[00:19] hyperair: What do you mean by "pixel perfect scrolling"? You mean the smooth-scrolling that gtk3 gets? [00:19] RAOF: yeah that [00:20] RAOF: i just wrote a small test program that dumps out the delta_x/y that it receives via GDK_SMOOTH_SCROLL_MASK, and will test it otu the next time my scrolling freezes up [00:20] I've noticed that it dies in evolution sometimes, but I chalked that up to evolution struggling on the challenging internet connection I have here. [00:20] hmm [00:20] in this case it dies in everything that uses Gtk3's GtkScrolledWindow [00:22] the last time this happened, i ran an evince window in gdb and found that it was generating + and - delta_y events which were cancelling each other out [00:22] Ah, right. [00:22] Yeah, now noticed that. [00:23] reloading the psmouse module fixes it [00:24] http://paste2.org/p/1983874 [00:24] Ra raw. [00:25] That's test-xi2 output. [00:25] Looks like the raw valuator is fine, but the cooked valuator is stuck at -2147483648.00 [00:26] what's this cooked valuator? [00:26] The thing that GTK listens to. [00:26] hm [00:27] ie: it's the value of the axis after any acceleration, processing, etc. [00:29] i see [00:45] Sarvatt: ping [00:45] what's the gesture that's hooked up to opening the dash? [00:49] hyperair: 4 finger tap [00:51] hmm [00:51] how does 2 finger scrolling end up as a 4 finger tap.. [00:51] and in the first place, how is utouch even working when the utouch patch has been disabled? [00:53] RAOF: is it safe to bring up the 117_gestures.patch yet? [00:53] perhaps that's the key to this [00:55] 117_gestures.patch is obsolete after real multitouch [00:56] i see [00:56] i'm not getting "real multitouch" on my laptop though [00:56] evince and eog no longer respond to pinch and rotate [00:57] could you try disabling coasting and seeing if that helps? [00:57] CoastingSpeed=0 [01:10] * RAOF suspects integer overflow [01:19] Yeah, something's overflowing. [01:23] well that was fun, took 20 minutes to get into a unity 3d session again trying to restart lightdm, so much for staying off the pc on a sick day :) [01:24] Sarvatt: You're sick? Boo :( [01:24] was just getting dumped back to the greeter, trying 0ubuntu5 out now [01:24] odd thing is, i couldn't reproduce it in a guest session [01:30] You've probably got an odd setting. [01:30] Somewhere. [01:31] hyperair: Have you filed a bug? [01:33] i'll invalidate out the bug if i still dont notice it tomorrow [01:34] hyperair: Also, I don't suppose you know whether the same problem occurs with evdev? [01:34] Hm, actually, evdev can't duplicate it because it won't have a scroll valuator. Bah. [01:35] RAOF: you mean regarding the valuator? [01:35] RAOF: i haven't filed a bug, actually. i was looking for a way to reproduce this [01:35] Yeah. [01:35] it randomly happens, but often enough to get annoying [01:35] hyperair: Well, I'm pretty sure you'll be able to reproduce it by just continually 2-finger scrolling down. [01:35] oh it's happened [01:35] O_o [01:41] hmm turns out that it's emitting GDK_SCROLL events with GDK_SCROLL_SMOOTH direction with delta_x=0 and delta_y=0 [01:45] Sarvatt, turns out the 295.40 blob doesn't work on anything less than or equal to the g80, which is what the 8800gts.gtx have. that's a minor snafu [01:48] hyperair: Yeah, that's exactly right. [01:48] And if you check out xinput test-xi2, you'll find that the scroll events all have a valuator of -2147483648.00 [01:49] And it's *stuck* on that value, so delta_x is obviously 0. [01:50] RAOF: no, delta_y [01:50] delta_x is still working [01:50] Bah, that's what I meant. [01:51] heheh [01:52] where's test-xi2 anyway? [01:52] xinput test-xi2 $DEVICEID [01:54] whoops, when i scroll on that window compiz detects it on the root window (i think) and switches the deskto === Sinnerman is now known as Cobalt [15:54] how are fglrx packages versioned? the amd driver on their website has version 12.3, but we are at 2:8.960 [15:54] "weirdly" [15:55] right [15:55] well, that I can see ; [15:55] ;) [15:55] the version we have is what the driver says it is. amd uses some year.release schema [15:57] tjaalton: that's.. confusing [15:57] people go to the amd website, see that their version is at 12.3 and go all like "omg, ubuntu drivers are so out of date - I have to update" and proceed to breaking their system [15:59] tseliot: maybe the fglrx package could mention the "marketing version" on the package description? [15:59] in all the time I've spent on my LoCo channel using upstream nvidia/amd drivers have been the biggest source of support questions [15:59] it's all part of an evil plan to make people stop wanting to install those drivers [16:00] also, because people are sort of smart they know they have to remove all the packages with nvidia in the name to install upstream nvidia driver [16:00] and by doing that they remove nvidia-common which would prevent nvidia installer from installing the driver ;) [16:00] good for them [16:00] :) [16:01] and then they go their merry way telling others how bad Ubuntu is ;) [16:01] which isn't that good ;) [16:01] let's move that to jockey then [16:01] dunno [16:03] I'm pretty sure the battle is completely lost [16:03] a lot of new ubuntu users are so called "power users" from windows [16:03] and power users know exactly enough to destroy their systems ;) [16:03] (especially how to copy and paste random scripts from the web) [16:04] yes [16:04] with 'sudo' on every line [16:09] tjaalton: do you think they would look for the driver version in the package description? [16:09] tseliot: where do they get the package version?-) [16:10] tjaalton: either Jockey or update-manager, I'd say [16:10] doesn't it show the description too [16:10] maybe Jockey does it [16:10] * tseliot can't remember [16:11] update-manager does not [16:12] just leave it as is [16:12] tjaalton: and that would require one more manual step in the packaging [16:12] software-center doesn't show it [16:12] no idea where the text is from [16:14] P.S. marketing versions are crazy :P [17:52] kklimonda: s/power users/tweakers/ ;) [17:55] JanC: potatoe potato ;) [18:09] nah, lots of power users don't have time to tweak all day [18:11] although maybe that would only affect the frequency of breaking their system ;) [18:29] Hi everyone, quick question: I am trying to run intel_gpu_dump on Precise and found that the program is missing. Was it renamed? [18:57] mars: you don't need it anymore. If the gpu hangs, it will dump contents automatically in /sys/kernel/debug/dri/0/i915_error_state [19:18] albert23, that's what I thought, thanks [20:06] bryceh: was there a kernel build with drm-intenl-next-queued somewhere? I only see -next on the mainline ppa [20:06] -n [20:08] tjaalton, yes [20:09] since Intel renames their branch-to-test so frequently we decided to just give it our own naming, which is drm-intel-experimental [20:09] currently that maps to daniel's drm-intel-next-queued [20:09] oh there it is [20:09] had to reload the page :) [20:10] when they switch to drm-intel-proposed-next-testing-queued-really, we can just update the mapping :-) [20:10] yeah I noticed the discussion you had the other day :) [20:19] hopefully if they change it again, we can catch it sooner [20:20] seems this time it changed in january, but I only realized it this past week [20:24] jesse asked poolie to test it to see if it fixes bug 745112, that's why I was asking [20:24] Launchpad bug 745112 in linux (Ubuntu Precise) "[arrandale] desktop is messed up with external monitors (x86_64)" [High,Triaged] https://launchpad.net/bugs/745112 [20:25] * bryceh nods [20:28] tjaalton, hey do you know offhand what package provides the scripts for dpkg-reconfigure keyboard-configuration? [20:30] bryceh: keyboard-configuration? :) [20:30] that's the package name being configured [20:30] duh, ok thanks :-) [20:31] oh, but source package is our old friend console-setup [20:31] yeah, ok [20:33] bingo, found the bug :-) [20:39] is it 985065? [20:54] tjaalton, that's right [22:15] bryceh, as soon as I get a couple more reviewed-bys on my input patches from whot I plan to push a big SRU for xorg-server with them [22:15] it would help to have someone else on the ubuntu side double check the patches just for sanity [22:16] I have them all here: http://cgit.freedesktop.org/~cndougla/xserver/log/?h=input-fixes [22:16] they were all found as I was attempting to fix bug 974887 [22:16] Launchpad bug 974887 in xorg-server (Ubuntu) "unity 3d touch stops registering clicks" [Medium,In progress] https://launchpad.net/bugs/974887 [22:17] so basically, I'm giving you a heads up, and if you have some spare moments and feel like reviewing them ahead of time, please do [22:17] RAOF, you too [22:18] oh, I forgot to mention that I won't be including the last two commits because they would be ABI breaks and I don't think anyone will use them in precise [22:19] cnd, ok, got a few plates in the air at the moment, but will review them when things calm down [22:20] thanks