<html> <head> <meta content="text/html; charset=UTF-8" http-equiv="Content-Type"> </head> <body text="#000000" bgcolor="#FFFFFF"> <div class="moz-cite-prefix">On 8/8/2013 2:58 AM, Nio Wiklund wrote: </div> <blockquote cite="mid:520341A8.2040206@gmail.com" type="cite"> <pre wrap="">On 2013-08-08 03:01, Aere Greenway wrote: </pre> <blockquote type="cite"> <pre wrap="">On 08/07/2013 05:04 PM, John Hupp wrote: </pre> <blockquote type="cite"> <pre wrap="">For what it's worth, I have just found that the workaround detailed in Comment #1 in the bug report (<a class="moz-txt-link-freetext" href="http://bugs.launchpad.net/ubuntu/+source/linux/+bug/1178982">http://bugs.launchpad.net/ubuntu/+source/linux/+bug/1178982</a>) does work. On my system there was no existing /etc/X11/xorg.conf, so I created it and added the specified lines as the sole content of the file. The colors and the proper window size were restored. This Dell has an Intel 845G chipset, so this workaround may fix this problem on any motherboard with the same chipset (or even other Intel chipsets that use the same Intel driver). ------------------------------------------------------------------------ Verging strictly off-topic, but remaining with the question of getting Flash to work decently well: On this former XP machine with a Celeron 2.4 GHz and 1 GB RAM, YouTube videos in the default window size and playing at 360p seemed to perform normally. Likewise, video from Hulu can be set to a lower quality to help assure continuous play. But video from Vimeo can only be set to HD-Off (if HD is available). And with video from the broadcast network sites CBS.com, NBC.com and ABC.com, you can only change screen size. So it seems that videos from Vimeo, CBS, NBC and ABC offer very little accommodation for lower-spec setups. And relatedly, I find that video that plays OK on a Windows PC with a dual core Intel E2200 @ 2.20 GHz -- even with just 1.3 Mbps download on my DSL service -- plays badly on the 2.4 GHz Celeron using the same Internet connection. So in this case processing power is more important than Internet connection speed. 2.4 GHz is the minimum required spec for Flash (the last I knew), but perhaps that merely means that you'll be able to play *something* (like YouTube or Hulu videos at a lower-quality setting), not that you'll be able to play everything. Does anyone know if there is a way to lower the quality settings for sites like Vimeo, CBS, NBC and ABC, even if there is no quality-setting tool in the player interface? (Or does anyone differ with the assessment I offer above?) </pre> </blockquote> <pre wrap="">John: Thank you very much for your reply. There's a lot of really good information in it. -- Sincerely, Aere </pre> </blockquote> <pre wrap="">+1 Thank you very much for describing and discussing this workaround :-) I was able to get good graphics with Saucy alpha 2 in my old IBM Thinkcentre desktop computer, that has suffered from bad graphics since Raring. I used your workaround, John, and it worked without issues with the log in screen. My Thinkcentre has the following graphics VGA compatible controller: Intel Corporation 82865G Integrated Graphics Controller (rev 02) I'll attach two text files for clarity, one with the tips explicitly stated, and one with the output of lspci on my Thinkcentre. I used the installed system for USB according to the following link <a class="moz-txt-link-freetext" href="https://help.ubuntu.com/community/InstalledSystemFakePAE">https://help.ubuntu.com/community/InstalledSystemFakePAE</a> I don't notice the bad graphics with my wall-paper (for Saucy alpha2), but switch to the default one, and you see it! After adding uxa acceleration, the default wallpaper is rendered as it should, without the sharp boundaries, 'jagged' as described by Aere concerning the Raring wallpaper. -o- I found also this link about UXA for Intel graphics <a class="moz-txt-link-freetext" href="http://task3.cc/135/intel-graphic-cards-linux-xorg-and-uxa-performance-boost/">http://task3.cc/135/intel-graphic-cards-linux-xorg-and-uxa-performance-boost/</a> -o- @ Phill: Where could this workaround be added into the Lubuntu Wiki? Best regards Nio </pre> </blockquote> I'm trying to find a workaround for the workaround and fix this garbled login screen. But to round out the knowledge of the problem and solution, I observe that uxa acceleration solved a different problem for Nio than the one I was first addressing (with Flash content displaying in green and purple in a compressed window). Nio wrote: "I don't notice the bad graphics with my wall-paper (for Saucy alpha2), but switch to the default one, and you see it! After adding uxa acceleration, the default wallpaper is rendered as it should, without the sharp boundaries, 'jagged' as described by Aere concerning the Raring wallpaper." And to pick out another bit worth noting, he observed that uxa causes no problem with the login screen, but his case deals with Saucy alpha 2 rather than Raring, so I merely underline those particulars. So now I'm wondering how uxa fouls up the LightDM login screen on Raring but not on Saucy. I'm hanging on by my troubleshooting fingernails here looking for clues, but <a href="https://en.wikipedia.org/wiki/LightDM">https://en.wikipedia.org/wiki/LightDM</a> notes that LightDM "uses various front-ends to draw login interfaces, so-called Greeters." The apparent Greeter of interest here is the GTK+ Greeter. Following the the references in the LightDM wiki, I see that 12.04 used lightdm-gtk-greeter (1.1.5-0ubuntu1). 13.04 uses v 1.5.1-0ubuntu1 of the Greeter (and v 1.6.0-0ubuntu3 of LightDM). What version of the Greeter is Saucy using? Are there some LightDM/Greeter settings I can play with to see if any of them yields a clean login screen? Or can we compare some Raring vs. Saucy configurations and learn something useful? <hr size="2" width="100%"> And closer to the original topic, it seems that no one has refiled the closed bug. If I can boot with an earlier kernel and the Flash problem does not appear as noted there (I have not tested this myself), is that enough to conclude that this bug should be filed against the kernel? </body> </html>