1.2-caja-2.0. It's better to make python-caja
depend
directly on gir1.2-caja-2.0.
Hi,
Looks like it's a problem in the French translation (malformed string) which
appeared
with our latest update from Transifex before 1.20.3 release. I'll try to fix it.
Hi,
Thanks, I'll try to figure out what's going on there. That legacy code is a
mess...
Hi,
I'm not able to reproduce it in my Debian Testing VM. Don't have it installed
on baremetal,
maybe it's relevant. Anyway, the full backtrace might give some hints... can
you obtain it?
You'll need to do the following:
1. Run the debugger with the latest dump from Caja:
$ coredumpctl debug
Hi,
Ok, seems like the bug is in Caja, not in libgio. It's not clear yet what
causes the crash...
Does it crash on every copy/paste of any folder? Is there something common
between all
the folders? For example, a lot of large files, a lot of small files, some
specific characters
in
Hi,
Ok, seems like the bug is in Caja, not in libgio. It's not clear yet what
causes the crash...
Does it crash on every copy/paste of any folder? Is there something common
between all
the folders? For example, a lot of large files, a lot of small files, some
specific characters
in
Hi,
Since the crash is in libgio, it could be a bug in it (or maybe libgtk) instead
of Caja.
A backtrace is needed to know for sure. You can use coredumpctl tool from
systemd-coredump package to see the backtrace.
You'll also need -dbgsym packages from
Hi,
Since we had dropped [1] the non-working Invest applet from the package before
1.20 release,
there are no Python applets inside anymore. Therefore, Python and GI
dependencies can be
safely removed. Oh, and the mention of the Invest applet in package description
and copyright
file can be
Control: retitle -1 new upstream release 2019.02.14 is available, please
package it
Hi,
Now upstream released new version 2019.02.14.
This one is completely migrated to Python 3.
https://github.com/dropbox/nautilus-dropbox/blob/895e1323a81581894438ee75a13785c1a013/ChangeLog
Source: mate-terminal
Version: 1.20.2-1
Tags: buster sid
Hi,
Please drop 0001_fix_setting_transparency.patch - it makes the menubar of
mate-terminal
transparent with some GTK+ themes. You can test it with Adwaita, for example.
See also [1]
for a screenshot of the problem, showing hardly
Source: nautilus-dropbox
Version: 2015.10.28-1
Severity: wishlist
Tags: buster sid
Hi,
Upstream released new version 2019.01.31. Can you please package it?
This will allow to close #885355 and #876985.
Hi,
> Only remove that patch? Or add some other code?
Yes, only remove it. My tests showed that it's enough to fix the menu (as I
wrote
in the upstream report).
Oh, I see it's already fixed by now, thanks :)
Hi,
So, everyone seems to have this issue exactly with version 1.20.0-2. It's the
version
where 2003_fake_xdg_current_desktop.patch had been added [1]. I've done some
testing
and I can say this patch does the wrong thing. See my comment with screenshot
at [2]
for more info.
Therefore I
Control: tags -1 patch stretch upstream fixed-upstream
Control: affects -1 guake network-manager-gnome
Control: fixed -1 3.24.3-1
Hi,
This is an issue in GtkStatusIcon, not Guake, so I've reassigned it. The same
corrupted/duplicated
icon problem also happens in nm-applet and skypeforlinux
Control: found -1 0.99.4-4
Control: tags -1 patch stretch upstream fixed-upstream
Control: forwarded -1 https://bugs.freedesktop.org/show_bug.cgi?id=100626
Control: affects -1 mate-power-manager gnome-power-manager
Hi,
This had been fixed in upower 0.99.5 upstream [1]. Therefore, the bug
Hi,
This is now fixed in all supported versions of Ubuntu [1].
Maybe it's time to fix it in Debian too?
[1] https://bugs.launchpad.net/intltool/+bug/1117944
MATE 1.20 is released now. This can be closed once the release reaches Debian
repos.
Hi,
Thanks, patch applied upstream.
https://github.com/mate-desktop/mate-user-share/commit/4a48e270808eb51d40a071ba43d1e37433109142
Hi,
Can you also install the debug packages [1] and use "bt full" command in gdb?
The debug packages are needed for libgtk-3, libglib-2.0, mate-panel.
This will show function names, line numbers, variable values and other useful
info
to determine where and why the crash happens.
[1]
Hi,
If I understand correctly, you want me to run `systemd-coredump --backtrace`
(or some similar command) the next time mate-panel crashes?
Apologies if I am quite wrong.
Yes, or use the coredumpctl tool. It can be used to run gdb, and then you can
print the backtrace from gdb's prompt.
Control: notfixed -1 1.18.3-1
Hi,
Should this report be reopened, as it's valid for Jessie?
Hi,
Thanks, patch applied upstream.
https://github.com/mate-desktop/engrampa/commit/6045a6d7ebeaf6d89cf32b16080b1409f063fb50
Hi,
A backtrace for this crash is needed to find out the cause of the crash.
It can be obtained, for example, if systemd-coredump is installed.
https://www.freedesktop.org/software/systemd/man/systemd-coredump.html
https://wiki.archlinux.org/index.php/Core_dump
Hi,
This can be safely dropped as libgksu isn't used by mate-system-monitor since
1.14.0.
It runs gksu binary instead [1]. Also since 1.18.0 it tries pkexec first before
trying gksu [2].
(Support for both is optional, so no hard dependency is required.)
[1]
Control: tags -1 stretch
Control: fixed -1 1.18.1-1
Hi,
I confirm this:
- the problem is fixed in 1.18.1, so it doesn't exist in Buster/Sid
- the mentioned commit fixes the problem in Stretch
Hi,
This bug report is valid only for Jessie where Pluma is GTK+2.
Adding python-gtksourceview2 dependency in Buster (or Stretch)
is completely useless... Pluma won't use it.
I've noticed it because it's also the cause of having unneeded
python-gtk2 package installed in Ubuntu MATE 18.04.
Hi,
The menu item is back, and the dialog works, but you still need to resize it
every time you switch between the categories inside it. Otherwise it will show
empty space instead of colors or emblems. That's why I didn't close the upstream
report yet.
That's what the original reporter here
Hi,
I don't remember fixing this either in git master or in 1.18 branch...
The upstream report is still open. Well, maybe I missed something.
Hi,
I'm pretty sure we didn't backport the commit [1] to 1.18. This change was
meant for 1.20,
and we also made similar changes in Caja and Pluma (in Pluma it's in
filebrowser plugin).
[1]
https://github.com/mate-desktop/mate-panel/commit/e04356360b9277fa6001f57df712752b315d1be3
Hi,
The panel-properties-dialog-gtk3.ui file is from 1.16. It was there when
mate-panel
supported both GTK+2 and GTK+3 builds. In 1.18 it's only GTK+3, and the file is
named
panel-properties-dialog.ui.
Both mate-panel executable and the mentioned .ui file are in the same package
(mate-panel).
Hi,
Can you actually reproduce the issue? For me guake always shows the icon.
I've tried with locally built mate-panel 1.19, then downgraded to 1.18.4-2
from the repos. Also tried both manual launch of guake and autostart with
the session. No luck with reproducing.
Control: fixed -1 1.18.4-1
Control tags -1 - buster - sid
Hi,
It had been fixed upstream in 1.18.3, so marking it as fixed in the first
1.18.x version that appeared in the repos.
We didn't make a patch for 1.16 yet, I hope to find some time to do it.
Not sure if this report should be closed
Hi,
> My suspicion is that guake does not use X11 embed anymore (which is
> the API provided by the notification area applet), but the Indicators
> API.
>
> Try installing and adding mate-indicator-applet and check if you find
> guake's icon in there.
It's not going to work in Debian...
Hi,
I think it's better to do something (i.e. apply the patch) instead
of leaving intltool broken. I know that it had been patched in Fedora,
and there weren't any issues with that.
https://bugzilla.redhat.com/show_bug.cgi?id=1318674
Hi,
Here's a quick patch from upstream commit, it should fix the crash:
https://github.com/mate-desktop/mate-power-manager/commit/7facec455d825d317caffb8cafc4e4fec60065d9.patch
Hi,
It's fixed in upstream releases 1.18.1 and 1.16.1.
For now, here's the patch from upstream commit:
https://github.com/mate-desktop/mate-utils/commit/550229e422cd88fe39473f41cebe6cef12f6c8fb.patch
Hi,
Sure, I'll provide the links to upstream issue, commit, etc.
Hi,
All the updates mentioned above are now in Stretch. The last one was Marco
which migrated there today.
We (upstream) have more fixes to be backported to 1.16 branch, but it will
happen later, and they'll have to be applied as patches after Stretch is
released. New upstream releases probably
Hi,
Is something else needed from upstream or maintainers to unblock marco?
It's already uploaded to Unstable, but still has the block and 10-day
delay...
I think this is the last of MATE bugfix releases that we wanted to get
into Stretch.
Hi,
This is great, I see all the listed packages are now in Unstable, and all
unblock
requests are approved. Niels even lowered the migration delay to 2-3 days for
these
packages. However, marco and mate-themes are left with 10 days... maybe contact
Niels about these two?
I also added a
Hi,
Thanks for unblocking. This is really a good improvement over version 3.22.6.
We (upstream devs) have been testing various themes from the package, and with
each next version they definitely get better. :)
Hi,
Should this report be left open then? If the menu item isn't there, the dialog
isn't accessible at all. It can't be re-enabled via dconf-editor, there's no
such option there... it should be done explicitly in xml file which is modified
by the patch.
Hi,
Thanks a lot :)
Is it currently possible to have a migration delay less than 10 days?
There's less than 10 days left until the deadline on June 9th...
Hi,
Yes, that pull request fixed it, and the fix is included in Caja 1.16.3.
Hi,
For the record, here's the current list of bugs fixed in various
1.16.x releases:
caja [1.16.6]
- https://bugs.debian.org/757851
- https://bugs.debian.org/826261
- https://bugs.debian.org/836352
- https://bugs.debian.org/828158 [partially fixed]
eom [1.16.1]
-
Hi,
Forgot to mention one thing: the backgrounds and emblems dialog is currently
completely disabled (menu item to call it is removed) by the patch:
_do-not-show-property-browser-in-menu.patch
So, that patch has to be removed in order to make the dialog available again.
Also if unblocks will be accepted, https://bugs.debian.org/862172 can be solved
along the way...
For the record, the fix is included in version 1.16.1 upstream.
Hi,
Thanks for getting back on this.
One correction on the changelog though: https://bugs.debian.org/828158 isn't
fully fixed,
as I mentioned in the comment there. Drag-n-drop of emblems/backgrounds works,
but you
still need to resize the dialog to make them appear in it. Sorry, we still have
Reassigning the bug to metapackage since fixing it requires upgrading
of both mate-desktop (to 1.16.2) and Caja (to 1.16.3). Also merged all
the duplicate reports since they're all about the same regression.
Actually, it's recommended to upgrade Caja to the latest 1.16.x version
which is
Ok, Marco 1.16.1 is released with the fix upstream.
Hi,
Turned out the needed commits were present only in the master branch,
they didn't make it to 3.22 branch. This is now fixed, and 3.22.11 is
released. Please upload it.
Side note: package tracking system still shows 3.22.9 as the latest
release available, despite the fact that 3.22.10 was
Hi,
It was probably not this library, but mate-settings-daemon. I think it got
updated
on the same day.
Source: mate-desktop
Version: 1.16.1-1
Severity: normal
Tags: stretch sid patch
Currently gir1.2-mate-desktop package lacks runtime dependencies on other needed
GI packages, like gir1.2-glib-2.0, gir1.2-gtk-3.0, libgirepository-1.0-1. This
can be solved by enabling dh_girepository helper:
diff
Control: tags -1 jessie
Hi,
Actually, it's possible to patch mate-panel in Jessie to fix this.
The trick is to make use of 01_tasklist_orientation.patch from libwnck
to set the proper orientation of the tasklist.
It was already done before for mate-panel 1.12 (see [1]), but the patch
never
Control: reassign -1 marco 1.16.0-1
Control: tags -1 stretch sid upstream
Control: forwarded -1 https://github.com/mate-desktop/marco/issues/251
Hi,
This is a problem in Marco, the window manager. Reassigning the bug to it.
We have a possible fix upstream, currently it's being tested.
See the
Hi,
This utility should not affect desktop colors. You can copy color code from it
to use in some other application, that's all.
Hi,
Bug #826636 is now fixed by mate-screensaver 1.16.1-1.
Do you still get this issue with that version?
This had been partially fixed upstream. While the backgrounds & emblems dialog
still
needs resizing to make the contents appear, the actual applying of colors,
backgrounds
and emblems to folders now works as before. You can drag-n-drop a color or a
background
pattern from the dialog to the icon
Source: mate-terminal
Version: 1.16.2-1
Severity: important
Tags: stretch sid upstream fixed-upstream
Control: forwarded -1 https://github.com/mate-desktop/mate-terminal/issues/171
Quick double-clicking with right mouse button can freeze Xorg in a way that
keyboard and mouse input will be
Various background issues, caused by changes in GTK+ 3.22, had been fixed
upstream. The following packages are needed:
mate-desktop 1.16.2
mate-settings-daemon 1.16.2
caja 1.16.3
Since Caja 1.16.3 requires build and runtime deps on mate-desktop 1.16.2,
the following change to debian/control is
Hi,
Looks like this one can be closed.
Forcemerge probably failed due to not reassigning the package first...
I would try it myself, but I don't know if it can be done for archived
bug reports (#844785 is archived).
Window Preferences -> Placement -> Enable side by side tiling -> uncheck
Hi,
This is not a Caja bug. This is an old problem that affects also text editors
(gedit, pluma), other desktops (Xfce), etc etc etc.
google: m files octave Objective-C
Some results:
https://bugs.launchpad.net/ubuntu/+source/gtksourceview2/+bug/451816
Control: found -1 1.16.2-2
Control: tags -1 fixed-upstream
This is finally fixed upstream in 1.16.5 release.
Incidentally, it's the same fix as for https://bugs.debian.org/836352
Control: tags -1 - moreinfo + sid stretch upstream fixed-upstream
Control: forwarded -1 https://github.com/mate-desktop/caja/issues/755
Control: found -1 1.16.2-2
So this was finally reproduced and subsequently fixed upstream.
The fix will be included in the upcoming 1.16.5 release.
Hi,
All runtime warnings should be fixed by now in the latest upstream release,
1.16.1.
Control: tags -1 upstream fixed-upstream jessie
Control: forwarded -1 https://github.com/mate-desktop/mate-panel/issues/444
Control: fixed -1 1.14.1-1
Hi,
I guess this is what was reported upstream at [1]. If it's so, we have it
fixed - at least in Stretch (as it was fixed in version 1.14.1).
Control: tags -1 moreinfo
Hi,
It might be caused by some issues with GTK+3 at that time. In MATE 1.16
most issues related to GTK+ had been fixed. Does this issue still happen?
Hi,
> Vlad, if possible, always add a fixed statement when tagging an issue as
> "jessie" only.
> Sometimes, I simply guess what version may have fixed a bug. But for the
> release team,
> this fixed statement on the bug report is important. They want to see that
> bugs are
> already closed
Source: poppler
Version: 0.48.0-2
Severity: normal
Tags: patch upstream fixed-upstream
Control: forwarded -1 https://bugs.freedesktop.org/97870
Control: affects -1 atril evince
Due to a bug in poppler [1], thumbnailers of atril and evince were crashing
on some PDFs. It had been fixed upstream in
Control: tags -1 jessie
This is not reproducible in 1.16, so it sounds like Jessie-only issue.
Source: intltool
Version: 0.51.0-3
Severity: important
Tags: stretch sid patch
This had been reported upstream (and for Ubuntu) at [1].
Out-of-tree builds, including 'make distcheck', are broken with automake >=
1.15.
Because of this, various projects that make use of intltool had to add weird
Hi,
So 4.2.14 is in Jessie now. But recently I've upgraded my Jessie machine to
Stretch,
so I can't test if 4.2.14 really fixes this. But I think it should since it
contains upstream
patches... so you probably can close it now - or should I close it?
Control: tags -1 moreinfo
Hi,
It was just another issue of Caja 1.14 not fully compatible with GTK+ 3.21/3.22
(which you probably got on upgrade at the time). Version 1.16 should work fine.
So, does the problem still happen for you?
Control: tags -1 moreinfo
Hi, upstream MATE developer here.
It's been a while, and MATE 1.16 is now in Unstable and Testing,
with various fixes for GTK+ 3.22 support. I can't reproduce this
bug now. Do you still see it? If it's so, which view (icon view or
list view) and which themes can be
Matt: oops, I completely forgot about that additional patch, thanks for
reminding.
Mike: to avoid FTBFS with tzdata >= 2016g, you'll also need a patch from
1.16.0-1... It was named
1001_adapt-to-timezone-namechange-for-Asia-Yangon.patch. Since it's removed now
(upstream 1.16.1
already
Control: reopen -1
Hi,
Unfortunately, new versions of tevent and samba are stuck in s-p-u repo for
almost two months. They're not in stable repos yet. I have no idea why.
https://tracker.debian.org/pkg/samba
https://tracker.debian.org/pkg/tevent
https://bugs.debian.org/836795
Control: reassign -1 src:libmateweather 1.8.0-2
Control: tags -1 jessie
Hi,
The problem is that libmateweather is still not patched in Jessie.
Maintainers need to apply this patch to fix it:
http://paste.ubuntu.com/23149947/
Hi,
> Sorry for being such a pain. There indeed is a .config/gtk-3.0/gtk.css
> but it doesn't tell where it is coming from and it does have
> :prelight in the .css file.
>
> I have attached the .css file although I don't know should I change it
> from :prelight to :hover or just delete it ?
Oh, forgot about one more dir to look in.
As described in GTK+ docs at [1], there might be gtk.css
in ~/.config/gtk-3.0/ dir ($XDG_CONFIG_HOME is ~/.config).
[1]
https://developer.gnome.org/gtk3/stable/GtkCssProvider.html#GtkCssProvider.description
Hi,
Now I'm out of clues as well.
I can only suggest system-wide searching for files named "gtk.css"
which contain ":prelight" text on lines 63 and 73.
Hi,
Just in case, check if you have something in these dirs too:
/usr/local/share/themes
~/.local/share/themes
1 - 100 of 431 matches
Mail list logo
