[00:30] I saw request for help with Nvidia Drivers I would like to help out [00:32] Which request? Nouveau or the binary nvidia drivers? [00:33] proprietary drivers, so I think binary [00:34] Sorry I failed to read who to email [00:46] yotux, contact ara [00:47] am going to do so thank you [03:54] bryceh: merged mesa 7.7-4, though it pulled from the stable branch and probably would need some paperwork [03:55] didn't upload of course due to a3 [03:58] tjaalton, ok [04:01] do you think we need standing FFe's for mesa & xserver? there's 7.7.1 coming late March and 1.7.6 "some time" [04:13] tjaalton, as long as we can justify them as bug-fix only releases, I think we're covered by the point release policy [04:14] tjaalton, although it couldn't hurt to do an FFe anyway if you think it'd help to have added visibility on them [04:16] bryceh: alright, maybe when those point-releases happen then :) [04:17] btw, I've started tagging bugs by the release they were reported against, so when looking at bugs we can query for has_tag==lucid [04:18] is there a tag for edgers= [04:18] ? [04:18] no, but there probably should be [04:18] do you like 'edgers'? I can add that to the processor [04:20] it's alright I guess [04:20] done [04:21] the script runs once a week on saturday [04:23] ok [04:36] 1.6.4 feels sooo old [04:38] ancient.. === |thade| is now known as Alexia [18:07] hi === toabctl_ is now known as toabctl [18:08] hi [18:08] can anybody help with bug #527912 ? [18:08] Launchpad bug 527912 in xf86-input-wacom "Wacom Bamboo CTH-661 not recognized" [Undecided,New] https://launchpad.net/bugs/527912 === Sinnerman is now known as Cobalt [19:09] toabctl: http://sourceforge.net/mailarchive/forum.php?thread_name=4B3A8A6B.20205%40openwebboard.org&forum_name=linuxwacom-discuss [19:11] tjaalton, why do you know that the kernel component doesn't support the tablet? [19:12] toabctl: because the id isn't listed there [19:12] ok [19:16] none of the new bamboo's are [19:22] hm. shit [19:34] tjaalton: i'm pretty sure mesa shouldn't require a FFe, i dont see any new features since our snapshot just bugfixes? xserver probably would though since it reenables the record extension though then again thats just fixing something that was broken.. :D kklimonda: got a dmesg from it failing with the newer nouveau? I added the no firmware required patch to that one [19:36] Sarvatt: what change in mesa? [19:36] 7.7-4 from 7.7-3 [19:37] ok [19:39] maybe we should drop mach64 since its not in the kernel and would save ~1MB of cd space [19:40] if squashfs compresses similar to gzip -9 that is [19:40] makes almost no difference in the deb size since its lzma but its about 930k at gzip -9 [19:41] it required building the drm kernel modules from libdrm 2.4.15 or older and all that was dropped from libdrm [19:41] Sarvatt: ok [19:42] is the cd-size an issue still? [19:43] i see them talking about dropping language packs all the time in -devel and not being able to fit french or german right now [19:43] Sarvatt: seems sane to me.. [19:43] heh, ok [19:46] i810 and sis i'm unsure about too, i never was able to get sis to work at all but I havent found any more info on it [19:57] Sarvatt: I got nothing in dmesg but there was a "CtxProg is still running" message printed on vt along with some other message - I didn't have a camera to make screenshot and couldn't get it to display again later so I guess it doesn't help :) [19:58] Sarvatt: no - wait, I have it [19:59] Sarvatt: http://pastebin.com/u3yLdEvW - SAK is from me trying to get some response from system :) [20:14] Sarvatt, -sis we should keep; I periodically get bug reports on it so someone must use it [20:14] i810 can go, that's long obsolete [20:15] I think with disk space, the dropping of Gimp bought a bit of breathing room, but we're adding nouveau bits now so if there's stuff we can excise from the cd it'd be a nice thing to do [20:15] and of course debdiffs welcome :-) === yofel_ is now known as yofel [20:24] tjaalton, in xf86-input-wacom-0.10.4/src/wcmUSB.c are some lines like:{ 0xD3, 2540, 2540, &usbBamboo }, /* CTL-660 */ [20:24] but there is no line for the cth-661 [20:25] i would like to add a line but i don't know what the 3 parameters stand for. [20:28] toabctl: so the cth-661 and ctl-660 share the same id [20:28] no need to change anything [20:29] linuxwacom has the needed kernel bits [20:29] too bad that the git repo doesn't [20:30] tjaalton, linuxwacom==linuxwacom-0.8.4 ? and git==xf86-input-wacom? right? [20:30] for the kernel, git == linux-2.6.git, i guess [20:31] tjaalton, i don't know where to find the kernel module for wacom [20:31] wacom-kernel git actually [20:34] toabctl: get the linuxwacom tarball [20:34] 0.8.5-10 [20:35] tjaalton, and there's the code for the kernel module? (the code organisation looks very confusing for me) [20:35] yes (and it is) [20:38] tjaalton, and where can i find the kernel driver in the ubuntu repository? [20:39] comes with your kernel. /lib/modules/*/kernel/drivers/input/tablet/wacom.ko [20:41] right [20:41] jcristau, and how to find out from which source this module was build? from 0.8.4? or 0.8.5.10 ? [20:41] toabctl: what came with the kernel [20:42] linuxwacom used to ship the code as well [20:42] see what version it declares [20:42] in the linuxwacom kernel source [20:42] tjaalton, so the module is shiped by the mainline kernel? it's not a ubuntu patch? [20:42] right [20:42] ah. ok. [20:45] the kernel code was split into wacom-kernel git, but seems like it's not being updated [20:52] tjaalton, and how do i know which kernel version i need for linuxwacom-0.8.5-10? this version doesn't compile for me.. [20:53] toabctl: you don't want or need to compile the whole of linuxwacom [20:53] diff the kernel code and patch the kernel sources, or just copy the files and then build the kernel [21:25] tjaalton, thanks for the help. but i think it's to complex for me to fix it. [21:25] toabctl: :) [21:26] tjaalton, do you think it's a big thing or will tha tablet be supported in some weeks? or no idea? [21:27] toabctl: I'm not sure it'll be supported in lucid at all, tbh [21:27] judging by what was said on #ubuntu-kernel earlier [21:27] tjaalton, ok. thanks a lot for the help and information! [21:31] tseliot, is it intended that users are switched from nvidia->nouveau on karmic->lucid upgrade? [21:32] i just upgraded one of my installs two days ago and this happened to me [21:32] superm1: I don't think so [21:32] would it be expected if I had nvidia-glx-195 installed (rather than -190) though? [21:33] er -190 rather than -185 that is [21:34] also, what happens to the nvidia-xxx-libvdpau package in such a situation? [21:35] i can file a bug with whatever logs came up for this to help diagnose [21:35] but i suspect it's the same thing that will happen to anyone using that ppa of yours bjsnider [21:35] would the appropriate thing be 'ubuntu-bug update-manager' to grab all such logs? [21:36] superm1, you were using the 190 driver out of my ppa? [21:36] bjsnider, i'm not sure if it was your PPA or the ~mythbuntu PPA. it's basically the same packaging either way though [21:37] superm1: we have 195 in Lucid now and, unless your package was epoched it should have gone well [21:37] certainly wasn't [21:38] tseliot, why not add a conflicts: nvidia-xxx-libvdpau to nvidia-current? [21:38] but nvidia-glx-185 might not have been installed [21:38] because of it being named differently in the ppa's [21:38] so the upgrade process might not have known what to do about it [21:38] superm1: I guess my packages only replace nvidia-glx-185 [21:39] well i'll file the bug when i get home and we can all look and see what's possible for a solution [21:39] i'm certain there's lots of users using these PPAs for newer drivers, so it would be best to try to cover as many of them as possible [21:39] even if the solution is pushing something new to those PPAs for people to upgrade to before the upgrade [21:40] i could do that [21:40] it was certainly a poor experience when i tried to go play a recording and the machine just hung there because it couldn't find a suitable vdpau backend anymore :) [21:41] superm1, you mean it went to nouveau without you realizing it at the time? [21:41] bjsnider, yup [21:41] i mean there wasn't any tell-tale signs since i didn't use 3d, and i still had 1080p being output [21:42] i only used the video playback acceleration of vdpau [21:42] nouveau must work pretty well on your hardware [21:42] you wouldn've had kms too [21:43] i didn't pay attention to the bootup - the tv was off, so i have no idea if KMS was working [21:43] i did notice the redraws were a little slow in the menus though [21:44] superm1, what is the difference between mythbuntu and mythtv? [21:44] bjsnider, mythbuntu is a preconfigured installation with lots of things to better integrate with the OS setup for you [21:45] and extra tools for configuration of other pieces that aren't part of mythtv either [21:45] you can certainly add any of them to an ubuntu install though [21:45] such as what extra things? [21:45] mythbuntu-control-center is the biggest [21:45] it's kinda a one-stop shop for configuration of anything HTPC related on the box [21:45] so it's supposed to integrate with ubuntu better than mythtv by itself would? [21:46] well all the stuff is in the ubuntu archive, so you can achieve the same results yourself by installing all the pieces if you want [21:46] the stuff on the PPA's is just daily builds of mythtv [21:48] i see [21:48] some of the goals of the ubuntu desktop clash with what you would normally use an HTPC for too, so this only gives you the components you're gonna need for HTPC usage [21:48] is it built for multiple distros? [21:49] No, it's ubuntu only [21:49] superm1: what's the PPA? [21:49] our cdimages are produced on cdimages.ubuntu.com etc [21:49] tseliot, the ~mythbuntu PPA? [21:49] superm1: the one with the nvidia packages [21:49] superm1, i mean different ubuntu distros, like jaunty, intrepid etc. [21:49] tseliot, it was either this https://edge.launchpad.net/~mythbuntu/+archive/trunk-0.22 or the vdpau ppa [21:49] BTW I made 195 available in my proprietary drivers PPA [21:49] bjsnider, oh yeah it's been since 7.10 [21:50] cool [21:50] project is a few years old now [21:50] nvidia-current is the 195 at this point isn't it? [21:50] i'm pretty sure our nvidia-graphics-drivesr-190 on that PPA came from the vdpau PPA [21:50] bjsnider: yep [21:51] it was just put here because mythtv had to be built against libvdpau for it to work [21:51] superm1, i just updated it last week or something. i hope you're staying current with it [21:51] superm1: maybe it was some libvdpau that caused the problem? I'm too tired to tell ;) [21:52] tseliot, i haven't even reviewed the logs yet. i was still sorting out other upgrade problems [21:52] against ibvdpau? [21:52] i'll get the bug filed tonight with all the pieces [21:52] my driver doesn't install it, i use the external package [21:52] bjsnider, no it hasn't been updated. is it something important? [21:53] which 190 is it? is it .42? [21:53] 190.53 [21:53] that driver does not come with libvdpau anymore [21:54] only the nvidia vdpau driver is packaged in that one [21:54] yeah, hence we have the external libvdpau on that PPA too [21:54] and that's what mythtv had to build against [21:54] oh, ok [21:54] particularly libvdpau-dev [21:55] wel, hte most recent update added execstack in karmic, updated prerm/postrm scripts, basically in line with the standard karmic package [21:55] okay so for the users, if they're up and working probably not critical stuff [21:56] no [21:56] but whatever, a lot of them probably use the vdpau ppa [21:56] bryceh: hey, shouldn't the apport hook links be named according to the source package? [21:56] we don't have any stats or numbers of any kind do we? [21:58] i dont think so, but that would be awfully nice [21:58] it would help to know which PPAs to give priority to keeping more stable [21:58] i know ppa usage stats were proposed at one point [21:58] i'd like to know how many people are using old distros, because i think it's a lot [21:58] tjaalton, for manual reporting, yes [21:58] i get emails from jaunty and even intrepid people all the time [21:59] tjaalton, for crash detection it needs to be against the binary package [21:59] tjaalton, did you spot an inconsistency? [22:00] bryceh: yeah, there's no xf86-input-evtouch, -wacom.. [22:00] probably others too [22:01] tjaalton, heh, I just happened to fix a whole bunch just yesterday [22:02] tjaalton, unreleased but check the xorg git tree [22:02] you removed xf86-input-evtouch and wacom-tools? [22:02] tjaalton, a second set of eyes to doublecheck that everything's correct would help [22:02] er, evtouch was renamed [22:03] well, replaced it with source_xserver-xorg-input-wacom.py [22:03] but the source package is xf86-input-wacom :) [22:03] so maybe they both need to be there, for crashers and manual stuff [22:04] wacom-tools is good to go, since lucid doesn't have it anymore [22:04] $ apt-cache search wacom [22:04] xserver-xorg-input-wacom - X.Org X server -- Wacom input driver [22:04] weird. [22:05] anyway yeah feel free to add whatever to the x11-common.links file [22:05] that's the binary package [22:05] I don't look at updating it as often as I probably should [22:05] apt-cache show xserver-xorg-input-wacom |grep Source [22:06] yeah I'll add those [22:21] boas noutes [22:27] toabctl: usb.ids is missing a bunch of wacoms, like all the intuos4's [22:28] but it has nothing to do with the funcionality of the driver [22:28] +t [22:28] tjaalton, but it's a first step, isn't it? [22:28] toabctl: no, it's a "good to have for a nicer lsusb output", but that's it [22:29] my intuos4 works just fine without that [22:29] tjaalton, ah.ok [22:29] tjaalton, your intuos4 works with lucid? [22:29] yes [22:29] then i know what to buy next :) [22:31] tjaalton, the newest kernel driver for wacom is available in the wacom-kernel git in drivers/input/tablet, right? [22:31] toabctl: no, it's not updated yet. it's in the latest linuxwacom tarball [22:32] tjaalton, in which directory? the names ended with 2.6.27 [22:32] there [22:33] diff it against the ones in 2.6.33 and you see what has changed [22:34] hmm, there's thel linuxwacom cvs too [22:34] -l [22:34] lol [22:35] http://linuxwacom.cvs.sourceforge.net/viewvc/linuxwacom/linuxwacom-dev/ [22:38] the commits are really messy.. "Updated new Bamboo support and fix a tablet rotation issue [22:38] " [22:38] tjaalton, i installe the latest xorg driver now (xf86-input-wacom-0.10.4) and the kernel module from linuxwacom-0.8.5-10 and when i insert the tablet now, i got in /var/log/message: [22:38] Feb 25 23:37:19 zitrone kernel: [ 8308.664191] usb 3-1: USB disconnect, address 3 [22:38] Feb 25 23:37:25 zitrone kernel: [ 8315.380103] usb 3-1: new full speed USB device using uhci_hcd and address 4 [22:38] Feb 25 23:37:25 zitrone kernel: [ 8315.546334] usb 3-1: configuration #1 chosen from 1 choice [22:38] Feb 25 23:37:26 zitrone kernel: [ 8315.553354] input: Wacom BambooFun 2FG 6x8 Pen2 as /devices/pci0000:00/0000:00:1a.0/usb3/3-1/3-1:1.0/input/input18 [22:38] Feb 25 23:37:26 zitrone kernel: [ 8315.568645] input: Wacom BambooFun 2FG 6x83 as /devices/pci0000:00/0000:00:1a.0/usb3/3-1/3-1:1.1/input/input19 [22:38] that also includes support for intuos4 OLEDS [22:39] ok, that's better [22:39] and something happen when i move my finger over the tablet and when i press anywhere. but it's not usable.. [22:39] what does udevadm info --export-db show now? [22:39] or the Xorg.0.log? [22:40] does the pen work? [22:40] tjaalton, btw do you know why it is that debian renames xf86 -> xserver-xorg ? [22:41] bryceh: no.. [22:41] takes longer to type?-) [22:41] seems to me that for consistency sake, -evtouch and -wacom ought to be renamed to match [22:41] the pen works more or less (maybe it's me who don't know how to use) [22:41] yes, but they are not maintained by xsf [22:41] otherwise I wonder why not just go with xf86- and become consistent with upstream [22:42] tjaalton, are they ubuntu-only packages? [22:42] they are going to be merged to the server anyway, so [22:42] bryceh: no, -evtouch is maintained by malattia, -wacom by ron [22:42] toabctl: ok, that's great then [22:43] tjaalton, here's the output of udevadm info [22:43] toabctl: do you have the module somewhere? if it's i386 and built for the -14-generic kernel I might try it on mine [22:43] http://paste.ubuntu.com/384000/ [22:44] tjaalton, yes arch=i386 and the newest kernel from repository [22:44] and i updated the package of xf86-input-wacom with uupdate [22:45] i'll write it down in the bug report how to fix it. [22:45] toabctl: hmm it loads synaptics and evdev too [22:45] do you want to have the kernel module? [22:45] so maybe the udev rules need an update [22:45] yes please [22:45] one more test before I pass out :) [22:49] toabctl: well? [22:49] tjaalton, https://code.launchpad.net/~toabctl/+junk/bamboo [22:49] tjaalton, can i finetune the response of the tablet anyway? [22:49] toabctl: it loads the wrong driver for it [22:50] the touch function that is. I see synaptics being loaded [22:50] could be that wacom doesn't support that yet [22:50] tjaalton, ah. so new udev rules should fix this? [22:50] I don't know, check the wacom lists [22:51] tjaalton, the pen only work if i don't touch the tablet. i have to "fly" over the tablet with the pen [22:51] that's probably evdev then [22:51] yes you could try to fix the udev rules [22:53] tjaalton, should i report a bug against the actual kernel? is it possible that the new driver from linuxwacom will be merged into the ubuntu lucid kernel? [22:53] I get an error "invalid module format" [22:53] tjaalton, i build with generic-pae kernel [22:53] toabctl: I added the linux task already [22:53] ok, I don't have pae here [22:53] anyway, that's ok [22:54] tjaalton, i filled already a bugreport to debian that they should update xf86-input-wacom to 0.10.4 [22:54] and I'm not sure if this will be added to lucid [22:54] heh, good luck [22:54] tjaalton, why? because ron don't answer? [22:55] toabctl: if he thought 0.10.4 was worth it, it would've been packaged already I think [22:55] but he's right most of the time too [22:56] ok [22:56] the current one is patched, since vanilla 0.10.3 didn't work properly [22:59] tjaalton, to get the pen working correct, what i have to do? [22:59] not load evdev for it [22:59] tjaalton, and how to do that? [22:59] ie. make it match in the wacom udev rules [22:59] with udev? [22:59] ok [23:00] so it doesn't fall through the cracks so that the evdev rules catch it [23:00] anyway, been up 20h already and can't sit straight anymore. night! [23:01] tjaalton, bye and thanks a lot!!! [23:01] tjaalton, you are my hero!:-) [23:05] bryceh: I was talking about the mesa dri drivers, sorry