The indicator blinks between what looks like an empty graph (in a vertically shorter box) and the actual graph (in a full-sized box). This is best demonstrated in a video so I’m attaching one below.

Both the official repository version (0.4) and the one from daily PPA (0.5) exhibit this. Different color schemes don’t seem to affect it (i.e. the problem remains). There’s further problems when I add a third graph, but this report is just about the blinking.

(Bug #1336828 is about the indicator missing entirely in 14.04, so I decided to report this slightly different issue separately.)

• Indicator (0.5) blinking in Lubuntu 14.10 Edit (1.1 MiB, video/ogg)

For the ’Incorrect padding’ issue, Bug #1012358. (Leaving this as non-duplicate, since the reporter has other ’damaged’ errors besides that one.)

I acknowledge that (and as I said, it hasn’t really caused any issues with the backups), I was just responding to this being ”fixed” (above). If it’s not really an issue in the code, maybe we should close this as invalid? (Although I could argue that the message makes it seem as if there is a problem with the code when in fact there isn’t, so the bug could be ”this warning is slightly misleading”. But that’s up for interpretation.)

Sorry, but 5.0.0-1ubuntu1 didn’t fix this. In Utopic, with the same ’The Snowman’ DVD as before, I still get:

*** Zero check failed in src/ifo_read.c:903
for pgc->subp_control[i] = 0x00000001
*** Zero check failed in src/ifo_read.c:903
for pgc->subp_control[i] = 0x00000001
*** Zero check failed in src/ifo_read.c:903
for pgc->subp_control[i] = 0x00000001

jani@ubudev:~/lumiukko$ apt-cache policy libdvdread4
libdvdread4:
Installed: 5.0.0-1ubuntu1
Candidate: 5.0.0-1ubuntu1
Version table:
*** 5.0.0-1ubuntu1 0
500 http://archive.ubuntu.com/ubuntu/ utopic/universe amd64 Packages
100 /var/lib/dpkg/status

I’ve worked around that enum error (since before owncloud 7 already) by adding this in 3rdparty/doctrine/dbal/lib/Doctrine/DBAL/Platforms/MySqlPlatform.php:

@@ -661,6 +661,7 @@
     protected function initializeDoctrineTypeMappings()
     {
         $this->doctrineTypeMapping = array(
+            'enum' => 'string',
             'tinyint'       => 'boolean',
             'smallint'      => 'smallint',
             'mediumint'     => 'integer',

Reopening. It didn’t go away after all, looks like the occurrence varies (didn’t see it yesterday), perhaps those Chromium settings just made it more probable.

I take that back, it seems to have been caused by Chromium all along (more specifically either #threaded-compositing-mode or #deadline-scheduling), despite those glitches appearing outside of Chromium too. Sot it’s a Chromium bug, but I’m too lazy to debug this further (the issue goes away with those configuration flags set to default) so I’ll just mark this as invalid.

• Kuvakaappaus-blink.ogv-1.png Edit (768.6 KiB, image/png)

Also, here’s a screenshot (from the video) with the artifacts visible. I notice they seem to cover mostly just Chromium’s content area, but just last night I had this occur just as I was logging out (with the lines remaining on screen until I tried to take a screenshot), with nothing but the log out confirmation running on top of my desktop, so it’s probably not just Chromium that’s causing this.

• blink.ogv Edit (1.4 MiB, video/ogg)

Ha, finally caught a couple of glimpses of this in a screen recording! Attaching it here.

Since upgrading HWE from -lts-saucy to -lts-trusty, I have had recurring graphical glitches on screen, with short black horizontal lines appearing briefly on screen (on top of normal contents), particularly when switching between applications. The lines appear only for a few 100 ms before going away (just enough to register in your eye that there’s something there), so I’ve been trying to make a screen recording to capture it, but of course the phenomenon goes away when I do (perhaps RecordMyDesktop does something with the screen that makes it less likely to appear).

I have another 12.04 system with desktop Radeon graphics, and there -lts-trusty has not produced this issue so far (the one with the issue and the one I’m attaching the logs here from has Intel).

I don’t recall seeing this with -lts-saucy or other previous Xorg packages. It was so obvious right after upgrading that either it never occurred before or was so rare or quick to flash away that it didn’t register.