|
Bugzilla – Full Text Bug Listing |
| Summary: | Gnome applet settings retention: "touchpad", "volume" and time/date | ||
|---|---|---|---|
| Product: | [openSUSE] openSUSE 11.0 | Reporter: | Jim Omura <jimomura> |
| Component: | Usability | Assignee: | E-mail List <bnc-team-screening> |
| Status: | RESOLVED NORESPONSE | QA Contact: | E-mail List <qa-bugs> |
| Severity: | Normal | ||
| Priority: | P5 - None | ||
| Version: | Final | ||
| Target Milestone: | --- | ||
| Hardware: | x86-64 | ||
| OS: | openSUSE 11.0 | ||
| Whiteboard: | |||
| Found By: | Customer | Services Priority: | |
| Business Priority: | Blocker: | --- | |
| Marketing QA Status: | --- | IT Deployment: | --- |
| Attachments: |
SaX.log 2008/06/27 15:31 Toronto time
xorg.conf 2008/06/27 15:33 Toronto time Sax.log 2008/06/27 16:06 xorg.conf 2008/06/27 16:09 Toronto time SaX.log 2008/06/28 11:30 Toronto time xorg.conf 2008/06/28 11:32 Toronto time |
||
|
Description
Jim Omura
2008-06-21 03:24:00 UTC
Looking at the open bug list, it appears that there are at least 2 bugs that are probably related to this one: #402864, opened on June 23, 2008 #402536, opened on June 23, 2008 These later reports were posted as "i586" version bugs, but the operation of the bugs seem to be similar to this one. The culprit seems to be "GSynaptics", though sax2 is also mentioned in the 402536 report. To where does "GSynaptics" write its output? I expect that there has to be a "xxx.conf" file somewhere in the HOME directory, but I did not see anything obvious. I found the output file which is:
/root/.gconf/desktop/gnome/peripherals/touchpad/%gconf.xml
My intention was to shut off "tap to click" and "virtual scrolling" leaving the touchpad to operate in a "plain" manner. The result file was:
<?xml version="1.0"?>
<gconf>
<entry name="accel_factor" mtime="1213956777" type="int" value="13">
</entry>
<entry name="horiz_scroll_delta" mtime="1213956780" type="int" value="0">
</entry>
<entry name="vert_scroll_delta" mtime="1213956782" type="int" value="0">
</entry>
<entry name="max_tap_time" mtime="1213929612" type="int" value="0">
</entry>
<entry name="off" mtime="1213929599" type="bool" value="false">
</entry>
</gconf>
Looking at it, I expect that if I remove the last line it might work. However, right now, I have simply renamed the file, so it is not being used. That way, I can use OpenSuSE 11.0 as-is until there is a fix.
Three Bugs In One? So far, I have encountered 3 bugs. The first one I noticed was the "touchpad" bug. The 2 others are as follows: The second is the "sound". When I first ran the "Live" disc, the sound worked, and when I installed 11.0, the sound worked on the first boot, but was far too loud on my computer, so I adjusted it using the program with the icon in the bottom bar right side. After that, it stopped working and has not worked since. The third bug is the "time/date". When I first installed 11.0, I set the time and date, and I think it was ok. At some point I noticed it was wrong and I readjusted it, probably by using the application on the bottom bar. And after that it kept on being wrong until I adjusted it with a different application. Eventually, I set the time/date in Windows (this is a dual-boot setup with Vista 32) and am no longer touching the settings on the OpenSuSE side. Eventually I will try it again. The 3 bugs have the following in common: 1. The original setting made during the installations worked. 2. After attempting to change settings, they "stopped working". The "time/date" clock still worked but some settings were clearly lost. 3. As far as I know, all are "GNOME". It seems to me that they also all have to do with saving setting data, and I expect probably in XML. So I think that it is possible that they might be related by a common library. If this is true, then fixing one might fix all of them. Because of this, I have decided not to open separate bug reports for the 2 other bugs at this time. If the "touchpad" bug is fixed and the others are still a problem, then I will probably open new bug reports for them later. If someone else opens up similar bug reports, well, that might not turn out to be a bad idea anyway, but for now, I think it might be a waste of time. I have been reading Bug #402536 which resolved on June 26. Unfortunately, I just recently installed the ATi drivers "fglrx" which required changes to the xorg.conf file through sax2, so both of my logs have been replaced and I have no copies of the originals (unless Linux archives them somewhere that I do not know about). Anyway, I am going to try to upload my current SaX.log and xorg.conf and later I will see if anything has changed. Actually, I can say right now that the sound is still not working, but maybe the touchpad problem has been solved. Created attachment 224913 [details]
SaX.log 2008/06/27 15:31 Toronto time
Current SaX.log after adding ATi "fglrx" driver
Created attachment 224915 [details]
xorg.conf 2008/06/27 15:33 Toronto time
This is the xorg.conf created by sax2 upon updating to use the ATi "fglrx" driver (which is working properly).
Well, I can now say that the result of adding "fglrx" was that the "touchpad" is being treated purely as a mouse now which means that "gsynaptics" is now gone and I have lost the ability to make most of the touchpad related changes. I will try "sax2 -r" this time and see what happens next. Created attachment 224919 [details]
Sax.log 2008/06/27 16:06
I ran "sax2 -r" at level 3 and this was the new log.
Created attachment 224920 [details]
xorg.conf 2008/06/27 16:09 Toronto time
This is the resulting xorg.conf after running "sax2 -r".
Running "sax2 -r" did resulted in roughly the same result as when I ran "sax2" to use the ATi "fglrx" driver. Maybe someone who knows more about these files might spot something, but I see nothing significant. The only thing I see is that "sax2" apparently does not know about "gsynaptics" and so I guess it might help if I re-install the "gsynaptics" package. I am too short of time to experiment like this, so for now I am going to leave this alone. I will probably look at it again after someone is assigned to this bug. I decided to take a chance and re-install "gsynaptics". Using "YaST2>Software Management" I un-installed "gsynaptics" Then I rebooted and did a fresh install. I found another package called "YaST2-mcs-plugin" which seemed to be related, but which was not a dependency, and decided to load it also. I rebooted to run level 3 and ran "sax -r" I then used "startx" and found that the "touchpad" was working again. I made my usual adjustments. I logged off from Gnome/X and used "shutdown -h now". I booted fresh into run level 5, and the mouse cursor locked up. I rebooted again to run level 3 and renamed the xml file and rebooted again. Now everything is the same as before (the touchpad is working but with the defaults). I will upload the SaX.log and xorg.conf files. Created attachment 225004 [details]
SaX.log 2008/06/28 11:30 Toronto time
This is the current SaX.log after re-installing "gsynaptics".
Created attachment 225005 [details]
xorg.conf 2008/06/28 11:32 Toronto time
This is the xorg.conf file after re-installing "gsynaptics".
It is interesting that "gsynaptics" seems to be working as usual, but that it is not mentioned at all in either the current "SaX.log" nor in the "xorg.conf". This is not what appears to be the case in bug #402536. I re-read the last comments in bug #402536 and saw that I had misunderstood it. The references were to "synaptics" which is a device type or driver and not the "gsynaptics" program. I did somewhat understand that there might be an issue whether the touchpad was set as "CorePointer" or "SendCoreEvents". Mine was the opposite from the settings that worked in the other bug report. So I decided to test this possibility. I edited the "xorg.conf" file with "gedit" and only swapped the 1 and the 3 in the "ServerLayout" section. I tested it in the usual way and after re-booting (after making changes to through "gsynaptics"), it still locked up. I undid the changes and everything is as before. I also found some old backup files of "xorg.conf" including the file just prior to installing ATi's "fglrx", and the sections regarding the mouse and touchpad were the same as the are in the "xorg.conf" files I have already uploaded. I will not upload them unless requested. I have separated the Sound problem as Bug #405088. It does not seem to be related to this touchpad problem. I now expect that the clock setting problem is also a separate problem. I might open another bug report for it later. On further thought, the "sound" problem could still be related to the "touchpad" since they are both losing settings from session to session. I have changed the summary of this bug, but will leave the other one as-is. Last Call: I am currently deciding whether to remove 11.0 and try 10.3. If anyone wants me to upload anything like a "conf" file or "log" file, now is a good time to ask. I have replaced 11.0 with 10.3 and cannot provide further informations about these problems. Best of luck! |