[07:29] * bryce waves to tseliot [09:16] hi bryce [11:54] well that was exciting, hit my brightness up keys a few times, my screen filled with purple garbage and X restarted \o/ [11:54] nothing in Xorg.0.log.old, or /var/crash/ :/ [14:28] hello [14:30] can someone please help me with bug 306970? [14:30] Launchpad bug 306970 in xserver-xorg-video-i128 "xorg i128_drv.so: undefined symbol xf86usleep" [Undecided,Confirmed] https://launchpad.net/bugs/306970 [14:38] hello, anyone here? [14:57] pcguy11: just try jaunty [14:57] I have to stick with intrepid [14:58] then try to build the jaunty version [14:58] can I just install the jaunty i128 driver on intrepid? [14:58] no [14:58] you have to build it [14:59] so I would have to change apt-get to point to jaunty, do a apt-get source xserver-xorg-video-i128 [14:59] apt-get build-dep xserver-xorg-video-i128 [14:59] right? [15:00] or use pbuilder [15:00] and grab the source pkg files first [15:00] directly from the archive, not with apt [15:01] What is pbuilder? Can you point me to some documentation please? [15:01] google :) [15:01] I'm not really here [15:02] I will give pbuilder a try. Thanks [15:03] np [15:03] bbq -> [16:05] tseliot, ping. wanted to talk to you about those scripts that are (eventually) going to live to make people's lives sane if they install an nvidia driver that isn't yet packaged [16:06] superm1: yes, I haven't worked on them yet [16:06] superm1: basically they should remove my packages first [16:06] tseliot, well i had some ideas for them, did you have a wiki page started? [16:06] when users run the nvidia installer [16:07] no, not yet. Feel free to create a new one [16:07] okay well let me pass them by you now at least then and then i can try to transcribe into a wiki page [16:07] so they should probably also 1) install DKMS [16:08] 2) register the driver with DKMS (and if possible use DKMS for the first build) [16:08] this wasn't the purpose of the helper scripts [16:09] maybe not the purpose, but can there be an extension to do these types of things? [16:09] basically I talked to Aaron about the problems which are caused when you use his installer and my packages are already installed [16:09] i guess I don't know the breadth of when these scripts get executed. I had assumed there was a preinst and postisnt equivalent [16:10] think of them as preinst, postinst, prerm, postrm scripts [16:10] they cannot change how the installer works [16:11] but they can still clear the environment before and after the installation/uninstallation of the driver [16:11] okay so then for preinst clean up any installed drivers with apt packages [16:11] yes, exactly [16:11] postinst, take the build stuff and register it with dkms [16:11] this would conflict with nvidia's current behaviour [16:12] how do they currently handle upgraded kernels though? [16:12] I thought you had to rerun the installer for those cases [16:12] as the installer builds and installs the module [16:12] they don't handle this case [16:12] that's what I thought; so it might have to be built twice then - once by their installer and once by DKMS [16:13] ok but this would require more than just a simple postinst [16:13] yeah [16:13] i'm just thinking it's a logical extension if people are going to be wanting to install the drivers from NV and expecting them to work like the drivers in the archive [16:16] unfortunately they don't want to put distro-specific stuff in the installer [16:17] i see. well then perhaps its worthwhile for them to do something similar to what AMD is doing in their rpm postisnt [16:17] where they check for if DKMS is installed, they build and register with it, but they don't require it [16:17] then in the preinst when you remove other packages, you could just add DKMS too? [16:18] so you wouldn't need any more complicated of a postinst then, and you wouldn't be changing their behavior (they would) [16:18] superm1: you can try to ask Aaron via email and CC me [16:19] well I've got a with our contacts on Monday, so let me mention it there first and see what they say [16:20] ok, great [16:20] i'm not sure how closely tied into aaron's group my contacts are. so maybe they'll just direct me at him [16:20] thanks tjaalton [16:21] oops tseliot [16:21] ? [16:22] thanks for discussion and what not. i'll let you know what happens [16:22] superm1: ok :-) === ubott2 is now known as ubottu [17:19] Hi, I tried the jaunty source code for my i128 video card driver and it still gives me the same error message about AddScreen/ScreenInit failed for driver 0 [17:21] see bug #306970 [17:21] Launchpad bug 306970 in xserver-xorg-video-i128 "xorg i128_drv.so: undefined symbol xf86usleep" [Undecided,Confirmed] https://launchpad.net/bugs/306970 === pcguy11 is now known as temp === temp is now known as pcguy11 [17:54] pastie: hi [19:14] howdy [19:15] I've got a strange situation here with intrepid where X works but logging into gnome gives us a blank screen. this is a fresh install. [19:15] does it ring any bell? we already tried removing compiz and that wasn't it [20:51] tjaalton: any thing holding us up from rolling out the xserver in git? [21:13] bryce_: nope [21:16] tjaalton: ok cool, mind if I upload right now, or would you like to do the honors? [21:17] bryce_: go ahead .) [21:17] :) [21:17] (alpha-6 freeze is coming up, so I want to make sure we get stuff in) [21:17] ok will do [21:17] shall I do libdrm as well, or did you have some plans with that? [21:18] not other than perhaps dropping the nouveau-backport, since it should be included in 2.4.5 [21:18] is that just a patch? [21:19] yes, IIRC [21:19] ok [21:19] in git [21:19] is libdrm managed in git? [21:19] yes [21:19] oh [21:20] ok, I'll give it a poke [21:20] mesa too btw ;) [21:20] sigh [21:20] right [21:20] not that I've been much help lately [21:21] I think I forgot to commit my last change to mesa [21:21] heh [22:10] tseliot: btw is the python-randr api code mentioned on https://wiki.ubuntu.com/X/Blueprints/ScreenConfigurationUI available somewhere? [22:11] not yet, I haven't touched that code from December [22:12] bryce_: I would like to write something in C with dbus so that we can call it from other languages such as python [22:13] tseliot: would you be interested/willing to mentor a GSoC student for this work? [22:13] (if ubuntu does GSoC this year) [22:14] bryce_: for which project? [22:14] the former? [22:15] for the C/dbus work. Or is that not something you think is appropriate for a student to do? [22:16] bryce_: it seems pretty tough for a student, at least IMO [22:16] I guess I should back up... would you be interested in / have time for being a mentor for a student on some X-related project? [22:20] bryce_: I don't know how it works and I haven't talked to smagoun yet about what he want me to do. Would it be ok if I gave you a reply next week after I talk to him? [22:21] certainly [22:22] ok [22:38] good night