=== med is now known as med_out [01:27] apw: hi; I'm curious if you think bug 729338 will have a fix before oneiric is released, or should we look for alternative workarounds ? [01:27] Launchpad bug 729338 in linux "yama hardlink restriction misbehaves under aufs" [Medium,Triaged] https://launchpad.net/bugs/729338 === smb` is now known as smb [07:30] * apw yawns [07:30] * smb waves (a coffee) [07:31] * apw reposts with a cup of english breakfast [07:44] Capitol! [07:49] A [07:53] B [07:54] No, capital is where the government lives. Capitol is the accepted response for english breakfast tea. [07:55] * smb was hoping to score 7 for doing a cab... :-P [07:57] RAOF, i have have heard the sound of "capital" used in the context of "capital idea" i didn't think it spelt differently though [07:58] My version is! [08:00] heh, well i assumed a miss-spelling and worked with that. but indeed it is a capit[oa]l idea, so much so i need another [08:04] RAOF: just to confuse some more, Capitol Hill is where the US government lives ;) [08:06] tjaalton, ahh but that really is a differnt spelling (with an o) [08:08] apw: yeah, just a reply to the "capital is where.." [08:09] * smb thought it was capital hill... well until they nearly went out of it now... :-P [08:09] hey RAOF, tjaalton, smb, and apw. [08:09] * apw waves to jk-- ----- [08:09] hi jk-- [08:09] jk--: o/ [08:09] * jk-- makes use of the Oxford Comma while it still exists === jk-- is now known as jk- [08:10] jk++ ? ;) [08:10] :D [08:11] smb: too similar to c++ [08:11] jk-, Oh, you did not accidentally type code when setting the nick then... ;-) [08:13] /nick intmain(void) [08:13] /nick intmain(void){return0;} [08:13] :) [08:14] * apw calls you initmain from now on [08:14] intmain even [08:14] only if I can call you EXIT_FAILURE [08:15] or maybe EXIT_SUCCESS, less bad connotations [08:17] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/821157 [08:17] Ubuntu bug 821157 in linux "intel 5300agn in Oneric leads w722v to reboot" [Undecided,New] [08:17] if i am reading this right oneiric causes their wireless router to reboot ... [08:18] nice [08:19] apw: how odd [08:20] indeed. and i am unsure how i am meant to help with it ... your router is clearly broken if a client can crash it [08:21] yeah [08:21] I mean like what, if an ATM went out of order when you put your card in, there is something wrong with the atm not your card [08:22] when my car drove over the bridge it collapsed, your car is clearly faulty please fix it [08:22] heh [08:23] "but it worked with my previous car!" [08:24] some people have a hard time with cause and effect [08:32] jk-, on the nail [09:35] got anymore info [10:04] Natty writes at startup in a virtual console: "[1.581365] mmc0: no vmc regulator found". I suppose that this message originates from the MMC kernel subsystem. What kernel routine or program does cause this message? [10:26] bullgard6, what h/w is that from [10:27] bullgard6, and is that verbatim copy? ie could it be vmmc in the middle ? [12:15] apw: The hardware is a Thinkpad T61. Yes, I made a mistake. It should read "vmmc". I beg your pardon. [12:18] apw: In the mean time I found out that /usr/src/linux-source-2.6.38/drivers/mmc/host/sdhci.c includes a printk line of this text. === _LibertyZero is now known as LibertyZero [12:42] bullgard6, i am trying to understand if there is an issue associated with this, or if you just are querying the error [13:08] apw: Once I tested this computer using an SD card, and it worked. I would like to know what this warning means to me. [13:16] bullgard6, from what i am reading your mmc slot can have an option power regulator, if that is found we enable it else we emit that message === med_out is now known as medberry [14:20] * ogasawara back in 20 === medberry is now known as med_out [14:58] cnd, I have a question for you about input multitouch protocol when you have a minute [15:01] sforshee, sure [15:03] cnd, I'm working with a touchpad driver, 2-finger touch seems to be working fine but 3-finger events aren't [15:03] comparing the output from this driver to that from synaptics, the only thing I see different is that this driver only reports 1 MT position whereas synaptics reports 2 [15:03] which touchpad? [15:03] elantech [15:04] the tripletap event is reported, but the device only gives a single position for 3 touches [15:05] cnd, in ubuntu a 3-finger tap just acts like a 1-finger tap [15:05] and a 3-finger drag like a 1-finger drag [15:06] sforshee, this is typical for multifinger touchpads [15:06] are you sure it's actually multitouch?> [15:07] yeah multi-finger is not the same [15:07] cnd, it reports the number of contacts (up to 3) and with 2 fingers it reports 2 positions [15:07] sforshee, oh, and then it drops down to one position for three touches? [15:08] yep [15:09] cnd, would it help to see a dump of the events? [15:09] hmm [15:09] it wouldn't hurt [15:10] so what questions do you have? [15:10] cnd, http://pastebin.ubuntu.com/659348/ [15:11] cnd, I guess my question is what events the driver needs to spit out to have 3 contact gestures properly supported [15:11] ok [15:11] sforshee, do you know that it can even do three touch [15:11] cnd, that dump is for dragging on the touchpad with 3 fingers [15:11] just cause it can tell you that there are three fingers touching doesn't mean i knows where they are [15:11] apw, it reports that there are 3 contacts, but only reports one position [15:11] right [15:11] sforshee, can you provide me the full dump that includes the transitions from 1 to 2 to 3 touches [15:11] that as they say tells you nothing [15:12] but isn't that enough for a 3-finger tap or drag? [15:12] sforshee, multifinger mode, which this looks to be, would be enough [15:12] now that is a different thing, and likely it should be [15:12] except we don't actually have the support for it [15:12] cnd, let me get that dump for you [15:12] c [15:13] we gave it a shot about a year ago, and it broke devices [15:13] cnd we gett [15:13] cnd we getting anything new for this cycle? [15:13] and we never tried again [15:13] apw, kernel patches or? [15:13] functionality in touch [15:13] from a user perspective [15:13] yes [15:13] so we're still working on eog and evince to get smooth scrolling and rotate and pinch to zoom [15:14] cnd, with a synaptics touchpad I'm getting 3-finger taps == middle click and 3-finger drag moves the window around [15:14] hopefully those land before FF next week [15:14] we just published utouch-qml to universe [15:14] that will make writing new qml applications with touch really easy [15:14] sforshee, yeah, that's with sem-multitouch [15:14] which is a step up from multifinger [15:15] apw, that's also something new this cycle [15:15] semi-multitouch synaptics trackpads have some gesture support [15:16] cnd, http://pastebin.ubuntu.com/659355/ [15:16] let me know if that isn't what you meant [15:18] sforshee, in this event stream, did you go from 0->1->2->3 or from 0->1->0->2->0->3? [15:18] sforshee, is that ... heh waht he said [15:18] cnd, the latter [15:18] I guess you want the former? [15:18] yep :) [15:18] okay, just a sec [15:19] but this does help too :) [15:19] cnd, http://pastebin.ubuntu.com/659356/ [15:19] now you should have plenty of data :) [15:20] indeed :) [15:21] sforshee, is this really 0->1->2->3? [15:21] cause the data is still the other way [15:21] sforshee, its the exact same data including timestamps [15:21] rats, let me try again... [15:22] heh [15:23] cnd, apw: http://pastebin.ubuntu.com/659357/ [15:23] I just got the wrong file [15:23] ok [15:25] sforshee, is this the upstream elantech driver, or from somewhere else? [15:25] sforshee: Hey dude, you know the fix that you put into the proposed kernel that made the atheros driver work, the latest natty kernel seems to of removed it again I think [15:25] cnd, it's upstream plus some work I've been doing to support a new model [15:26] sforshee, does the hardware cease to give two points when you put three fingers down? [15:26] davmor2, do you have a bug number that I can refer too? [15:26] cnd, yes [15:26] or is it something in the driver that is causing that to happen [15:26] hmmm [15:26] sforshee: just tracking it down [15:27] cnd, the v2 support in the upstream driver is pretty similar to what I have, the difference largely being how 2-touch coordinates are reported [15:27] by the hardware that is [15:28] sforshee, if the hardware ceases to give two coordinates when there are three touches, then I suppose this is the best you can do [15:28] I don't know what our stack would do in this case, but I'm guessing it won't see three finger gestures [15:29] sforshee: https://bugs.launchpad.net/bugs/710738 [15:29] Ubuntu bug 710738 in linux "Regression latest kernel breaks my Atheros AR5001 wifi" [High,Fix released] [15:29] you can run geistest (in utouch-geis-tools) to see what gestures are emitted [15:29] cnd, okay, that's what I wanted to know -- whether I needed to do something different in the driver [15:29] sforshee, if it's something you are interested in, you could also add support to our stack for it :) [15:29] cnd, thanks -- that answers another question I have, what tools to use to test multitouch :) [15:29] there are multiple tools, but that's the easiest to use and read [15:30] cnd, thanks a lot! [15:30] sure, np [15:30] sforshee: nevermind I've just fired it up again and now it's working so meh [15:32] ok, time to actually get up for the day [15:32] davmor2, I just checked and the patches are definitely still in natty [15:33] hopefully it keeps working for you :-/ [15:33] sforshee: yeah I think there might just of been a glitch after the initial install, seems fine now :) === kentb is now known as kentb-early-lunc === med_out is now known as medberry [16:17] * apw takes a risk and updates one of his machines [16:17] if you don't see me again you know why === medberry is now known as med_eatz [17:52] ogasawara, apw: sorry about the ipv6 patch, it was done Friday at the sprint and somehow I dropped it. /me swears it was sent then :/ [17:53] jjohansen: no worries. I had it on my todo list to review, so thanks for doing the work! === kentb-early-lunc is now known as kentb [17:53] jjohansen: I'll get applied for the upload I'm preparing [17:54] ogasawara: I am going to drop an apparmor patch next week, would you prefer revert patches for current stuff, or do you plan to just rebase? [17:55] jjohansen: probably easier for me if you revert patches for current stuff [17:56] ogasawara: okay, then that is what I will do [18:14] pgraner, do you have root on tangerine? can you install unzip ? === bjf is now known as bjf[afk] [18:42] anyone around that can approve my nominations on bug 583760 ? [18:42] Launchpad bug 583760 in gentoo "[PATCH] Mouse cursor dissappears with nouveau" [High,Confirmed] https://launchpad.net/bugs/583760 [18:43] sforshee: done [18:43] ogasawara, thanks! === med_eatz is now known as med_afk === med_afk is now known as med_out [19:52] * ogasawara lunch === bjf[afk] is now known as bjf [20:10] manjo, nope I sure don;t [20:11] cool np [20:17] sforshee, I remembered one other thing that you need to be sure to do with the elantech driver [20:17] be sure to set the maximum slot value at 1 so that it matches what the device gives, and set the INPUT_PROP_SEMI_MT property for the device [20:19] cnd, thanks for the tips [20:19] testing with geistest didn't go so well [20:21] no? [20:21] most gestures were just detected as drags [20:23] cnd, the driver already does both of those [20:34] sforshee, ahh, great :) [20:34] sforshee, what did you expect for gestures? [20:34] you should only be able to get two touch drags and pinchs [20:35] cnd, oh, really? then it didn't do so bad [20:35] cnd, I saw this page and tried to run through those tests: http://testcases.qa.ubuntu.com/Hardware/Touchscreen/Multitouch [20:37] sforshee, SEMI_MT alone limits you so you can't do rotate [20:37] cnd, that explains it then :) [20:37] SEMI_MT means you have a bounding box, but you don't actually know where the touches are [20:38] the actual touches could be at (x1, y1) (x2, y2) or (x1, y2) (x2, y1) [20:38] and if you don't know where they are, then you can't do rotation [20:38] but you can do pinch by watching if the bounding box's area is changing [20:39] cnd, that makes sense, and seems entirely appropriate for this device [20:39] sforshee, can you get three touch drags? [20:39] you may not see them in geistest [20:39] because unity may be grabbing them [20:39] but if you run geistest in gnome 2 then you could get them [20:40] cnd, no [20:40] ok [20:40] it's behaving like a one touch drag [20:41] there's some logic in utouch-frame that emits the number of touches based on BTN_TOOL_*TAP events for SEMI_MT devices [20:41] if you are interested, you could poke at it to generate the correct number of touches for these devices [20:41] and then three finger drags may work [20:41] cnd, it's sending the tripletap event [20:41] but only one touch point [20:41] yeah, that would be enough for drags, but not pinches of course [20:42] there's probably a bit of logic in utouch-frame that needs to be tweaked [20:42] cnd, in the pastebin from earlier you can see the events: http://pastebin.ubuntu.com/659357/ [20:42] but no triple touch drag [20:47] sforshee, the driver is working properly [20:47] it's a matter of "fixing" utouch-frame to make it emit three touches when the driver emits BTN_TOOL_TRIPLETAP [20:47] utouch-frame does this already for synaptics, but synaptics emits 2 slots from the driver when there are more than 2 fingers [20:48] elantech emits 1 slot from the driver when there are more than 2 fingers [20:48] cnd, okay, I get you now [20:48] something about that is throwing off utouch-frame [20:48] you don't have to look into utouch-frame if you're busy, I'm not trying to push work on you :) [20:48] but if it interests you, that's where you would look [20:49] cnd, I probably won't have time to anytime soon, but I might take a look later on [20:49] ok [20:51] sforshee, I'm just happy to have one more person who speaks the language of the input subsystem :) [20:51] cnd, I've got more touchpad work queued up for next week too :) [20:52] you better watch out though, you could get slurped into the multitouch team :) [20:52] * sforshee shudders [20:52] that's an appropriate response [20:52] :) [20:52] cnd, i know where you live [20:53] bjf, I need to be careful too, cause my only get-away vehicle is a bike... === yofel_ is now known as yofel === med_out is now known as medberry