Note: This is a public test instance of Red Hat Bugzilla. The data contained within is a snapshot of the live data so any changes you make will not be reflected in the production Bugzilla. Email is disabled so feel free to test any aspect of the site that you want. File any problems you find or give feedback at bugzilla.redhat.com.
Bug 1226604 - background slide show timing no longer correct with respect to system time zone and background file's start time
Summary: background slide show timing no longer correct with respect to system time zo...
Keywords:
Status: CLOSED ERRATA
Alias: None
Product: Fedora
Classification: Fedora
Component: mate-control-center
Version: 22
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: Wolfgang Ulbrich
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2015-05-31 00:55 UTC by April
Modified: 2015-07-23 09:00 UTC (History)
5 users (show)

Fixed In Version: mate-desktop-1.10.1-3.fc21
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2015-07-23 08:55:13 UTC
Type: Bug
Embargoed:


Attachments (Terms of Use)

Description April 2015-05-31 00:55:59 UTC
Description of problem:
I'm not sure if I'm reporting this on the correct component, sorry ...

The Fedora Infinity background has 4 images which transition depending upon the time of day ... sunrise, day, sunset, and night ... these seem to be shown at the wrong times in Fedora 22 ... I'm getting sunrise when sunset should appear, then day when night should appear, sunset when sunrise should appear, and night when day should appear. This works correctly in Fedora 21 but not Fedora 22.

I'm using the Mate Desktop environment, if that's relevant.


Version-Release number of selected component (if applicable):
0.0.5-10

How reproducible:
100%

Comment 1 Mathieu Bridon 2015-05-31 08:44:38 UTC
I can't reproduce this issue here, with GNOME on Fedora 22. It works just fine.

Maybe it is a bug in MATE instead?

Comment 2 April 2015-05-31 11:01:13 UTC
I guess I just don't know what Component to file this bug under. The XML file for the background looks fine; I assume this is an issue with whatever software in the Mate Desktop takes care of the background transitions....

Comment 3 Wolfgang Ulbrich 2015-05-31 14:10:56 UTC
Sorry, mate-backgrounds ships only backgrounds and nothing more.
Why you always think it is abug if something don't work?
Changing backgrounds works fine in rhel7 with mate.
So there is no bug in mate.
@ Mathieu Bridon
reasign back to default commponent
close or do what ever you want with the report

Comment 4 Wolfgang Ulbrich 2015-05-31 14:28:20 UTC
PS:
/usr/share/backgrounds/infinity/1-infinity-sunrise.png
/usr/share/backgrounds/infinity/2-infinity-day.png
/usr/share/backgrounds/infinity/3-infinity-sunset.png
/usr/share/backgrounds/infinity/4-infinity-night.png
/usr/share/backgrounds/infinity/infinity.xml
/usr/share/doc/fedorainfinity-backgrounds/COPYING
/usr/share/gnome-background-properties/desktop-backgrounds-infinity.xml
This directory will never work with mate!

Comment 5 April 2015-05-31 14:37:17 UTC
Wolfgang,

I understand that the mate-backgrounds package only ships backgrounds, but I am not sure which component of Mate is responsible for background transitions (these are called "Slide Shows"). Gnome2 and Mate have a feature which allows an XML file to serve as a background, transitioning through different images automatically throughout the day. A couple such slide shows are included in the backgrounds that ship with Fedora Mate Spin ("Cosmos", for example). The Fedora Infinity background is one of these XML files which is specifically set up to show different images based upon the time of the day (actually, it looks like maybe all of these are supposed to align with specific times of day, but the fact that this stopped working isn't as obvious for photo slide shows as it is for "Fedora Infinity" which has distinct time-of-day backgrounds). This works correctly up to Fedora 21, but does not work correctly in Fedora 22. In Fedora 22, whatever Mate software controls these transitions is not aligning those transitions to the correct time of day.

This issue is not about manually changing the background, it is about the break to time of day alignment for background slide shows.

The Fedora Infinity background does in fact work with Mate ... go to Appearance Preferences->Background->Add ... for file type, switch to "All files" ... navigate to /usr/share/backgrounds/infinity, the select the XML file. This is the same for other slide shows such as "Cosmos".

Comment 6 April 2015-05-31 15:08:37 UTC
Ah, I see you've pointed out there's an unrelated issue with the packaging -- ideally, desktop-backgrounds-infinity.xml would also be added to /usr/share/mate-background-properties so that the user doesn't have to browse for the file manually.

Comment 7 Wolfgang Ulbrich 2015-05-31 15:11:54 UTC
I'm currenty testing it.
But is this enabled?
[rave@mother ~]$ gsettings get org.mate.background background-fade
true

PS: because of the tons of reports you opend in the last 5 days against mate packages, i'm thinking more and more that something is complete wrong with you installation or you use wrong settings or use old configurations in home folder
I can't reproduce any of your report in f22 and i've feedback from a lot of others peoples that all is running well in f22.
Maybe you have an issue with you graphic driver.
PLease post provide 'systemctl -b 0', .xsession-errors to one of the other reports. I will dig into it to see what's wrong with your system.
And please provide hardware informations and which architecture you use, and complete list of which packages are installed.

Thank you

Comment 8 Wolfgang Ulbrich 2015-05-31 15:13:42 UTC
(In reply to April from comment #6)
> Ah, I see you've pointed out there's an unrelated issue with the packaging
> -- ideally, desktop-backgrounds-infinity.xml would also be added to
> /usr/share/mate-background-properties so that the user doesn't have to
> browse for the file manually.

No, fedorainfinity-backgrounds should ship this file to
/usr/share/mate-background-properties/
too.

Comment 9 Wolfgang Ulbrich 2015-05-31 15:14:41 UTC
(In reply to Wolfgang Ulbrich from comment #8)
> (In reply to April from comment #6)
> > Ah, I see you've pointed out there's an unrelated issue with the packaging
> > -- ideally, desktop-backgrounds-infinity.xml would also be added to
> > /usr/share/mate-background-properties so that the user doesn't have to
> > browse for the file manually.
> 
> No, fedorainfinity-backgrounds should ship this file to
> /usr/share/mate-background-properties/
> too.

Sorry, missread your comment

Comment 10 Wolfgang Ulbrich 2015-05-31 15:17:04 UTC
(In reply to Wolfgang Ulbrich from comment #7)
> And please provide hardware informations and which architecture you use, and
> complete list of which packages are installed.
mate packages , of course

Comment 11 April 2015-05-31 15:18:32 UTC
Also, if desktop-backgrounds-infinity.xml does in fact get copied to /usr/share/mate-background-properties, that copy should have "gnome-wp-list.dtd" changed to "mate-wp-list.dtd".

Comment 12 April 2015-05-31 15:40:45 UTC
(In reply to Wolfgang Ulbrich from comment #7)
> I'm currenty testing it.
> But is this enabled?
> [rave@mother ~]$ gsettings get org.mate.background background-fade
> true

background-fade is enabled.

> 
> PS: because of the tons of reports you opend in the last 5 days against mate
> packages,

I only have 4 open reports against Mate packages. I withdrew my first report.

> i'm thinking more and more that something is complete wrong with
> you installation or you use wrong settings or use old configurations in home
> folder

I verified the checksum of the installation media (F22 Mate x86_64), and I have performed a clean basic install inside of VirtualBox for testing which replicates the same issues that affect my primary system. The VirtualBox install has all default settings, and I reset it to the fresh installation snapshot before testing different issues. The VirtualBox install has no content in the home folder carried over from any other installation.

> I can't reproduce any of your report in f22 and i've feedback from a lot of
> others peoples that all is running well in f22.

Could you please indicate specifically what you can not reproduce? Are you referring to this bug? You should be able to reproduce by installing fedorainfinity-backgrounds, then selecting the background as I outlined in a prior comment, then looking at your background and seeing what color it is. It is *supposed to be* blue at 1:00 p.m. (13:00), pink at 6:00 p.m. (18:00), dark purple at 12:00 a.m. midnight (00:00), green at 7:00 a.m. (07:00), and a blend of two colors between these times.

> Maybe you have an issue with you graphic driver.
> PLease post provide 'systemctl -b 0', .xsession-errors to one of the other
> reports. I will dig into it to see what's wrong with your system.

Please ask me for the information in whatever bug you believe it applies to. I think that will be less confusing.

> And please provide hardware informations and which architecture you use, and
> complete list of which packages are installed.

As far as general hardware information is concerned, I don't think it's relevant as I am able to reproduce the issues in a default VirtualBox installation of F22 Mate Spin. The architecture is x86_64. The complete list of packages is the default that comes with F22 Mate Spin, as I am able to reproduce the issues in that setup.

> 
> Thank you

Comment 13 Mathieu Bridon 2015-05-31 18:32:28 UTC
(In reply to Wolfgang Ulbrich from comment #8)
> (In reply to April from comment #6)
> > Ah, I see you've pointed out there's an unrelated issue with the packaging
> > -- ideally, desktop-backgrounds-infinity.xml would also be added to
> > /usr/share/mate-background-properties so that the user doesn't have to
> > browse for the file manually.
> 
> No, fedorainfinity-backgrounds should ship this file to
> /usr/share/mate-background-properties/
> too.

Ah, so I need to add that file there, then?

Do you have a pointer to a documentation explaining how this works, and why the file is needed in that folder? (just to make sure I don't do anything wrong)

Comment 14 Wolfgang Ulbrich 2015-05-31 18:47:34 UTC
(In reply to Mathieu Bridon from comment #13)
> (In reply to Wolfgang Ulbrich from comment #8)
> > (In reply to April from comment #6)
> > > Ah, I see you've pointed out there's an unrelated issue with the packaging
> > > -- ideally, desktop-backgrounds-infinity.xml would also be added to
> > > /usr/share/mate-background-properties so that the user doesn't have to
> > > browse for the file manually.
> > 
> > No, fedorainfinity-backgrounds should ship this file to
> > /usr/share/mate-background-properties/
> > too.
> 
> Ah, so I need to add that file there, then?
> 
> Do you have a pointer to a documentation explaining how this works, and why
> the file is needed in that folder? (just to make sure I don't do anything
> wrong)
Good question, but mate is very similay to old gnome2.
I quess it's the same reason why you installed it for gnome directory.
But
[root@mother rave]# ls /usr/share/mate-background-properties/
cosmos.xml                        f22.xml            mate-nature.xml
mate-abstract.xml                 f21-extras.xml

All this themes are visible out of box in mate-appearence-properties.

Comment 15 Mathieu Bridon 2015-05-31 19:17:57 UTC
(In reply to Wolfgang Ulbrich from comment #14)
> (In reply to Mathieu Bridon from comment #13)
> > (In reply to Wolfgang Ulbrich from comment #8)
> > > No, fedorainfinity-backgrounds should ship this file to
> > > /usr/share/mate-background-properties/
> > > too.
> > 
> > Ah, so I need to add that file there, then?
> > 
> > Do you have a pointer to a documentation explaining how this works, and why
> > the file is needed in that folder? (just to make sure I don't do anything
> > wrong)
> Good question, but mate is very similay to old gnome2.
> I quess it's the same reason why you installed it for gnome directory.

To be honest, I have no idea why that file is there, I took over this package because I like the background, and its original maintainer had orphaned the package. :)

In fact, I think when I got that package GNOME 3 had already been released for some time. ^_^

> But
> [root@mother rave]# ls /usr/share/mate-background-properties/
> cosmos.xml                        f22.xml            mate-nature.xml
> mate-abstract.xml                 f21-extras.xml
> 
> All this themes are visible out of box in mate-appearence-properties.

I'm not sure that would be related to this issue, though, given that this file doesn't actually contain the time for the transitions, they are in /usr/share/backgrounds/infinity/infinity.xml

So let's try something.

Can any of you test this scratch build, to see if it fixes this issue?

    http://koji.fedoraproject.org/koji/taskinfo?taskID=9898394

A test would confirm, though.

If it fixes the issue, I'll send it to Rawhide and F22.

Comment 16 April 2015-05-31 19:38:20 UTC
That file just has some metadata and tells the desktop environment's GUI for changing the background about the existence of Fedora Infinity at /usr/share/backgrounds/infinity/infinity.xml.

This means that when the user opens the background changing GUI, Fedora Infinity shows up as an option, so the user doesn't have to browse for the XML file manually.

You are correct that this is an entirely separate issue from this bug report.

I tested your scratch build and can confirm that it has the intended result of adding Fedora Infinity to the Mate GUI for changing the background.

Comment 17 Mathieu Bridon 2015-05-31 20:01:09 UTC
(In reply to April from comment #16)
> That file just has some metadata and tells the desktop environment's GUI for
> changing the background about the existence of Fedora Infinity at
> /usr/share/backgrounds/infinity/infinity.xml.
> 
> This means that when the user opens the background changing GUI, Fedora
> Infinity shows up as an option, so the user doesn't have to browse for the
> XML file manually.
> 
> You are correct that this is an entirely separate issue from this bug report.
> 
> I tested your scratch build and can confirm that it has the intended result
> of adding Fedora Infinity to the Mate GUI for changing the background.

Ah, ok, I had misunderstood.

So adding that file doesn't fix this bug report, then... :(

So here's a stupid idea, April... How is the time displayed on your desktop? If it's in the 12 hours format (i.e "8:30pm" instead of "20:30"), are you sure that you haven't misconfigured your clock so that am and pm are inverted?

Cause that could very well explain the issue. :)

If not, then I'm really clueless. The /usr/share/backgrounds/infinity/infinity.xml contains the proper times for the transitions, as far as I can see, so I don't think the bug is in the fedorainfinity-backgrounds package.

Comment 18 April 2015-05-31 20:12:54 UTC
I have tested with both time formats, and my clock is indeed using the correct time :-)

I agree that the bug is not in fedorainfinity-backgrounds. The question is which part of Mate controls this. Maybe mate-settings-daemon?

I've also tested offsetting the start time in /usr/share/backgrounds/infinity/infinity.xml by 12 hours, and this doesn't make any difference. It seems that the issue is that the start time specified in the XML file is no longer recognized by the software which updates the background throughout the day.

Comment 19 April 2015-05-31 20:40:47 UTC
I did some further testing to get more precise info regarding the amount of the offset ... it looks like it's 11 hours ahead (or 13 hours behind). It shows the sunrise image at full force starting at 20:00 rather than starting at 07:00.

I tested by changing my system clock, then changing the desktop background and changing it back so that it immediately updates it's timing relative to the clock.

Comment 20 Wolfgang Ulbrich 2015-05-31 20:48:56 UTC
It's not worse to give me some time to test if using .xml's files with timinigs works here. As i said in comment 3 , all is working well with using default xml background files in rhel7 or centos7. i never test this in fedora.
So please relaxe, i will reasign it to right commponent if it is a mate bug.
Hey, don't foget it's late sunday.

Comment 21 Mathieu Bridon 2015-05-31 20:55:34 UTC
(In reply to Wolfgang Ulbrich from comment #20)
> It's not worse to give me some time to test if using .xml's files with
> timinigs works here. As i said in comment 3 , all is working well with using
> default xml background files in rhel7 or centos7. i never test this in
> fedora.
> So please relaxe, i will reasign it to right commponent if it is a mate bug.

Thanks Wolfgang!

In any case, should I still push an update adding that XML file in the mate directory, then?

Comment 22 April 2015-05-31 21:22:18 UTC
Wolfgang,

I'm just providing information I feel may be relevant. I am not pressuring anyone to work on this right now :-)

I've determined that there are two pieces of logic which seem to have gone missing which are causing the 11-hour (in my case) offset ... whatever software controls the background slide shows no longer accounts for the user's time zone, and also no longer accounts for the start time specified in the background XML file. In my case, Eastern Standard time is UTC-04:00, and the Fedora Infinity XML file specifies that the animation starts at 07:00, so if you add 4+7 hours you get the 11-hour offset that I'm observing.

You should be able to reproduce this 11-hour offset by:
1) switching to a different background and then closing Appearance Preferences
2) switching time zone to America/New York
3) opening Appearance Preferences and selecting Fedora Infinity

The result should be that the backgrounds are offset by 11 hours.

I have confirmed this result by changing to London time on my system (at which point the offset is reduced accordingly).

Comment 23 Wolfgang Ulbrich 2015-06-01 11:44:49 UTC
For better testing i have set all duration values to 150s (3min), with a resulting cycle of 24 min.
And here the backgroungs changes, After every 6 min the next background is ful applied and after 24 min i have seen all backgrounds and the game starts again.
So, changing/fading backgrounds is working in mate session.

So reseting all values back to default and started the background animation again, it's 2pm here. And i see the the sunset background (maybe in duration to night, not shure), here which duration should start at 3pm. :-/

So reading the correct time seems to be quite broken.
I remember that something is new in f22 with contolling the UTC times, maybe systemd or something else.
All i can say is that nothing is changed in mate-control-center for timming backgrounds during developing from mate 1.8 to 1.10.
see https://github.com/mate-desktop/mate-control-center/commits/master

I will reasing it to mate-control-center for the momment but i'm pretty shure that the issue is caused by a change in a more general system application.

(In reply to Mathieu Bridon from comment #21)
> (In reply to Wolfgang Ulbrich from comment #20)
> > It's not worse to give me some time to test if using .xml's files with
> > timinigs works here. As i said in comment 3 , all is working well with using
> > default xml background files in rhel7 or centos7. i never test this in
> > fedora.
> > So please relaxe, i will reasign it to right commponent if it is a mate bug.
> 
> Thanks Wolfgang!
> 
> In any case, should I still push an update adding that XML file in the mate
> directory, then?
Yes, that would be a great improvement.

@ April
Since the anoying audacious bug all my activties at redhat bugzilla are observed by this email address bugs.michael.
I do not aggre with that.
I'm living in a free country and that's a violation of my privacy rights.
For this reason i decided not to use redhat bugzila any more until redhat admins
revert this decision by cwg.

I you have probs with mate please file out a report directly at upstream.
https://github.com/mate-desktop
I will help you there.

Comment 24 Wolfgang Ulbrich 2015-06-01 11:48:47 UTC
@ April
Please file out a report for this at https://github.com/mate-desktop/mate-control-center
Let us working there.

Comment 25 April 2015-06-01 13:13:44 UTC
Reported here: https://github.com/mate-desktop/mate-control-center/issues/159

The new report gives a clearer outline of what you should see at what times.

Comment 26 Mathieu Bridon 2015-06-01 17:31:10 UTC
(In reply to Wolfgang Ulbrich from comment #23)
> > In any case, should I still push an update adding that XML file in the mate
> > directory, then?
> Yes, that would be a great improvement.

Ok, done:

https://admin.fedoraproject.org/updates/fedorainfinity-backgrounds-0.0.5-11.fc22

If any of you could give it some karma, it would be great. :)

Comment 27 Fedora Admin XMLRPC Client 2015-06-17 22:16:26 UTC
This package has changed ownership in the Fedora Package Database.  Reassigning to the new owner of this component.

Comment 28 Fedora Update System 2015-07-20 15:22:19 UTC
mate-desktop-1.10.1-3.fc21 has been submitted as an update for Fedora 21.
https://admin.fedoraproject.org/updates/mate-desktop-1.10.1-3.fc21

Comment 29 Fedora Update System 2015-07-20 15:22:29 UTC
mate-desktop-1.10.1-3.fc22 has been submitted as an update for Fedora 22.
https://admin.fedoraproject.org/updates/mate-desktop-1.10.1-3.fc22

Comment 30 Fedora Update System 2015-07-21 08:15:50 UTC
Package mate-desktop-1.10.1-3.fc22:
* should fix your issue,
* was pushed to the Fedora 22 testing repository,
* should be available at your local mirror within two days.
Update it with:
# su -c 'yum update --enablerepo=updates-testing mate-desktop-1.10.1-3.fc22'
as soon as you are able to.
Please go to the following url:
https://admin.fedoraproject.org/updates/FEDORA-2015-11861/mate-desktop-1.10.1-3.fc22
then log in and leave karma (feedback).

Comment 31 Fedora Update System 2015-07-23 08:55:13 UTC
mate-desktop-1.10.1-3.fc22 has been pushed to the Fedora 22 stable repository.  If problems still persist, please make note of it in this bug report.

Comment 32 Fedora Update System 2015-07-23 09:00:52 UTC
mate-desktop-1.10.1-3.fc21 has been pushed to the Fedora 21 stable repository.  If problems still persist, please make note of it in this bug report.


Note You need to log in before you can comment on or make changes to this bug.