[00:09] rickspencer's hitting a synaptics crash too - lp #937695 [00:09] Launchpad bug 937695 in xserver-xorg-video-intel "Display Lockup" [High,New] https://launchpad.net/bugs/937695 [00:16] It's not clear that's actually a synaptics crash, though? [00:16] RAOF, [ 8161.471] 8: /usr/lib/xorg/modules/input/synaptics_drv.so (0x7fb8190c3000+0x4d9b) [0x7fb8190c7d9b] [00:17] it's something in pointer event handling code [00:18] Oh, it's in the signal handler, which is why there's intel above it in the stack. [00:18] bryceh: ah.. I think greyback hit the x freeze too [00:26] hrm.. minimize then unminimize in latest unity results in an all white window [00:26] known issue? [00:26] It doesn't here. [00:27] I have an nvidia chipset with twinview.. [00:27] ah [00:27] its only doing it in terminal [00:28] and terminator [00:28] RAOF, any progress on the recent nvidia issue? [00:29] not doing it here either, just checked with gnome-terminal, please file a bug against xorg and we'll reassign to nvidia or compiz as appropriate [00:30] DBO: Yup - it's mostly fixed, but the barrier behaviour on the screen edge is still slightly different. I'm hunting down wh. [00:30] Sarvatt: will do [00:30] DBO: nvidia issue? are you on the nvidia calls? it might make sense for you to join them if not, usually we cant fix actual bugs with the drivers [00:30] oh barrier problem yeah thats RAOF [00:31] RAOF, disabling it all together makes it act identical here [00:31] Also since rebooting into the latest, the dock isn't auto-hiding.. that seems weird [00:31] RAOF, can we just disable it [00:32] Sarvatt, Im not interested in getting in touch with nvidia, I'll let the business people do that [00:32] DBO: We can, but it's entirely possible that other things will break this in the same way. [00:32] DBO: So we might as well fix it properly now. [00:33] (For sufficiently ugly versions of “properly”, due to ABI stability requirements) [00:33] RAOF, so long as a proper fix hits distro sooner rather than later [00:33] currently we have every single nvidia user unable to reveal their launcher [00:33] DBO: If it's not today we can futz with the xorg.conf settings. [00:33] RAOF, thanks, I know it sucks but it needs to happen that way [00:35] RAOF: alberto can change jockey to apply xorg.conf options to nvidia globally if that helps [00:36] theres also disabling the barrier stuff globally if making binary drivers we cant change work is a priority [00:38] Well, the problem is in the barrier implementation. The binary drivers are well within their rights to do what they're doing. [00:41] RAOF, the driver is allowed to constrain cursor motion? [00:41] doesn't that seem like a broken abstraction to you? [00:41] Well, yes. But it's not a terribly unreasonable thing for the driver to want to do. [00:43] RAOF, okay, well, whatevs :) [00:43] Because (in this case) the driver is the only thing which knows where the CRTC boundaries are, and clamping the cursor to the CRTC boundaries is a perfectly reasonable thing to want to do :) [00:43] Ah. Of course. My old friend, floating point rounding. [00:44] Further info on the terminal thing.. [00:44] it only happens if gnome-terminal or terminator are *maximized* [00:46] Is there any way to turn off this *weird* thing where the mouse pointer gets stuck in the middle of the screen on dual-monitor btw? [00:46] RAOF: more barrier issues ^^ :) [00:46] same thing? [00:48] SpamapS: No, different thing. In what way stuck? [00:48] DBO, is the launcher problem new to the 295.20 driver? [00:48] Well, similar thing. [00:48] bjsnider: Yup. [00:49] so unity and gnome-shell are both broken by that driver [00:49] i thought it was only gnome-shell [00:50] In what way is gnome-shell broken? [00:50] amazing that so many problems happened to a driver they marked as long-lived support [00:50] RAOF, last I heard its a weee bit crashy [00:51] bjsnider, you can tweak your xorg.conf to fix the issue [00:51] if you want [00:51] gnome-shell crashes when trying to type in the search box in overview [00:51] bjsnider, http://ubuntuforums.org/showthread.php?t=1928096&page=2 I posted a fix in there [00:51] RAOF: as I'm moving the mouse from right screen to left screen, the mouse stops briefly at the left edge of the right screen. [00:52] RAOF: it does not happen if the mouse is moving fast [00:52] SpamapS, thats normal [00:52] SpamapS: Ok, that's by designish. [00:52] its *horrible* [00:52] can I turn it off or at least configure the delay to be shorter? [00:53] SpamapS, there is a responsiveness slider in the appearance preferences for the launcher reveal [00:53] tweaking the slider might give you results you prefer [00:53] Its like somebody just came and stood right in the middle of desk between my monitors and started humming Old MacDonald. Very distracting. [00:54] SpamapS, its so you can reveal the launcher on the left edge of the right hand screen [00:54] if you have launchers not-hiding [00:54] its so you can half-maximize windows [00:55] It needs some tuning. :) [00:55] That is true. [00:55] so [00:55] being able to reveal the launcher of the left edge of the left screen would be nice [00:56] DBO: thanks the slider helped a tiny bit, at least the launcher does not pop up immediately, but it still stops the mouse from moving. [00:56] SpamapS, where did you move it to [00:56] lifeless: nvidia? I'm fixing that right now. [00:56] out of curiosity [00:56] DBO: lowest sensitivity [00:57] Actually [00:57] RAOF: yes [00:57] no.. its still annoying no matter where I slide it to [00:58] SpamapS, its always going to stop your mouse [00:58] yeah, thats my main gripe [00:58] the further right you move it, the more it will respond to fast movement [00:58] it doesn't stop mine [00:58] I'd like it to not ever stop my mouse. [00:58] unless I'm moving it slowly [00:58] I'll just get used to fast moving I guess [00:59] SpamapS, there is a way you can turn it off, but only in ccsm [00:59] I think I'd be happier if it simply required slower movements. [00:59] and I cant ensure someone wont remove that feature or make me remove it in the future [00:59] DBO: I swore an oath that I would never kill again... the day I apt-get purged ccsm. ;) [00:59] yeah [00:59] I was hoping you would say that :) [01:00] SpamapS, if you present your case clearly, well articulated, and without any form of vitriol in a bug report [01:00] Really, I just think its requiring slightly too fast a movement to work in the intended way. [01:00] you might get listened to [01:00] wow, now thats a confident statement :) [01:00] SpamapS: What's your input device? [01:00] I can't speak for others lifeless :) [01:00] DBO: vitriol is the liquid schwartz of the internet... [01:01] RAOF: an optical mouse [01:01] SpamapS, yeah but vitriol is a quick way into the bin with your bug report [01:01] Ah. One of those ones which very conveniently claims it resolution is 1/meter [01:01] I ignore all bugs, no matter how critical, reported by cocks [01:02] DBO: Hm. How much do we care about receiving events with very low velocity on the barrier? [01:02] RAOF, define very low [01:03] * SpamapS runs a few seds on his bug report .. s/a....le/kind sir/;s/f...ing/fine/; [01:03] DBO: Small enough that the pointer would move less than a pixel if it were allowed to. [01:03] RAOF, yeah we dont care about those [01:04] the velocity calculation would always be minimal right? [01:04] Right. [01:04] If I move with ninja-like precision it works.. but my usual lazy wandering mouse pointer is trapped in its current screen.. [01:04] yeah those have a negligible effect on the end result [01:04] Good, because we don't generate them on the right hand barrier because we trunc() the thingy. [01:05] SpamapS, is your mouse sensativity very low [01:05] RAOF, you mean left hand barrier? [01:05] I mean the right hand barrier. [01:05] let me look [01:05] As in: the barrier on the left hand of the right screen. [01:05] DBO: as low as it can go [01:06] and acceleration is all the way Low too [01:06] RAOF, oh, okay [01:07] SpamapS, thats probably why its so hard for you [01:07] DBO: that and the numerous head injuries ;) [01:07] * DBO is glad you took that innuendo that way [01:08] DBO: putting it in the middle (which now makes using the mouse very annoying in general) doesn't really make it any better though.. [01:08] ah yeah actually mine seem to be all the way down too [01:08] I retract my original statement [01:09] well I'm already getting used to sort of "chopping" to get to the other side of the screen.. [01:10] its not exactly intuitive, but it is a quick muscle memory change [01:10] you should be able to use a even smooth slow push [01:10] it should let you through [01:11] DBO: right but I don't even want to get stuck there so I just accelerate over it [01:13] DBO: I think perhaps if the pause were smaller, or the speed required to skip the pause just a little bit slower.. I would love the new feature [01:14] SpamapS, there is a bit of a tradeoff to be had [01:14] so you have 2 monitors yes [01:14] which one do you consider the "primary" monitor? [01:14] left [01:14] which I actually look at and use a lot less than the secondary. ;) [01:14] I use the right hand monitor as my main monitor [01:15] when I say primary I dont mean whatever X considers the primary [01:15] whatever *you* do [01:15] My left monitor is actually my laptop screen [01:15] and nvidia is very unhappy with me if I change the primary [01:15] okay so we use the same setup actually [01:15] having a heavier barrier allows me to much more easily access the hidden launcher on the right hand monitor [01:15] the fact that you dont use hidden launchers probably has a lot to do with the difference [01:15] I do use hidden [01:16] actually on upgrade they went to fixed [01:16] you can turn hidden back on [01:16] Yeah I did that [01:17] DBO: true, I may find that after a day or two, I'm not even traversing screens anymore because thats the main reason I click. Not sure. [01:18] SpamapS, or you may find RSI ;-) [01:19] right now I'm finding that with default settings restored.. its not showing me the launcher... [01:19] ahh if I approach slowly, no show [01:21] which I like actually [01:22] Yeah. The barrier stuff *does* make the autohidden launcher much more usable. [01:23] Still ... when I *do* want to go from screen to screen... I have to drag *a lot* [01:24] SpamapS, its about an inch of mouse movement here [01:24] more like 6 for me [01:24] O_O [01:24] just was measuring it actually [01:24] really? [01:24] SpamapS: get a higher res mouse ;) [01:24] Its almost the entire width of my ubuntu mousepad [01:25] I admit, this *is* a crappy mouse [01:25] but that has never mattered to me. :) [01:25] I wish mice gave us useful information about their capabilities in this regard. [01:25] it's fun with a trackpad and a projector attached [01:26] if I use my trackpad.. which is *really* big (macbook pro 5,1) .. I have to start on the very edge of it, and it just flips on the other side [01:26] trackpads from apple currently have a known issue with this :/ [01:26] it still shouldn't be that hard [01:27] try moving the mouse a little faster? [01:27] RAOF: help ... my -intel is crashing like crazy :/ didnt upgrade today (2 days ago). [01:27] *Trackpads* we can do something about; they generally give actual information about their stuff. [01:27] DBO: yeah, I bumped up my sensitivity a tiny bit, and move a little faster, it seems to work better. [01:27] RAOF: i was told that precise is production quality... now i am unhappy :) lol [01:27] asac: Do you have an Xorg.0.log.old available for debugging? [01:27] RAOF: this happens with unity 2d and unit 3d [01:28] RAOF: i have filed a bug [01:28] one sec [01:28] RAOF: not sure if its always the same, seems that apport was a bit buggy and failed to submit 5 times... so this one i got in: bug 937762 [01:28] DBO: seems like there's a time component here. If somebody is moving in the same direction for 3 seconds... perhaps they don't mean to stop. :) [01:28] RAOF: opened it up now [01:29] SpamapS, predicting user intention is difficult [01:29] yep, another input crash [01:29] SpamapS, predictable behavior is WAY better than smart but unpredictable behavior [01:29] input crash? yes, i think it gives me a correlated feeling to typing [01:29] asac: Updating now should stop that. [01:29] byt i am typing so fast :) [01:29] so i cant tell ... because i am always typing so chances for typing while crashing are high :) [01:30] RAOF: cool. very bood [01:30] i will check... see i get something xorg'ish right now [01:30] input synaptics i got [01:30] xserver-xorg-input-synaptics 1.5.99~git20120220-0ubuntu3 has the fix. [01:30] RAOF: guess relogin? [01:30] great [01:30] RAOF, I need these events for barriers to eventually send down mouse button state [01:30] let me relog in again [01:30] cu in a bit [01:30] DBO: Really? [01:31] mmmmm [01:31] no [01:31] but I'd like it [01:31] otherwise I have to query the server [01:31] bryceh: RAOF: great. i am back :) ... so lets see [01:31] * asac goes on a typing rampage [01:31] oh ... [01:32] asac: The trigger would be the transition between typing and touching the touchpad. [01:32] seb128: ... why did you remove alt-F10 keyboard shortcut to maximize my gnome-terminal in unity? [01:32] RAOF: interesting :) ... guess i didnt see it before then as i just reenabled it two days ago :) [01:32] e.g. the touchpad [01:32] DBO: Please query the server :) [01:32] i cant remember why i enabled it ... ah i know... someone claimed it might have automatic multitouch here :) [01:32] asac, it's not me but there is a bug open about it I think, didrocks might know better, ping him tomorrow [01:33] seb128: ouch thought you were sleepig [01:33] asac, lol, if IRC is running I'm not sleeping :p [01:33] I don't let IRC running during night [01:33] seb128: now i remember [01:33] :) [01:34] ;-) [01:34] but I'm about to go, it's late here! [01:35] DBO: I obviously *could* send you button state in the events, but this protocol wants to be generically useful; adding random other bits of state that you happen to want is probably out-of-scope :) [01:35] seb128: i know. sleep well [01:35] thanks ;-) [01:36] RAOF, understood [01:40] DBO: So unless it's going to be a lot of overhead, you need atomic updates, or it's to do with pointer motion, I'll be pushing back on adding additional data to the events :) [01:41] RAOF, hey so long as you know Im always going to ask for whatever seems to be the most convenient thing [01:41] we are in a good state [01:41] you protect the protocol, I will advocate Unity's interests [01:43] This sounds fine :) [01:44] You know what's not a great use of space? Shown-by-default launcher on this netbook. [01:47] DBO: Did you write some XTest-using tests for the barrier stuff? [01:47] RAOF, yes [01:47] Where are they? [01:47] uhm we dont know if they work on all machines really [01:47] consistency is still a problem [01:48] but they are in the autopilot test suite [01:48] Where's the source? Ideally I'd get a bunch of them run as a part of the xserver build. [01:49] RAOF, well you would be depping on unity to use my tests [01:49] all we are doing is using relative motion events with xtest to cause the launcher to reveal [01:49] and testing that it is in fact revealed [01:51] I was rather more thinking of stealing your code and wrapping it up in a shiny new xorg-gtest wrapper. [02:03] RAOF, its written in python [02:04] Curses. Oh, well. [02:07] DBO: What do you need for a bamf bug where it consistently fails to believe that chromium is, in fact, on *this* workspace? [02:07] Until I switch workspace away and back again. [02:07] 2 things [02:07] first if you are experiencing the bug now [02:07] lets skype [02:07] second [02:08] if you have a way to cause the bug [02:08] let me know [02:08] waaaait [02:09] Im experiencing it right now [02:09] hold on [02:16] RAOF, I think I found the bug [02:18] Do you need some testing? [02:18] un-momento [02:19] RAOF, https://code.launchpad.net/~jassmith/unity/unity.fix-monitor-tracking/+merge/94078 [02:22] * RAOF pulls, builds, and installs. [02:35] RAOF, so just to be sure, by the time I wake up tomorrow, one way or another this nvidia reveal issue will be a thing of the past [02:35] DBO: Yes. [02:35] In fact, it should be in a couple of hours. [02:36] Oh, arse. [02:36] DBO: How much do you want testing of that branch? It would appear to need a newer nux. [02:36] DBO: err, whoops, accidentally globally approved your branch [02:36] RAOF, im relatively certain it works [02:37] I un approved, it, hope that the merge bot didn't see that :/ [02:37] we're good Im sure [03:08] DBO: There you go. One nvidia fix. [03:13] RAOF: bah you fixed your already obsolete patch? :P [03:13] Yes :) [03:14] push to git plz! [03:15] need to refresh it for xserver master [03:15] not tonight though, sorry to hound you minutes after you uploaded [03:17] hurm, -synaptics is still on the ubuntu+1 branch [03:18] bryceh: And cnd hasn't pushed his latest changes, either. [03:18] Sarvatt: Done! [03:33] so... [03:34] why can't I install skype on my laptop? [03:34] I have the partner repo checked [03:34] running 64 bit [03:35] Is skype actually in the Precise parter repo? [03:36] No, it isn't. :) [03:38] thumper: sudoedit /etc/apt/sources.list; change the partner distro from precise to oneiric. Tada! skype. [03:42] RAOF: really? [03:42] ok [03:42] Really. [03:43] Partner's kinda weird. I'm not sure if there's anything in there for precise yet. [03:43] flash is there [03:43] * thumper does it [03:48] micahg: Is it? It's not flashplugin-nonfree in multiverse? Which, true, downloads from partner or something, but the package itself is in the main archive. [03:48] RAOF: adobe-flashplugin [03:48] Ho, hey! Wow! [03:49] We have too many ways to install flash ;) [03:49] pull flash out of the installer and we'll happily drop it from multiverse ;) [03:52] the package in multiverse slurps down a hard-coded url on archive.canonical.c, right? [03:52] (i just got bitten by that - i have a deliberately frozen apt mirror, and the version of flash it's trying to fetch isn't on archive.c.c anymore) [03:52] RAOF: thanks, installing skype again [03:56] broder: well, it pulls the .orig.tar.gz which changes "location" with each version and verifies the sum on it [04:01] You know, gnome-online-accounts would be more useful if any of its features worked. [04:39] RAOF, is it possible this update has changed the velocity values? [04:39] things feel a little more twitchy than before [04:41] Good morning [04:42] morning pitti [04:42] Morning pitti. [04:55] DBO: It will have changed velocity values, but only by a small amount; < 1 px/dt. === m_conley is now known as m_conley_away [04:56] RAOF, probably me being overly sensitive then [04:57] I did check that the values coming out were comparable to before (except on the edges on nvidia, obviously). [04:57] And, actually, it should only change the values on the left-most edge. [04:58] Since dt ~10^-3, that means velocities may be different by up to about a hundred, mostly less. [05:01] Sweetshark: not uploading the new LibO; armel and armhf failed on "debug output level 11 is too high", too [05:02] argh, someone binNEWed the new LibO on i386/amd64 [05:02] oh well [05:18] is it deliberate that the time remaining on battery isn't shown anymore ? [05:18] (and that my battery just shows 'charged' after 2 hours on it ? [05:18] hm, it worked fine for me over the weekend [05:27] pitti: hi, does LibO build successfully on armhf, modulo the most recent FTBFS you mention? [05:28] achiang: only thing I can say is that it failed with the same error on armel and powerpc, so that "debug output level 11" breakage is not armhf specific [05:28] achiang: it does have janimo's armhf patches, so it ought to build otherwise [05:28] pitti: ok, thanks. [05:40] the new color chooser looks nice, there's some screenshots at http://iloveubuntu.net/new-redesigned-color-chooser-dialog-landed-precise-pangolin [05:42] hey jbicha [06:07] wow I never realized how much I move back and forth between screens on my multi-monitor setup until this new reveal pause thing started making it awkward [06:16] RAOF, any word on the "press against screen edge to bypass barrier" issue we chatted about a week or so ago? [06:16] * SpamapS installs new xorg to get back the launcher on the left side [06:16] DBO: Which one? The one where some people have barrier ids that increase? [06:17] no [06:17] the one where if you push on the top edge of the screen [06:17] Oh, that one. [06:17] you can move past the barrier without it stopping you [06:17] I thought that you didn't care, because you're not defining a full-height barrier anyway. [06:18] (And, yes. I tried to get my head around the algorithm that's used to calculate barrier hitting, and it's madness. It's the perfect example of something that should have copious comments) [06:18] RAOF, no we care now because the barrier has to go to the panel [06:18] for the case of corner reveal [06:19] Ok. I'll bump up the priority of fixing it then. [06:20] DBO: I know I asked before, but there's no way to turn off the thing that stops the mouse in the middle right? 3 hours of trying to build new habits and I'm about ready to set my mouse on fire. :-/ [06:21] SpamapS, there are some minor changes in the latest update [06:21] xorg update [06:21] pull them down [06:21] and see if its still driving you up a wall [06:21] 2:1.11.4-0ubuntu4 [06:21] just installed [06:21] will try it out [06:21] did you log out and back in? [06:21] no I literally just installed it 60 seconds ago :) [06:21] :) [06:28] good morning [06:28] morning didrocks [06:28] bonjour didrocks, ca va? [06:29] hey DBO, guten morgen pitti, I'm fine thanks (and got the repaired heater after a month in the store \o/)! Wie geht's? [06:29] didrocks: wow, just after the days of -15 degrees are over ;) [06:29] didrocks: gut, danke! [06:30] pitti: yeah, I know, this is really the bad timing :) [06:35] morning didrocks :) [06:40] morning didrocks and pitti ! [06:40] hey jasoncwarner_ [06:40] jasoncwarner_: I thought you were on holidays? [06:40] pitti: next week, this week i'm trying to get through all my admin, taleo and other stuff...need to poke Laney about his application ;) [06:41] hey smspillaz, jasoncwarner_ [06:45] didrocks: is it possible somehow to get back intellihide? [06:45] (gsettings, etc.) [06:46] DBO: for the record, the new X did re-enable my left-side-revealing for the main display. But the overall experience still inspires musicide (the act of killing a mouse) [06:46] pitti: no, the code has been completely removed [06:46] pitti: executive decision [06:46] ah, too bad; thanks [06:47] DBO: what package should I report a bug in? [06:47] SpamapS, unity [06:47] pitti, prepare for lots more options to die [06:48] DBO: ok. Actually it does seem like the mouse moves a little sooner between the two screens. [06:48] basically every option thats not in the gcc applet [06:48] is going to die this week [06:48] DBO: especially going left -> right [06:48] DBO: still too bad; intellihide was the one that by far made most sense :( [06:49] as long as nobody kills FFM :) [06:49] smspillaz: any luck with my https://bugs.launchpad.net/ubuntu/+source/compiz/+bug/936818 ? [06:49] Launchpad bug 936818 in compiz "Visual Corruption on Desktop" [High,New] [06:50] DBO: btw, did you get complaints about the new "push harder" behaviour to reveal the launcher? with a touchpad it's really hard work to make it appear [06:50] didrocks: Have you been able to play with barrier-test? [06:50] pitti, FFM? [06:50] DBO: focus follows mouse (not an unity setting, it's a compiz one) [06:50] pitti, touchpad is going to improve [06:50] pitti, ah yeah you can keep that [06:50] DBO: ah, nice [06:50] RAOF: unfortunatly, I only work 12 hours a day, I hope to have time this week though :) [06:51] didrocks: :) [06:51] RAOF: I need first to see why I can't upgrade my nvidia [06:51] * RAOF starts preparing dinner; back later. [06:51] as I think I can't rebuild the previous working package for ever :) [06:51] jasoncwarner_: I thought you said you still got that on unity-2d and gnome-shell ? [06:52] two different bugs [06:52] smspillaz: one was before desktop [06:52] smspillaz: one was on desktop [06:52] smspillaz: this is on the desktop (see screenshots) [06:53] jasoncwarner_: what happens if you disable unity ? [06:54] smspillaz: don't know, I'll test that in a bit (just finishing up some work) [06:55] ok [06:58] smspillaz: I'm trying out classic with effects so I'll let you know...it takes some time for it to appear [07:00] OMG classic feels soooooo old after Unity...like, seriously...where is dash? where is my launcher? [07:00] jasoncwarner_: it sounds like you have an overheating gpu [07:01] jasoncwarner_: my old gpu used to overhead a lot and it wouldn't be uncommon for me to get artifacts like that one [07:01] *overheat [07:01] jasoncwarner_: I could be wrong though, it could just be a bug in nux [07:01] or worse, compiz [07:01]