[00:00] bryceh: where are the ubuntu git branches? if I was doing this for the kernel, I would just do git tag --contains 5725849a1b427cd4a72b84e57f211edb35838718 [00:00] and see which releases had it [00:00] but I have no idea where to find the ubuntu X git trees :) [00:01] debcheckout [00:01] I guess reading https://wiki.ubuntu.com/X/GitUsage might help [00:02] RAOF: ah, yeah, that works too. I'm so used to only doing that with bzr that I didn't even try with git. durrr [00:02] It works with anything that's got a VCS-* tag :) [00:02] yeah :) [00:04] kees, yeah they're kept at alioth [00:15] RAOF: that appears to only be the debian/* tree... [00:16] kees: You've run ‘debcheckout xorg-server’, right? [00:18] and xserver-xorg and xorg. [00:19] s/^and /only / [00:19] doing xorg-server now :) silly names! [00:20] \o/ much nicer [00:21] Yeah. xorg is just a wrapper :) [00:21] * kees nods [03:28] I got the sandybridge laptop today. Mostly good so far. Let me know if there's stuff you need tested. [08:22] hi! I have problems with x-updates ppa [08:22] I did purge [08:23] but under synaptic when I check e.g. libdrm-nouveau1, I still see maintainer Ubuntu X-SWAT [08:23] and it is [08:23] I also tried to add ppa again, update [08:23] and then again remove [08:23] but it's still there [08:23] what's the problem then? [08:24] how to get ubuntu 10.10 default drivers [08:24] 10.10 does have that package you know [08:25] also the maintainer is the same [08:25] My problem is that I have GF GTS 250. Prop drivers does not work good with comput (very slow!) and opensource driver is just too bad [08:25] but it's fast [08:25] here is also my ubuntuask question http://askubuntu.com/questions/27907/compiz-slow-under-proprietary-nvidia-driver [08:26] yes, it sucks [08:26] oh, I didn't know [08:26] I mean x-updates nouveau driver is FAST on compiz [08:27] 300PFS with blur [08:27] 150 FPS in expo [08:27] BUT, after time I get texture errors [08:27] blinking etc... [08:27] it's useless [08:27] and it's doesn't even run google earth [08:27] yes, it's not mature yet [08:27] it uses mesaSW fallback [08:28] even less so on 10.10 [08:29] any idea why nvida prop driver works bad for compiz? (it is good for 3D games) [08:29] nope [08:30] but it's not only me who had problems? [08:30] no [08:32] darn... what do you suggest. What about xorg-edgers? I didnt spot difference between x-updates and edgers for my graphics [08:33] there are unresolved crashers with nouveau, doesn't matter what version you run [08:37] problem is with "newer" GPU-s, right? my GF8600GT works very well with prop driver on ubntu 10.04 [08:38] i've got the same, and it's sluggish on 11.04 [08:40] GTS 8600 or GTS 250? [08:42] gf8600gt [08:47] but under 10.04 works just great? did it worked good for you under 10.04? [08:47] dunno, i guess [08:47] any idea where is the problem for slow compiz? maybe old driver version could help? [08:47] not that interested tbh :) [08:47] old version wouldn't work with the server [08:48] darn... [08:50] you said that many people has problem with slow nvidia driver (compiz)? [08:50] no i didn't [08:50] just confirmed it's not just you [09:05] sorry I confused [09:06] tjaalton, can you help me with this? http://askubuntu.com/questions/27907/compiz-slow-under-proprietary-nvidia-driver [09:07] gsedej_work: how? [09:14] how to get nvidia prop driver work fast under compiz? [09:15] I get like 5 FPS when in expo screen [09:16] i've no idea, ask nvidia [09:19] ok, thx [09:42] Hi! I wanted to use xorg-edgers. when I did apt-get upgrade I get next line. What does it mean? how to solve it [09:42] E: /var/cache/apt/archives/ia32-libs_20090808ubuntu10+maverick~xorgedgers0.20110325.1_amd64.deb: trying to overwrite '/lib32/libncursesw.so.5.7', which is also in package lib32ncursesw5 5.7+20100626-0ubuntu1 === yofel_ is now known as yofel [13:39] tseliot, I've got a dell mini 10 with the "trackpad from hell" [13:39] in natty, it now has multitouch support [13:39] I've found it to be sooo much better with the jumpycursorthreshold [13:39] but that setting is disabled if the trackpad is multitouch [13:39] why is that? [13:40] I've hacked around that for now so I could play with it on my machine === cnd` is now known as cnd [14:11] cnd: I didn't know that feature would be disabled with multi-touch. I guess it only made sense with single-touch touchpads [14:12] tseliot, why does it only make sense with single-touch touchpads [14:12] cnd: also when I wrote that patch there was no such thing as multi-touch so I'm wondering if some other patch prevents it from working [14:12] hmmm [14:13] tseliot, the patch actually disables it for multi-finger [14:13] but it calls that multitouch in the man page [14:13] cnd: because in theory putting 2 fingers on the touchpad shouldn't cause a jump [14:13] yeah [14:13] though it does on mine :) [14:13] cnd: maybe someone modified my original patch or maybe my memory is really bad ;) [14:13] let me check === lilstevie|lagche is now known as lilstevie [14:16] tseliot, I'm thinking that we should enable it for all touchpads, especially since the touchpads its meant for now are multifinger/multitouch too [14:20] cnd: if that's still a problem then I agree [14:21] cnd: btw where does it say that it's not supposed to work with multi-touch? [14:22] cnd: I'm looking at 114_jumpy_cursor_first_part.patch [14:22] tseliot, look at the man page diff [14:22] "+Set the threshold above which movement events are ignored on single-touch touchpads." [14:23] and at line 127 is the check for has_double [14:26] cnd: good catch. I guess I really forgot about my code ;) [14:26] I'll fix it [14:27] I tseliot, ok, thanks! [14:27] cnd: thanks for reporting the problem [14:27] tseliot, btw, with this fix and the additions I made with the multitouch data, these trackpads actually work quite well now :) [14:27] cnd: excellent :) [14:28] we even can do some three finger gestures in unity, though it won't be pushed to ubuntu till oneiric [14:29] tseliot, you tried to push the jumpy cursor patch upstream right? [14:29] cnd: yes and it's not going to happen [14:30] cnd: have you seen the upstream bug report? [14:30] no [14:30] could you paste a link? [14:31] cnd: https://bugs.freedesktop.org/show_bug.cgi?id=21614 [14:31] Freedesktop bug 21614 in Input/synaptics "Touchpad cursor jumps when two fingers are used" [Normal,Assigned] [14:32] it seems that Yan Li (Intel) even refreshed my patch [14:34] long story short Peter said no and that's it ;) === lilstevie|lagche is now known as lilstevie [14:40] tseliot, don't upload the fixed package quite yet [14:40] I want to review things some more [14:42] cnd: ok [15:07] hi [15:07] any idea why a call to XSyncInitialize can fail? [15:08] is there anything that I should enable/install? [15:12] rodrigo_: one of the cases listed here perhaps? http://www.x.org/releases/X11R7.6-RC1/doc/libXext/synclib.html [15:13] hmm, maybe [15:14] although this is gnome-session 3.0, which has the same code as 2.32, and 2.32 works, but 3.0 fails [15:14] so, which case you mean? different protocol and lib versions? [15:16] rodrigo_: yes. Is this in Natty? [15:16] yes [15:17] hmm, maybe the gnome-session 3.0 package got compiled with a different version than the one I have, that might be [15:17] * rodrigo_ upgrades [15:50] tseliot, I think the issue is actually resolved by the addition of semi-multitouch functionality in the kernel [15:50] allow me to explain :) [15:50] the devices that were jumpy were likely all semi-multitouch devices [15:51] but we didn't know at the time how to turn that functionality on [15:51] in natty, the semi-multitouch mode is turned on for these devices, which also turns on multi-finger mode [15:51] ah [15:51] with multi-finger mode, the jumpy cursor issue should be handled by http://cgit.freedesktop.org/xorg/driver/xf86-input-synaptics/commit/?id=a6ca4d2523904b7ce49edc29ba408979bdf0d45e [15:52] however, in my multitouch patches I have the wrong ordering of a key function [15:52] right, I remember that commit [15:52] and this causes the change in number of fingers to be reported one frame too late [15:52] so here's what I suggest: [15:52] 1. I push a commit to fix the ordering of the function that affects the finger changes [15:53] 2. leave the jumpy cursor patch in for natty just to be safe [15:53] 3. remove the jumpy cursor patch for oneiric going forward [15:53] my hope is that the jumpy cursor patch is now a no-op if all the jumpy devices were really multitouch devices in disguise [15:54] cnd: ok, it sounds like a good plan. Have you tested your fix together with the jumpycursor patch? [15:55] to see if there are regressions of any kind because of my patch? [15:55] yeah, but it's just a noop here because has_double is now true [15:58] cnd: oh, right [15:59] cnd: if your fix makes the Dell mini 10 work correctly then you have my +1000 vote ;) [15:59] it does! [15:59] I guess it's a kernel fix isn't it? [15:59] with the jumpy cursor threshold it caused the cursor to stall if you moved it around too quick [15:59] and this gets around that issue [16:00] it's both a kernel fix and a synaptics fix [16:00] I added multitouch support to synaptics [16:00] and the functionality to make this work properly is built on top of that [16:00] cnd: yes, the threshold patch was meant to be a workaround since I couldn't find the real cause of the problem [16:01] yeah [16:01] it would have been sooo much easier if synaptics would just release their documentation [16:01] what have they gained by keeping the semi-multitouch mode a secret? [16:01] cnd: I doubt that will happen [16:02] cnd: I can't talk about that in public but I agree with you ;) [16:02] heh [16:02] thanks a lot for your work, I'm glad we can get rid of that hack and have things work out of the box :) [16:03] yep :) [18:02] Hello, I'm giving "fglrx" from natty a try again and try how to get my dual-screen setup back. Do I need a xorg.conf to have a virtual size bigger than 1920x1920? (I need 3200x1920) [18:33] bryceh, I just pushed a fix for synaptics to git [18:34] if you get a chance please upload it [18:38] cnd, ok [18:39] bryceh, I'm so excited cause the trackpad from hell is fixed now :) [19:43] hrm, is there a standard way to tag CVE's in debian/changelog? http://git.debian.org/?p=pkg-xorg/app/x11-xserver-utils.git;a=blobdiff;f=debian/changelog;h=62bf1e211dc7d39097d25e683a27ba26d99b2ce1;hp=9f37d5a08c71030e9c2a11ee0af62b37fe3c338e;hb=52b85a64989483df69812ea78e00666dbdcec993;hpb=a976b76500dafd6274b40bc255f93064b66f7559 [20:10] Sarvatt, does the cve string really have to be 450 characters long? [20:11] that seems like it may be overkill [20:11] huh? [20:11] isn't that the version you just posted? [20:12] + * xrdb 1.0.9 (Fixes CVE-2011-0465) [20:12] Sarvatt: ** RESERVED ** This candidate has been reserved by an organization or individual that will use it when announcing a new security problem. When the candidate has been publicized, the details for this candidate will be provided. (http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2011-0465) [20:12] ? [20:13] no, the a976b76500dafd6274b40bc255f93064b66f7559 [20:13] i mean the commit [20:13] maybe i'm thinking of cvs [20:13] which one is the version control system that's been abandoned? [20:28] jcristau: gnome3 requires the pointer barrier patches that arent even in xserver master yet, libx11 1.4.3 isn't as crazy as that :P === maxb_ is now known as maxb === soreau_ is now known as soreau === soreau_ is now known as soreau