| === zhxt_afk is now known as zhxt | ||
| === xiinotulp is now known as plutoniix | ||
| === chihchun is now known as chihchun_afk | ||
| === benonsoftware is now known as bdunn | ||
| === chihchun_afk is now known as chihchun | ||
| === bdunn is now known as benonsoftware | ||
| === benonsoftware is now known as MerryChristmas | ||
| === MerryChristmas is now known as benonsoftware | ||
| Mathisen | hello is it worth installing ubuntu on a samsung s4 ? i saw it should be suported.. been using cuyagenmod so far | 07:38 |
|---|---|---|
| Mathisen | is it working okej on that device ? | 07:39 |
| jgdx | mariogrip, hey, where can I buy an one+ with ubuntu? | 07:40 |
| Stanley00 | Mathisen: you can check this site: https://wiki.ubuntu.com/Touch/Devices/i9505 | 07:41 |
| jgdx | mariogrip, (re: https://forums.oneplus.net/threads/wip-ubuntu-touch-for-oneplus-one.266170/page-80#post-13192766 ) | 07:41 |
| Mathisen | Stanley00, thx sound camera does not seem to great so i think i pass on this :) ... maybe some other day | 07:42 |
| Stanley00 | Mathisen: that wiki updated on Dec 2013, on the link to xda, the last update time is on 2014-03-24, so you should check the xda link too :3 | 07:43 |
| === dlan_ is now known as dlan | ||
| JamesTait | Good morning all; happy Monday, and happy Miniature Golf Day! 😃 | 09:09 |
| === vrruiz_ is now known as rvr | ||
| CaptainHeavy | Hello :) I've just asked the following question regarding the Messaging app for Ubuntu Touch: https://answers.launchpad.net/ubuntu/+source/messaging-app/+question/271616 | 10:02 |
| CaptainHeavy | I didn't know whether this had already been addressed and then realised I should have double-checked by asking the question in the IRC here before posting it. Apologies if its already been answered. | 10:02 |
| guest42315 | so guys.. it;s like a week or so since we can't review apps on ubuntu store o_O | 10:07 |
| guest42315 | ogra_, the strange thing is i managed to review "match the colour", but i can't review new apps like Monster Wars ^^ | 10:08 |
| popey | JamesTait: ^ | 10:08 |
| guest42315 | https://bugs.launchpad.net/ubuntu/+source/unity-scope-click/+bug/1497305 | 10:10 |
| ubot5 | Ubuntu bug 1497305 in unity-scope-click (Ubuntu) "Can't review Monster Wars" [Undecided,Confirmed] | 10:10 |
| JamesTait | Oh, that's weird. I thought that was fixed. | 10:10 |
| guest42315 | https://bugs.launchpad.net/canonical-devices-system-image/+bug/1497161 | 10:10 |
| ubot5 | Ubuntu bug 1497161 in Canonical System Image "Can't add or edit reviews in Ubuntu Store" [Undecided,New] | 10:10 |
| guest42315 | i just tried to reivew monster wars and it didn't worked :D | 10:11 |
| guest42315 | strange enough is that i've manage to review match the colour | 10:11 |
| guest42315 | i'll try to review another old and new app | 10:11 |
| CaptainHeavy | Related to my question, what is the purpose of being able to select messages in the messaging app? There doesn't appear to be any actions that can be performed after selecting them (other than deletion) | 10:12 |
| guest42315 | same with dinosaur, can't review | 10:14 |
| guest42315 | JamesTait, look at the new added apps (0 reviews) https://uappexplorer.com/apps | 10:15 |
| guest42315 | monster wars should have had at least 20 reviews by now | 10:15 |
| JamesTait | That page does look suspiciously devoid of reviews. | 10:18 |
| guest42315 | yep | 10:18 |
| JamesTait | Narrowing it down to just new apps is helpful though. | 10:18 |
| JamesTait | IIRC it was an "old" app I reviewed (Dekko, I think). | 10:19 |
| guest42315 | try reviewing monster wars :D https://uappexplorer.com/app/monsterwars.t-mon | 10:19 |
| guest42315 | "10/10 make more levels! or else" something like that | 10:20 |
| * JamesTait installs another game on his phone, "Just for testing". 😝 | 10:21 | |
| guest42315 | sure :)) | 10:23 |
| JamesTait | "2015-09-21 11:27:13,694 - CRITICAL - Network error submitting a reviews for: monsterwars.t-mon | 10:28 |
| JamesTait | OK, we're looking into it again. Sorry about this. | 10:30 |
| guest42315 | JamesTait, no :P | 10:35 |
| guest42315 | JamesTait, np | 10:35 |
| WebVisitor-6 | Hi | 10:44 |
| WebVisitor-6 | I have a laptop with multi-touch support. Out of the box some of the gestures described in https://wiki.ubuntu.com/Multitouch#Supported_Gestures works perfectly. | 10:45 |
| WebVisitor-6 | There is one specifically which used to work on my old laptop but which does not on my new one | 10:45 |
| WebVisitor-6 | "3 finger press and drag to move window" | 10:45 |
| WebVisitor-6 | The laptop has a touchscreen, and that gesture works with the touchscreen if i put three fingers on the screen | 10:46 |
| WebVisitor-6 | but it does not work on the mousepad | 10:46 |
| WebVisitor-6 | does anyone know what the relationship is between the touchpad and the screen in the synaptic configuration? Why are some of the gestures available on the screen and not on the touchpad, and vice versa? | 10:46 |
| brendand | WebVisitor-6, it depends on the type of touchpad and the driver | 10:50 |
| brendand | WebVisitor-6, a lot of touchpads do not support true multi-touch | 10:50 |
| brendand | WebVisitor-6, they just interpret the gesture in hardware, which may be two finger drag or perhaps side edge drag | 10:51 |
| brendand | WebVisitor-6, whereas a touchscreen is a touchscreen | 10:51 |
| WebVisitor-6 | Ah ok but some of the other three- and four-finger gestures work on the touchpad | 10:52 |
| WebVisitor-6 | Its a synaptic device as far as i can tell | 10:52 |
| brendand | WebVisitor-6, like which ones? my touchpad is also synaptic but no 3-4 finger gestures work | 10:55 |
| WebVisitor-6 | xinput says "ETPS/2 Elantech Touchpad" | 10:59 |
| WebVisitor-6 | "Virtual core XTEST pointer", "Atmel Atmel maXTouch Digitizer", "ETPS/2 Elantech Touchpad" | 11:00 |
| === xavigarcia is now known as xavigarcia_lunch | ||
| === om26er_ is now known as om26er | ||
| jgdx | WebVisitor-6, #ubuntu might be of help. “touch” in this channel refers to ubuntu for phones. | 11:22 |
| jgdx | Elleo, that maliit restart to get introspection requirement for autopilot is a time sink. Is there a way around it? | 11:50 |
| Elleo | jgdx: not that I can think of, other than always running maliit with introspection, but I'd guess there's a runtime cost for that which we don't want to place on all our users | 11:54 |
| === xavigarcia_lunch is now known as xavigarcia | ||
| jgdx | Elleo, right, that's not wanted. :) | 12:01 |
| === alan_g is now known as alan_g|lunch | ||
| === chihchun is now known as chihchun_afk | ||
| robin-hero | Hi! I think I found a very serious "bug" with the unity8 and the indicator-datetime. In the translators message two strftime(3) variables are mixed. | 12:29 |
| robin-hero | This is from Launchpad indicator-datetime 29th string: | 12:30 |
| robin-hero | en_US example: "%a %d %b %l:%M %p" --> "Fri Oct 31 1:00 PM" en_GB example: "%a %b %d %l:%M %p" --> "Fri 31 Oct 1:00 PM" | 12:30 |
| robin-hero | But this is wrong. %d is for the day, not for month. And %b is for month, not for day. | 12:30 |
| robin-hero | Sorry, If it is not enough clear, the problem is with the Translators Comment on Launchpad | 12:31 |
| === alan_g|lunch is now known as alan_g | ||
| jgdx | robin-hero, hm, have you suggested the new translation? | 13:01 |
| daniel_ | Hello Everyone | 13:01 |
| daniel_ | i'm thinking about getting an Ubuntu Touch phone, and i have tested it on my Nexus 7 (flo), but i would like to know a bit more about Ubuntu for phones | 13:01 |
| jgdx | daniel_, go ahead! | 13:01 |
| robin-hero | jgdx: Yes, But need to modify the Translators Comment in the code | 13:02 |
| daniel_ | Once i have ubuntu phone, when a new update is released, do i have to wait for my network company to released just like google or simpler like Ubuntu for pc? | 13:03 |
| ogra_ | the latter | 13:03 |
| ogra_ | all updates for all devices come from a centra ubuntu server currently | 13:03 |
| ogra_ | *central | 13:03 |
| ogra_ | usually around the same time | 13:03 |
| ogra_ | (every 6 weeks) | 13:03 |
| daniel_ | so it means everyone gets it at once? | 13:03 |
| ogra_ | right | 13:04 |
| ogra_ | vendors are free to set up their own servers indeed, but today there is no such vendor yet | 13:04 |
| daniel_ | so i could say it is similar to an ubuntu pc update? | 13:04 |
| cwayne | it may be phased out over 24hrs or so, but it's certainly nothing like "wait until my carrier says its okay" | 13:04 |
| lotuspsychje | ogra2: so vendors are paying canonical for the server system rental? | 13:05 |
| daniel_ | that is really good to know | 13:05 |
| conall | Hi All, Im looking to install ubuntu touch on a google nexus one. Is there anywhere I can get an image for it? | 13:06 |
| lotuspsychje | !devices | conall | 13:06 |
| ubot5 | conall: You can find the full list of devices, official images, community images, and works in progress at https://wiki.ubuntu.com/Touch/Devices | 13:06 |
| daniel_ | one other question i got is, i think it is a bit hard to know because depends on other companies but, is canonical working on providing android apps or porting them? | 13:06 |
| ogra_ | lotuspsychje, no idea, i dont work in sales :P | 13:06 |
| ogra_ | perhaps thats part of the contracts | 13:07 |
| lotuspsychje | ok | 13:07 |
| conall | Thanks for that, yes. However, the nexus one is depricated, so I cant use the "ubuntu-device-flash" method | 13:07 |
| conall | So I think I need to download the image and use fastboot to flash it? | 13:07 |
| popey | i dont think we ever supported the nexus one | 13:08 |
| popey | so no, there is no image for it. | 13:08 |
| lotuspsychje | conall: you can try the XDA forums for an existing project maybe | 13:08 |
| daniel_ | please someone answer me? | 13:08 |
| popey | daniel_: no, we're not | 13:09 |
| ogra_ | no plans, no ... and dont be so impatient :) | 13:09 |
| daniel_ | i understand | 13:09 |
| conall | thanks lotuspsychje | 13:10 |
| lotuspsychje | daniel_: wich app would you need to see ported? | 13:10 |
| popey | conall: there's pretty much no chance of porting to the nexus one | 13:10 |
| daniel_ | skype because i use it everyday | 13:10 |
| popey | daniel_: ask microsoft :) | 13:10 |
| daniel_ | popey: lol as if they would listen | 13:11 |
| popey | if enough people ask, they listen | 13:11 |
| daniel_ | are most of ubuntu apps over html5? | 13:11 |
| daniel_ | skype also runs for browsers, what if your browser supports it? | 13:12 |
| lotuspsychje | daniel_: you can try out an alternative telegram | 13:12 |
| popey | some are html5, some are qt/qml, some are c++ | 13:12 |
| lotuspsychje | daniel_: no video chats for now though.. | 13:13 |
| popey | skype web doesn't fully work on our browser yet, maybe in the future | 13:13 |
| daniel_ | awwn :( | 13:13 |
| daniel_ | but still an alternative :) | 13:13 |
| ogra_ | once the browser has proper camera and mic integration (as well as WebRTC) it will perhaps | 13:13 |
| lotuspsychje | daniel_: convince all your friends to change to telegram :p | 13:13 |
| daniel_ | lotuspsychje: that will be a challenge :p | 13:14 |
| lotuspsychje | daniel_: lot of users already use it | 13:14 |
| daniel_ | lotuspsychje: when someone ports ubuntu touch to a sony device for example, if a new release is available, will this person have to make a new build or just update from canonical? | 13:15 |
| popey | if it's an officially supported device, we make the updates | 13:17 |
| t1mp | doesn't the nexus4 charge when connected to a laptop via a usb cable? | 13:18 |
| popey | it does, but not fast | 13:18 |
| t1mp | seems like mine is draining instead | 13:18 |
| ogra_ | well, and probably doesnt when you actually use it | 13:18 |
| t1mp | from 13% to 10% | 13:18 |
| t1mp | ah, ok | 13:18 |
| ogra_ | a USB port can only supply 500mA max | 13:19 |
| t1mp | I'm reflashing and installing packages and running autopilot tests | 13:19 |
| ogra_ | if your device uses more it will drain more than it charges | 13:19 |
| t1mp | ok | 13:19 |
| ogra_ | yeah, that surely draws more than 500mA | 13:19 |
| jgdx | kenvandine, morning :) ubuntu_system_settings.tests.test_security.SecurityTestCase.test_lock_security_focus_on_entry is flakey.. I'm trying to figure it out. But… do we need that test? | 13:19 |
| ogra_ | (and beyond that these 500mA are theoretical ... if your hub inside the laptop is shared the power might be shared too) | 13:19 |
| kenvandine | jgdx, does that test just verify the entry has focus? | 13:20 |
| jgdx | kenvandine, oui | 13:20 |
| kenvandine | so i think we had a bug where that entry didn't get focus automatically, so this test was added to prevent regressions | 13:20 |
| kenvandine | i suspect there is a real bug keeping that entry from getting focus | 13:21 |
| kenvandine | or... it's racy | 13:21 |
| jgdx | kenvandine, the dialog does not appear at all | 13:21 |
| kenvandine | like not allowing enough time for the entry to get focus? | 13:21 |
| kenvandine | so that's a real problem then right? | 13:21 |
| jgdx | i don't know yet, but that's the reason the test fails | 13:21 |
| t1mp | isn't this supposed to work? | 13:22 |
| t1mp | tim@ubuntu:~/.cache$ ubuntu-device-flash --clean-cache | 13:22 |
| t1mp | Unknown command `ubuntu-device-flash'. Please specify one command of: core, query or touch | 13:22 |
| ogra_ | well, you didnt specify a command | 13:22 |
| ogra_ | (like the message tells you) | 13:23 |
| jgdx | kenvandine, so I'm looking at the test being run on mako, and when it fails, the “Change passcode…” button is tapped (it gets darker) but nothing happens. | 13:23 |
| kenvandine | jgdx, that's not cool | 13:24 |
| jgdx | kenvandine, no it aint.. happens 5-10% of the times run | 13:24 |
| kenvandine | jgdx, but you can only reproduce it in the tests? | 13:24 |
| t1mp | ogra_: ahh.. I expected --clean-cache to just wipe my cache, even if I'm not flashing a new image | 13:25 |
| ogra_ | t1mp, indeed it will, but it still needs to knwo if it shoudl wipe a snappy cache or a phone cache ... you didnt tell it ;) | 13:25 |
| t1mp | ogra_: I'll just rm -rf ~/.cache/ubuntuimages ;) | 13:26 |
| ogra_ | why dont you just add the command ? | 13:26 |
| jgdx | kenvandine, i can reproduce the ultimate result (no click and click-effect on the button) by swiping on the button | 13:27 |
| jgdx | kenvandine, but not if I tap it. | 13:27 |
| kenvandine | jgdx, oh... maybe autopilot is broken? | 13:27 |
| jgdx | kenvandine, so I wonder if an autopilot tap is intepreted as a swipe | 13:27 |
| t1mp | ogra_: right.. I already deleted the cache directory, and now ubuntu-device-flash --clean-cache touch seems to work. But iirc before it was asking for a device id, which I could have figured out but it is less work for me to just delete the directory | 13:29 |
| jgdx | kenvandine, that would explain why we get failures like the last one here https://jenkins.qa.ubuntu.com/job/generic-deb-autopilot-runner-vivid-mako/3486/ | 13:30 |
| jgdx | kenvandine, the first one is real and I have a fix. | 13:30 |
| kenvandine | jgdx, cool | 13:31 |
| kenvandine | jgdx, what was the fix for the first one? | 13:31 |
| jgdx | kenvandine, it was broken but good. A new page was added and strings the test use were changed, but the test wasn't changed. | 13:32 |
| kenvandine | ah | 13:32 |
| robin-hero | Hey! Sometimes when I push the power button on my device to unlock it, it shows the power off/restart dialog. Anybody experience this? | 13:38 |
| peter-bittner | Me too, on bq Aquaris E5. This did not happen before the last OTA. | 13:39 |
| robin-hero | peter-bittner: Did you fill a bug report on launchpad? | 13:39 |
| robin-hero | I searched for it on Launchpad, but didn't find any releated | 13:40 |
| peter-bittner | No, there are more important things to complain about! :-) | 13:40 |
| ogra_ | there was a bu about it and that was supposed to be fixed | 13:41 |
| ogra_ | *bug | 13:41 |
| robin-hero | ogra_: Could you give us a link? | 13:41 |
| ogra_ | so worth opening it again or filing a new one | 13:41 |
| ogra_ | no | 13:42 |
| robin-hero | OK, I'll filled a new on | 13:42 |
| * ogra_ doesnt have it handy and is super busy with other stuff atm ... but i know we had one about two OTAs ago ... just file a new one | 13:42 | |
| robin-hero | ogra_: And for which project? Unity8? | 13:42 |
| ogra_ | either that or powerd ... in the end it needs to show up inder canonical-system-image anyway, so you could even file it directly against that | 13:43 |
| robin-hero | Filled: https://bugs.launchpad.net/canonical-devices-system-image/+bug/1498006 peter-bittner: please mark the "affects you". | 13:46 |
| ubot5 | Ubuntu bug 1498006 in Canonical System Image "Sometimes the power off/restart menu appears when I push the power button to unlock my phone." [Undecided,New] | 13:46 |
| peter-bittner | Done. | 13:47 |
| robin-hero | Thanks | 13:47 |
| peter-bittner | You're my hero! ;-) (just kidding) | 13:48 |
| robin-hero | :D | 13:49 |
| daniel_ | Thanks lotuspsychje popey ogra_ wish you all good luck with ubuntu touch :) | 13:50 |
| * mterry will buy cyphermox a beer if https://code.launchpad.net/~phablet-team/network-manager/lp1480877-wifi-rm-dup-scan-signals/+merge/271575 gets reviewed in time for OTA7. It's killing me when I walk around my neighborhood :) | 13:54 | |
| lotuspsychje | daniel_: good luck yourself purchasing ubuntu phone :p | 13:57 |
| robin-hero | sil2100: Hey! You mentioned in one of your email the translations automatically update every week. But as I see the last update was two weeks ago. | 14:03 |
| === chihchun_afk is now known as chihchun | ||
| === dandrader is now known as dandrader|afl | ||
| === dandrader|afl is now known as dandrader|afk | ||
| sil2100 | robin-hero: hey! Let me look into it later, it should have been all automated now | 14:11 |
| sil2100 | Thanks for mentioning | 14:11 |
| robin-hero | sil2100: Thanks | 14:11 |
| cyphermox | mterry: it was on my list of things from Tony to review already, but it's not that obvious a fix | 14:18 |
| cyphermox | in any case, I added a comment | 14:18 |
| mterry | cyphermox, fair. I don't know anything about the code itself. I just hate that bug :) | 14:19 |
| cyphermox | mterry: yeah, I understand | 14:19 |
| peter-bittner | A question on Ubuntu desktop (sorry!). Which one is the current default IM client? Still Empathy? | 14:19 |
| robin-hero | peter-bittner: Yes | 14:20 |
| cyphermox | maybe even not making this come up in propertiesChanged,, since it's not quite something up to the AP itself, more like internal NM data that users might use, but maybe don't need to be reminded everytime it changes | 14:20 |
| cyphermox | mterry: ^ | 14:20 |
| peter-bittner | Thanks! -- Are there any plans to package Empathy for ARM? (i.e. Ubuntu Touch) | 14:20 |
| mterry | cyphermox, the last-seen property? ::shrug:: I don't know the rationales there. but sure | 14:21 |
| cyphermox | yeah | 14:21 |
| cyphermox | it's something that already existed in NM, but we exposed it for GPS/Wifi triangulation so you could use very recent APs to figure out where you are | 14:21 |
| awe_ | cyphermox, my mp fixes a bug that was introduced by the location service | 14:24 |
| awe_ | there's no reason to send the same signal 2-3 times in time-span of less than a minute | 14:25 |
| awe_ | prior to the location service change | 14:25 |
| cyphermox | awe_: of course not, still *shrugs* have you discussed it with tvoss or whomever maintains location-service now? | 14:25 |
| awe_ | the code all used "schedule_cull_list" | 14:25 |
| awe_ | which prevented dups | 14:25 |
| awe_ | I added mandel to the review | 14:25 |
| cyphermox | I would simply rather make sure we make this as least intrusive as possible, and that upstream knows about the changes if they apply to master | 14:26 |
| cyphermox | (which I think they might, now) | 14:26 |
| awe_ | again, this was a change that *we* introduced | 14:26 |
| === dandrader|afk is now known as dandrader | ||
| cyphermox | sure, but you're proposing to remove some large parts of upstream code | 14:28 |
| cyphermox | in the end, I defer to your knowledge of the problem matter. I don't look at NM that much nowadays and I know you tested this | 14:29 |
| cyphermox | I'm just saying that if there's some parts of upstream code that you'd like to change, it would be good to bring it up upstream, resubmit patches and whatnot | 14:29 |
| awe_ | the only big removal was bss_updated_cb, and the only thing it was used for was scheduling a scan_list_cull and updating last_seen | 14:30 |
| awe_ | since scan_done_cb does this already, it's really not adding anything useful | 14:30 |
| cyphermox | then merge it and send the patch upstream, please | 14:30 |
| awe_ | again... I'll check, but I'm not sure you pushed the change upstream that caused the problem | 14:31 |
| awe_ | but let me check | 14:31 |
| cyphermox | awe_: I did send it to the ML, no hurt in pushing it again with updates so it's not broken. | 14:38 |
| awe_ | k, as mentioned I'll take a look | 14:38 |
| om26er | H! is there a way to change screen brightness through powerd ? | 14:59 |
| === chihchun is now known as chihchun_afk | ||
| mterry | awe_, does meizu mx4 work with 3g here in the states? I'm having a hard time finding info online about it | 15:27 |
| awe_ | mterry, yes. no lte | 15:27 |
| mterry | ah... looks like yes for at&t anyway... | 15:27 |
| awe_ | I use AT&T, and it works fine | 15:28 |
| mterry | awe_, I have metropcs... let me see if those use the same bands... | 15:28 |
| awe_ | not sure if I've tried my T-Mobil SIM, but pretty sure I have | 15:28 |
| awe_ | yea.. I think it has 'em all | 15:28 |
| awe_ | ( for 3g ) | 15:28 |
| mterry | awe_, I'm using a nexus 4 right now for personal use, so I'm used to no 4g | 15:28 |
| awe_ | wow, old man's phone | 15:29 |
| awe_ | ;D | 15:29 |
| mterry | awe_, it's not a great phone these days | 15:29 |
| === nuclearbob is now known as nuclearbob_lunch | ||
| pmcgowan | ogra2, or Laney should I be able to disable apport simply by changing /etc/default/apport to enabled=0, that file is empty on the phone | 16:01 |
| ogra_ | stop pining my snappy testuser :P | 16:02 |
| ogra_ | pmcgowan, you should just be able to disable it in the UI as i understand | 16:03 |
| jibel | pmcgowan, but a different setting for stable and proposed would mean a new image to publish to stable | 16:03 |
| jibel | and no way for the user to enable it | 16:03 |
| pmcgowan | ogra_, I want apport to not run the UI just controls whoopsie | 16:05 |
| popey | jibel: pmcgowan do we have a clear plan for silo 55 (media hub mpris stuff)? Was any decision taken to land it after we looked at the list of gotchas the music guys found? | 16:06 |
| ogra_ | pmcgowan, whoopsie controls apport | 16:06 |
| Laney | I don't think that whoopsie turns apport off | 16:08 |
| Laney | pmcgowan: I think so | 16:08 |
| ogra_ | ah, i thought it should | 16:08 |
| Laney | It just turns off uploading | 16:08 |
| popey | the wording in system settings certainly only implies the tickbox is responsible for controlling "sending to canonical" not "filling /var/crash" | 16:09 |
| davmor2 | ogra_: Laney is right it turns off uploading which means you disk slowly fills with bug report that sit there | 16:09 |
| popey | (interestingly it's worth noting that the machine ID is fixed, if you sell your phone to someone else and they do a factory reset or wipe, then your crash reports appear with theirs before you bought the phone) | 16:10 |
| popey | (on errors.ubuntu.com) | 16:10 |
| popey | (I have someone elses phone, and I can see their crashes) | 16:11 |
| jibel | ogra_, turning off whoopsie addresses privacy concerns but doesn't turn off apport. | 16:13 |
| ogra_ | k | 16:14 |
| ogra_ | i thought bdmurray had said that was handled internally by whoopsie | 16:14 |
| ogra_ | but i obviously mis-remember that | 16:14 |
| ogra_ | ( Laney is definitely deeper int the code than i ever was :) ) | 16:15 |
| Laney | not that deep | 16:15 |
| ogra_ | deeper :) | 16:15 |
| popey | hm, neverball no longer starts on my rc-proposed phone | 16:15 |
| popey | did we break mir/sdl? | 16:15 |
| Laney | like maybe the dermis | 16:15 |
| popey | works on my ota-6 phone | 16:15 |
| ogra_ | yeah, better than the goosebump surfaces i know :) | 16:16 |
| ogra_ | popey, confirmed | 16:16 |
| popey | filed https://github.com/pseuudonym404/neverball-touch/issues/11 | 16:16 |
| Laney | hey bart, your epidermis is showing | 16:17 |
| popey | this is worrying, we broke apps in the store | 16:17 |
| ogra_ | seems to crash really early | 16:17 |
| ogra_ | like even before the start animation comes up properly | 16:17 |
| popey | hmm, mediascanner is constantly crashing here on rc-proposed | 16:27 |
| popey | in a loop -> mediascanner -> apport -> mediascanner -> apport | 16:27 |
| popey | every 2 seconds | 16:28 |
| popey | so, tedg I have a counter-argument to your position that crashing apps aren't a problem :) http://paste.ubuntu.com/12515201/ | 16:29 |
| tedg | popey: ? | 16:30 |
| tedg | I think crashes are a problem :-) | 16:30 |
| popey | pft :) | 16:31 |
| guest42315 | OTA7 is a week away, right? | 16:32 |
| popey | or so | 16:32 |
| guest42315 | ish :P | 16:32 |
| guest42315 | c00l | 16:32 |
| davmor2 | popey: I see no mediascanner crash | 16:33 |
| popey | probably media specific | 16:33 |
| popey | but there's no log | 16:33 |
| ogra_ | guest42315, if there are serious bugs found during QA the OTA will be delayed ... so dont take the date as set in stone, it is just an estimate ... quality > being on time | 16:33 |
| pmcgowan | ogra2, Laney hey I lost connection, what was the answer to turn apport off? as it didnt work for me setting that flag | 16:33 |
| ogra_ | pmcgowan, Laney claimed (obviously rightly) that whoopsie doesnt turn off apport | 16:34 |
| pmcgowan | but how does one turn it off on the phone | 16:34 |
| ogra_ | by editing /etc7default/approt atm i guess | 16:34 |
| davmor2 | pmcgowan: it turns off uploading not apport | 16:35 |
| ogra_ | yay, typos galore | 16:35 |
| pmcgowan | ogra_, I did that it the service is still running | 16:35 |
| pmcgowan | unless I misunderstand | 16:35 |
| ogra_ | you need to reboot | 16:35 |
| pmcgowan | did | 16:35 |
| ogra_ | its not a dynamic toggle | 16:35 |
| ogra_ | hmm | 16:35 |
| pmcgowan | service --status-all shows apport ? | 16:35 |
| pmcgowan | does that mean its running | 16:35 |
| ogra_ | the upstart job of apport shoudl respect it | 16:35 |
| pmcgowan | popey, abeato is making fixes to media hub, then need to retest I assume | 16:36 |
| davmor2 | popey: I can confirm your neverball issue, possibly one of the many update to mir/qtmir/unity8/ however we do not shockingly start every app in the store to see if we regressed. | 16:37 |
| popey | ok, thanks pmcgowan | 16:37 |
| popey | davmor2: shouldn't have to, should be backwards compatible | 16:37 |
| popey | davmor2: shall i file a mir bug do you think? | 16:38 |
| davmor2 | popey: well it seems ot be crashing before the app opens, so that to me would suggest that the app calls it's import maybe and dies there. Just checking to see if I see a log file | 16:39 |
| davmor2 | popey: Failure to initialize SDL (Failed to connect to the Mir Server) | 16:40 |
| popey | davmor2: yeah, thats in the github issue | 16:40 |
| popey | uh. I just disabled apport by putting "enabled=0" in /etc/default/apport and rebooted the phone, now my phone is in a reboot loop! | 16:40 |
| jibel | pmcgowan, this flag controls the upstart job, you first have to stop apport as root, then set enabled=0 | 16:40 |
| Laney | pmcgowan: cat /proc/sys/kernel/core_pattern ? | 16:41 |
| jibel | not trivial to do from the UI as phablet | 16:41 |
| ogra_ | popey, thou shalt not disable bug reporting ! | 16:43 |
| popey | clearly | 16:43 |
| popey | well this is annoying | 16:44 |
| popey | phew, managed to ssh in and remove that line from /etc/default/apport before it rebooted again | 16:44 |
| ogra_ | and that fixed it ? | 16:44 |
| === pstolowski is now known as pstolowski|eod | ||
| popey | dunno, rebooting | 16:45 |
| ogra_ | curious how this could break (or fix) the boot | 16:45 |
| ogra_ | sounds honestly more like a coincidence | 16:45 |
| popey | yup | 16:45 |
| popey | fixed | 16:45 |
| ogra_ | insane | 16:46 |
| popey | yes | 16:46 |
| popey | pmcgowan: don't set that :) | 16:46 |
| jibel | popey, can you try again, it's weird. | 16:46 |
| popey | okay | 16:46 |
| popey | the only thing I will do is edit /etc/default/apport, add enabled=0, save and reboot | 16:46 |
| popey | okay? | 16:46 |
| jibel | popey, yes | 16:47 |
| popey | ok, done | 16:47 |
| popey | it gets as far as unity, the indicators etc, and then reboots. | 16:48 |
| Laney | I just tried it, worked for me | 16:48 |
| Laney | forgot what I'm running though | 16:49 |
| pmcgowan | let me look | 16:49 |
| popey | i have things crashing though | 16:49 |
| popey | which might be what triggers it | 16:49 |
| jibel | popey, it works fine here, however the core pattern is set to /bad_core_pattern | 16:49 |
| Laney | unity works so probably ...vivid | 16:49 |
| popey | (mediascanner is crashing in a loop for me) | 16:49 |
| jibel | instead of core | 16:49 |
| Laney | phablet@ubuntu-phablet:~$ cat /proc/sys/kernel/core_pattern | 16:49 |
| Laney | core | 16:49 |
| popey | so maybe if one of you can turn it on and try and orchestrate a crash of some kind | 16:50 |
| popey | to see if that triggers the reboot | 16:50 |
| popey | s/on/off/ | 16:50 |
| pmcgowan | Laney, I get "/bad_core_pattern" | 16:50 |
| popey | same here | 16:51 |
| pmcgowan | popey, jibel what is working for you/ | 16:51 |
| pmcgowan | ? | 16:51 |
| Laney | I don't even know what that means | 16:51 |
| Laney | sorry | 16:51 |
| popey | pmcgowan: if I set enabled=0 and reboot, my phone goes into a reboot loop. I suspect because I have a mediascanner problem (it's crashing constantly) which is probably triggering the reboot. | 16:51 |
| popey | This is on krillin, my rc-proposed phone | 16:52 |
| pmcgowan | Laney, popey how do I check that apport is actually disabled, as it shows up in the service list | 16:52 |
| jibel | pmcgowan, apport doesn't generate a crash report | 16:52 |
| Laney | if you don't have apport in the core_pattern then crashes won't be passed to it | 16:52 |
| pmcgowan | ok | 16:52 |
| Laney | check it on your desktop | 16:53 |
| Laney | that's how it works | 16:53 |
| popey | bah, undone the edit to /etc/default/apport but now it's kept /proc/sys/kernel/core_pattern as /bad_core_pattern and my phone keeps rebooting :( | 16:53 |
| ogra_ | popey, i still wonder if its a coincidence ... there is a watchdog that cyphermox implemented that force-reboots if upstart processes go into a loop | 16:56 |
| popey | nice | 16:56 |
| popey | that seems somewhat unuseful | 16:56 |
| ogra_ | yeah, it should stop at some point and pop up "call the support" or some such :) | 16:56 |
| popey | i have barely a few seconds during the boot process in which I can ssh in and fix stuff | 16:57 |
| popey | :( | 16:58 |
| popey | any idea how I can make this not-broken? | 16:58 |
| * ogra_ has no clue about the watchdog | 16:58 | |
| ogra_ | (*if* that is what kicks in) | 16:58 |
| === alan_g is now known as alan_g|EOD | ||
| === ogra2 is now known as ogra-snappy-test | ||
| sergiusens | popey, ogra_ flash an adb enabled recovery image and edit from there maybe? | 17:19 |
| ogra_ | yeah, if all else fails | 17:19 |
| popey | ugh | 17:26 |
| popey | editing the file doesn't make any difference now, it's just in a constant reboot loop | 17:27 |
| ogra_ | yeah, i guess its the watchdog rather than the file | 17:27 |
| ogra_ | there should be messages soemwhere ... either in syslog or dmesg | 17:28 |
| popey | why on earth would rebooting be the right thing to do when you have a crashy process? | 17:28 |
| popey | from what? | 17:28 |
| ogra_ | popey, dont ask me :P | 17:28 |
| * ogra_ never liked that idea ... | 17:28 | |
| ogra_ | i think android does something similar | 17:28 |
| ogra_ | cyphermox, is there a way to intercept the reboot loop you get when watchdog kicks in ? | 17:29 |
| popey | http://paste.ubuntu.com/12515647/ that's all I get from syslog | 17:30 |
| popey | Sep 21 17:29:43 ubuntu-phablet session-watchdog: 'mediascanner-2.0' (instance '') hit respawn limit - asking logind to reboot | 17:30 |
| ogra_ | Sep 21 17:29:43 ubuntu-phablet session-watchdog: 'mediascanner-2.0' (instance '') hit respawn limit - asking logind to reboot | 17:30 |
| ogra_ | hah | 17:30 |
| popey | :) | 17:30 |
| popey | echo manual >> ~/.config/upstart/mediascanner-2.0.override | 17:31 |
| popey | might do it | 17:31 |
| ogra_ | yeah | 17:31 |
| ogra_ | degraded mode :) | 17:31 |
| ogra_ | i guess you could do the same for the watchdog | 17:32 |
| ogra_ | iirc it is an upstart job too | 17:32 |
| dobey | jdstrand: hi. what's the landing process for apparmor-easyprof-ubuntu changes? is it appropriate to add an MP for that to a silo in ci train? | 17:32 |
| popey | right, got my phone to stop rebooting with that | 17:33 |
| pmcgowan | jamesh, why would this be after disabling apport? cat /proc/sys/kernel/core_pattern | 17:34 |
| pmcgowan | "/bad_core_pattern" | 17:35 |
| ogra_ | pmcgowan, i think thats an android kernel default | 17:42 |
| ogra_ | to prevent crating any core files at all | 17:42 |
| jibel | pmcgowan, what ogra said, it is set by the kernel depending on the platform | 17:42 |
| jibel | ie different devices will have different default pattern | 17:43 |
| jdstrand | dobey: we don't do MP builds for apparmor-easyprof-ubuntu, but an MP is fine. but this needs discussion because it requires changing the default template | 17:43 |
| ogra_ | it usually says "core" in other kernels | 17:43 |
| jdstrand | dobey: which will trigger a recompile of a lot of app policy | 17:43 |
| ogra_ | which would create a /tmp/core file if somethionng crashes | 17:43 |
| ogra_ | (including a 1:1 copy of your ram) | 17:43 |
| jdstrand | dobey: and that is something we try to minimize in stable updates | 17:44 |
| pmcgowan | ok | 17:44 |
| jdstrand | there is something else that people want that will trigger rebuilds, so they should land together if they are supposed to land on the device | 17:44 |
| jdstrand | (related to frameworks) | 17:45 |
| pmcgowan | ogra_, jibel I tried setting the pattern to core but on reboot it resets to the default again | 17:45 |
| jdstrand | but, long story short, the MP to the project is fine (I have it on my todo list already) | 17:45 |
| ogra_ | pmcgowan, yeah | 17:45 |
| ogra_ | and you really dont want it set to core | 17:46 |
| pmcgowan | ok | 17:46 |
| ogra_ | would eat your disk | 17:46 |
| dobey | jdstrand: hmm, ok. does jenkins monitor MPs for that project? what's the best way to get it installed on a device for testing? | 17:46 |
| pmcgowan | let me see if enable apport will set it back | 17:46 |
| ogra_ | it does, i just checked the apport upstart job | 17:46 |
| jdstrand | dobey: it does not. we upload to the ppa | 17:48 |
| dobey | ok | 17:49 |
| jdstrand | dobey: so, just prepare a deb like you would for the archive, then dpkg -i it | 17:49 |
| dobey | jdstrand: right. i was just hoping that i could point at one in jenkins or such. would it make sense for you to manually upload a package to our silo ppa, so we can test the whole stack there, prior to it landing in the overlay ppa? | 17:51 |
| cyphermox | ogra_: nothing I can think of right this minute. You'd have to get in recovery and modify the job I think | 17:54 |
| jdstrand | dobey: I can, but this package may be changing in other ways for the eventual landing (the aforementioned landing together) | 17:55 |
| jdstrand | problem is, not sure what that is going to look like yet | 17:55 |
| ogra_ | cyphermox, yeah, i think popey solved it differently | 17:55 |
| dobey | jdstrand: sure | 17:55 |
| cyphermox | yeah, I see | 17:55 |
| popey | well, worked around it | 17:56 |
| jdstrand | dobey: so, can you add a comment in the MP on what you'd like from us/me for you to test this? | 17:57 |
| dobey | jdstrand: sure. | 17:58 |
| jdstrand | thanks | 18:00 |
| george_e | Quick question... is there a way to build the click package for an app from the command line instead of Qt Creator? I'd like the click package to target armhf. | 18:23 |
| george_e | I'd also like to build the armhf chroot from the command line as well. | 18:23 |
| george_e | Is this possible? | 18:23 |
| peter-bittner | $ click build . | 18:25 |
| george_e | I tried that but it targets amd64 (my host) instead of armhf. | 18:25 |
| peter-bittner | What's in you manifest.json? | 18:27 |
| peter-bittner | There is an "architecture" property, not sure though whether it makes a difference for the build tool. | 18:28 |
| pmcgowan | ogra_, still around? | 18:29 |
| ogra_ | pmcgowan, yep | 18:29 |
| george_e | peter-bittner oh, oops. Indeed there is and it's set to amd64 :P | 18:29 |
| pmcgowan | ogra_, why is /etc/default/apport on the phones empty vs containing the same contents as desktop? | 18:29 |
| pmcgowan | is it coming from a tarball ? | 18:29 |
| ogra_ | pmcgowan, because we dont use apport but apport-noui ... | 18:30 |
| ogra_ | so it isnt the same as on desktop | 18:30 |
| pmcgowan | oh | 18:30 |
| pmcgowan | ogra_, apport is installed though | 18:31 |
| pmcgowan | and dpkg tells me the file is from that package | 18:31 |
| pmcgowan | oh maybe not | 18:31 |
| pmcgowan | yeah it does | 18:31 |
| peter-bittner | Has porting Empathy (the instant messenger) to armhf / Ubuntu Touch been discussed here or on the mailing list before? | 18:32 |
| ogra_ | pmcgowan, | 18:35 |
| ogra_ | phablet@ubuntu-phablet:~$ mount|grep default/apport | 18:35 |
| ogra_ | /dev/mmcblk0p16 on /etc/default/apport type ext4 (rw,relatime,discard,nodelalloc,data=journal) | 18:35 |
| pitti | pmcgowan: if you put "enabled=0" into /etc/default/apport it should not start at boot, and consequently not trigger whoopsie either | 18:35 |
| ogra_ | because we mount an empty writable file on top | 18:35 |
| ogra_ | phablet@ubuntu-phablet:~$ grep default/apport /etc/system-image/writable-paths | 18:37 |
| ogra_ | /etc/default/apport auto persistent none none | 18:37 |
| ogra_ | phablet@ubuntu-phablet:~$ | 18:37 |
| pmcgowan | pitti, yeah that seems to work, I would like to do it conditionally based on the channel | 18:37 |
| ogra_ | the first "none" there should have been "transition" | 18:37 |
| popey | this could make things worse for some people | 18:37 |
| pmcgowan | how so | 18:37 |
| popey | we should consider making this default after caring for the watchdog | 18:37 |
| ogra_ | popey, not really | 18:37 |
| popey | well, you could get into the situation I'm in | 18:38 |
| popey | something crashy, watchdog jumps in and reboots the device | 18:38 |
| ogra_ | sure, but the default file wouldnt have helped there | 18:38 |
| ogra_ | thats a watchdog thing | 18:38 |
| popey | that didn't happen with apport enabled because there was time bwteeen each restart of process | 18:38 |
| popey | no, the default _harmed_ me here | 18:38 |
| popey | thats my point | 18:38 |
| ogra_ | how would it harm you there ? | 18:38 |
| popey | (if apport disabled had been default, I'd be in a reboot loop and no way to get out easily) | 18:38 |
| ogra_ | empty is not different from enabled=1 | 18:39 |
| popey | which is arguably worse than crash dumping which slows the phone | 18:39 |
| popey | no, I'm not making this clear. | 18:39 |
| ogra_ | it is absolutely identical | 18:39 |
| popey | no no | 18:39 |
| pmcgowan | enabled=0 caused his grief | 18:39 |
| popey | I'm saying if you make the default enabled=0, it will make it _worse_ for some people | 18:39 |
| ogra_ | ah | 18:39 |
| ogra_ | yeah, agreed | 18:39 |
| popey | unless watchdog is cared for also | 18:39 |
| pmcgowan | so yeah we need that fixed | 18:39 |
| ogra_ | because the watchdog will force-reboot you all the time | 18:40 |
| pmcgowan | popey, do you have aenough for a bug for that? | 18:40 |
| ogra_ | by design | 18:40 |
| popey | sure | 18:40 |
| popey | doing now | 18:40 |
| popey | against system image? | 18:40 |
| ogra_ | well, it is multiple bugs | 18:40 |
| popey | sure, apport is one, I assume pmcgowan has a bug for making that off by default already? | 18:40 |
| popey | "my" bug is watchdog being to agressive | 18:40 |
| ogra_ | /etc/default/apport being empty is one ... someone needs to review the wtachdog concept | 18:41 |
| * ogra_ never really got the rationale behind this | 18:41 | |
| pmcgowan | what implements this watchdog? | 18:42 |
| ogra_ | upstart-watchdog | 18:42 |
| ogra_ | iirc | 18:42 |
| * popey files a bug | 18:43 | |
| ogra_ | pmcgowan, https://launchpad.net/ubuntu/+source/upstart-watchdog/0.3 | 18:43 |
| ogra_ | err | 18:43 |
| ogra_ | https://launchpad.net/ubuntu/+source/upstart-watchdog | 18:43 |
| popey | https://launchpad.net/ubuntu/+source/upstart-watchdog/+bugs zarro boogs! | 18:43 |
| ogra_ | bugfree \o/ | 18:43 |
| * popey feels blessed to file the first one | 18:43 | |
| ogra_ | boo | 18:44 |
| ogra_ | just make your system writable and it wont reboot :P | 18:44 |
| ogra_ | (is what the changelog for 0.3 says) | 18:44 |
| popey | ugh | 18:46 |
| pmcgowan | hmm | 18:46 |
| popey | pmcgowan: bug 1498133 | 18:47 |
| ubot5 | bug 1498133 in upstart-watchdog (Ubuntu) "Watchdog is too aggressive, can lead to unusable device" [Undecided,New] https://launchpad.net/bugs/1498133 | 18:47 |
| pmcgowan | thanks | 18:47 |
| pmcgowan | will mark the other with a pointer | 18:47 |
| pmcgowan | ogra_, where does the bug go about the empty file? | 18:49 |
| ogra_ | lxc-android-config | 18:49 |
| ogra_ | it ships /etc/system-image/writable-paths | 18:49 |
| pmcgowan | ok I have a task for that already | 18:49 |
| ogra_ | and looking closer the implementation is completely wrong | 18:50 |
| pmcgowan | oh my | 18:51 |
| ogra_ | it should use the same setup as hostanme and timezone use ... (being a link to /etc/writable, then teh content would have been there ...) | 18:51 |
| ogra_ | (i think i noted that in the original bug) | 18:51 |
| pmcgowan | ogra_, where did you note it? | 18:53 |
| ogra_ | in the original bug about "not being able to swithc of bug reporting" i think | 18:54 |
| pmcgowan | ok will find it | 18:54 |
| ogra_ | well, if not, just add it as pointer for the dev implementing it | 18:55 |
| ogra_ | i might misremember | 18:55 |
| pmcgowan | so should it not be in writeable_paths at all? | 18:56 |
| pmcgowan | ogra_, ? | 18:56 |
| pmcgowan | ogra_, seince its not directly in /etc isnt it ok as it is in the paths file? | 18:59 |
| ogra_ | well, you want it in /etc/writable (thats a writable dir) and have /etc/default/apport being a link to it ... putting it into /etc/writable and creating the link actually needs to happen at build time (livecd-rootfs task) | 19:03 |
| ogra_ | pmcgowan, http://bazaar.launchpad.net/~ubuntu-core-dev/livecd-rootfs/trunk/view/head:/live-build/ubuntu-touch/hooks/03-etc-writable.chroot is the code that handles the other files (feel free to link in the bug) | 19:03 |
| pmcgowan | ogra_, ok thanks, so does it still need to live in writeable_paths? I do not see hostname there | 19:05 |
| ogra_ | pmcgowan, no, actually not | 19:06 |
| pmcgowan | ack | 19:06 |
| === dandrader is now known as dandrader|afk | ||
| pmcgowan | popey, so, why would you not get a boot loop if apport was enabled? | 19:13 |
| popey | because apport would slow down the interval between one crash and the next | 19:13 |
| popey | as it spits out junk in /var/crash | 19:13 |
| pmcgowan | so help me, why does slowing it down fix the boot loop? | 19:14 |
| pmcgowan | wouldnt it just reboot later | 19:14 |
| popey | i think from what ogra_ was saying, the watchdog thing notices something respawning _really_ rapidly | 19:14 |
| popey | which, without apport, mediascanner (for me) was doing | 19:14 |
| popey | _with_ apport it crashes the same amount, but there's a delay between one and the next | 19:15 |
| popey | so I presume watchdog is look for N crashes per X seconds | 19:15 |
| popey | and apport makes X longer | 19:15 |
| popey | so doesn't trigger the reboot | 19:15 |
| pmcgowan | hmm ok | 19:16 |
| pmcgowan | so what was this fixing I wonder | 19:16 |
| popey | good question :) | 19:17 |
| popey | i have no email anywhere about it | 19:18 |
| popey | grepping irc finds https://launchpad.net/bugs/1394350 | 19:18 |
| ubot5 | Ubuntu bug 1394350 in Canonical System Image "[ubuntu-touch] system not recovering automatically when a critical service reaches the upstart respawn limit" [Undecided,In progress] | 19:18 |
| ogra_ | /usr/share/upstart/sessions/session-watchdog.conf | 19:18 |
| ogra_ | thats the thing that forces the reboot | 19:18 |
| ogra_ | start on stopped RESULT="failed" PROCESS="respawn" | 19:18 |
| pmcgowan | yeah I see that | 19:19 |
| popey | so it fixes an issue with upstart not restarting thing | 19:19 |
| ogra_ | means essentially: if any process respawns | 19:19 |
| pmcgowan | oh this is broken | 19:19 |
| ogra_ | well, there is a respawn limit built into upstart | 19:19 |
| ogra_ | so it will only restart after "n" occurences | 19:20 |
| ogra_ | i.e. if a process goes into a loop | 19:20 |
| pmcgowan | but it must assume it will not fail on that reboot, which is not the case | 19:20 |
| ogra_ | a single crash with respawn wouldnt cause it | 19:20 |
| pmcgowan | sure | 19:20 |
| popey | which is what gives me ~30 seconds before my phone reboots | 19:21 |
| popey | as mediascanner restarts that many times until watchdog says "no, reboot" | 19:21 |
| ogra_ | so it worked corectly as designed :) | 19:21 |
| pmcgowan | so it needs more smarts to see the respawns were all on the same boot or something | 19:21 |
| popey | no | 19:21 |
| popey | the respawns _were_ all in the same boot | 19:22 |
| ogra_ | the design was made between awe_ cyphermox and ricardo ... | 19:22 |
| popey | it _assumes_ a reboot would "fix" it | 19:22 |
| ogra_ | no | 19:22 |
| cyphermox | popey: correct, it *assumes* a reboot would fix it | 19:22 |
| ogra_ | iirc there were plans to actually notify recovery about that and to have a crash handler in the recovery img | 19:22 |
| popey | cyphermox: :) | 19:22 |
| cyphermox | and if it doesn't, time to go in recovery and reflash, basically | 19:22 |
| ogra_ | which tells you to call the support | 19:22 |
| popey | right | 19:22 |
| ogra_ | that would kick in if it notices a reboot loop | 19:23 |
| cyphermox | that never went in, because some disagreed it was the right thing to do | 19:23 |
| ogra_ | i think the recovery part was never implemented | 19:23 |
| popey | all this because I have some dodgy artwork on an mp3 :S | 19:23 |
| awe_ | pmcgowan, basically we had an instance where an indicator would crash and respawn and hit the limit | 19:23 |
| cyphermox | (or wasn't a priority, I don't recall) | 19:23 |
| pmcgowan | but help me out, if we just let it boot without the service, the user could get an update that maybe fixes it | 19:23 |
| ogra_ | cyphermox, i think we never got the UI bits for it | 19:23 |
| ogra_ | and it was low prio | 19:23 |
| awe_ | and we decided that we should monitor system jobs and if any of them got into this state | 19:23 |
| awe_ | that rebooting was the best solution | 19:23 |
| popey | s/best/least worst/ :) | 19:24 |
| awe_ | pretty sure ricmm added the job | 19:24 |
| ogra_ | awe_, the prob is that the rebooting never stops today :) | 19:24 |
| ogra_ | awe_, cyphermox did | 19:24 |
| pmcgowan | but in this case the cure was worse than the symptom | 19:24 |
| cyphermox | right. one thing would be to whitelist specific services that are absolutely required, but we had initially agreed everything on the system was necessary, IIRC | 19:24 |
| cyphermox | popey: dodgy artwork shouldn't crash mediascanner though :/ | 19:25 |
| awe_ | yes, pretty sure we were aware that in certain instances the rebooting could continue... and at that point, it was time to send the phone back | 19:25 |
| ogra_ | i think for now turning it off, getting the recovery side implemented and then re-enabling it is the proper path | 19:25 |
| popey | cyphermox: true, but this chain of events leads to "take the phone for service" | 19:25 |
| pmcgowan | yeah we dont want that | 19:25 |
| awe_ | popey, yes... we understood that | 19:25 |
| cyphermox | indeed; this is meant more for the finished product than when we're developing on it | 19:25 |
| pmcgowan | we want the user to get the next update that fixes it | 19:25 |
| awe_ | at some point, the phones may be broken enough that they need to get sent back | 19:25 |
| ogra_ | cyphermox, this is happening on the finished product :) | 19:25 |
| cyphermox | ogra_: is it? | 19:25 |
| pmcgowan | not really it happened on proposed | 19:26 |
| awe_ | if the phone boots every time and a criticial service like NM won't start. How's the user supposed to fix that? | 19:26 |
| pmcgowan | but still seems dangerous | 19:26 |
| ogra_ | pmcgowan, but only popey sees it ... | 19:26 |
| cyphermox | awe_: flashing a clean image should never have NM not start. | 19:26 |
| ogra_ | pmcgowan, what makes you sure it isnt a file that makes mediascanner crash | 19:26 |
| popey | ogra_: how many people have disabled apport? | 19:26 |
| cyphermox | the fix is to flash a clean, non-buggy image. | 19:26 |
| awe_ | cyphermox, I'm talking about a customer device | 19:26 |
| ogra_ | popey, well, that the watchdog doesnt work when apport is enabled is a bug | 19:27 |
| cyphermox | awe_: then I'm not sure I follow | 19:27 |
| * guest42315 snappy <3 | 19:27 | |
| awe_ | if a phone gets in such a state that a critical service can't start, there's not much we can do | 19:27 |
| cyphermox | awe_: right | 19:27 |
| ogra_ | popey, you should actually have seen it even with apport enabled | 19:27 |
| pmcgowan | depends on which service o? | 19:27 |
| awe_ | and yes, we discussed that you could get into an infinite reboot scenario | 19:27 |
| ogra_ | awe_, but there were plans to make you end up in recovery | 19:27 |
| popey | ogra_: no, because apport slowed down the respawns | 19:27 |
| cyphermox | awe_: but rebooting *might* fix it if the user has just been doing someting funny. | 19:27 |
| awe_ | but I also recall discussing some kind of throttling | 19:27 |
| awe_ | cyphermox, correct, which is why we reboot | 19:28 |
| ogra_ | popey, yes, thats a bug, preventing watchdog from doing what it is supposed to do | 19:28 |
| cyphermox | yep | 19:28 |
| ogra_ | popey, you would sit forever with a slow phone | 19:28 |
| popey | I did :) | 19:28 |
| ogra_ | and no media access :) | 19:28 |
| cyphermox | awe_: ok, so it looks like you all have a firm grasp on it, I'll let y'all work and get back to my grub bugs :) | 19:28 |
| awe_ | cyphermox, updated all the MPs by the way... | 19:28 |
| cyphermox | ok | 19:28 |
| pmcgowan | awe_, throtttling the reboots or what? | 19:29 |
| ogra_ | cyphermox, haha ... nobody will work on it ... at least nobody from the people discussing here | 19:29 |
| cyphermox | i will review again $very_late_tonight | 19:29 |
| awe_ | pmcgowan, yea... throttling reboots | 19:29 |
| awe_ | cyphermox, thanks! | 19:29 |
| awe_ | pmcgowan, but not sure that got implemented | 19:29 |
| pmcgowan | I dont see it | 19:29 |
| awe_ | pmcgowan, I think I said we thought about it, but it didn't get implemented | 19:30 |
| cyphermox | throttling reboots might prove difficult, but if we had the recovery mechnism when in a loop, that might fare better if we also provide the user a way to disable the watchdog | 19:30 |
| pmcgowan | and we knew of cases where are process respawned and rebooting fixed it? | 19:30 |
| ogra_ | awe_, we were waiting for design to come back with UI elements for recovery and then it was supposed to boot into recovery after n reboots | 19:30 |
| ogra_ | that never happened | 19:30 |
| awe_ | pmcgowan, this whole thing stemmed from an indicator-network bug | 19:30 |
| awe_ | and I asked what happens if upstart stops a critical process due to respawn limits | 19:30 |
| * pmcgowan thinks about reports of boot loops in the wild | 19:31 | |
| ogra_ | i dont think throttling was on the plate in the end | 19:31 |
| awe_ | ogra_, I think it slid off the plate | 19:31 |
| awe_ | ;/ | 19:31 |
| ogra_ | pmcgowan, yeah, popey's respawn could be caused by a bad file mediascanner cant handle | 19:31 |
| ogra_ | and could also already be on customer ohones out there | 19:32 |
| ogra_ | *phones | 19:32 |
| pmcgowan | yeah this is scaring me | 19:32 |
| awe_ | ogra_, by the way, for the syslog problem mentioned earlier. The fix is just to ensure that /var/log is owned by group="syslog", correct? | 19:32 |
| === dandrader|afk is now known as dandrader | ||
| ogra_ | awe_, yes | 19:32 |
| cyphermox | pmcgowan: can we get stats about that? | 19:32 |
| awe_ | ogra_, thanks | 19:32 |
| pmcgowan | cyphermox, I can try, its usually anecdotal | 19:32 |
| * popey looks at his mediascanner logs to see when this first happened | 19:32 | |
| cyphermox | yeah :/ | 19:32 |
| popey | grrr, developers who don't put date/timestamps in logs grind my gears :) | 19:34 |
| ogra_ | popey, when you copied the nasty movie onto your phone ... it is the new parental-control in mediascanner, have your partents type in the right code and you can use it again ;) | 19:34 |
| popey | ok dad | 19:34 |
| awe_ | pmcgowan, I thought we'd only added the watchdog for system services, but I see there's a session version too | 19:35 |
| popey | my first mediascanner crash was 5am yesterday. | 19:36 |
| popey | been crashing every 3 minutes since | 19:36 |
| pmcgowan | awe_, is there two? I only see one get installed | 19:36 |
| ogra_ | did you use it at 5am ? | 19:36 |
| popey | me? 5am? are you serious? | 19:36 |
| ogra_ | lol | 19:37 |
| mcphail | ogra_: that's when his phone downloads the Ubuntu-UK podcast | 19:37 |
| ogra_ | well, i wonder how it can happen out of the blue | 19:37 |
| popey | indeed | 19:37 |
| ogra_ | ah | 19:37 |
| popey | he's joking :) | 19:37 |
| ogra_ | so these guys spread malicious files !! | 19:37 |
| ogra_ | tainted oggs | 19:37 |
| ogra_ | popey, disk full ? (just to state the obvious) | 19:37 |
| popey | good call | 19:38 |
| popey | quite full | 19:38 |
| popey | [M#pC/dev/mmcblk1p1 30G 20G 9.6G 68% /media/phablet/2541-1C26 | 19:38 |
| popey | [M#e+/dev/mmcblk0p7 4.1G 3.8G 82M 98% /home | 19:38 |
| ogra_ | hmm | 19:38 |
| popey | there have been log rotations, so could have blown up at that point, for sure | 19:38 |
| popey | would disk full be reported somewhere else? | 19:38 |
| ogra_ | yeah, 82M isnt much | 19:39 |
| ogra_ | in the UI | 19:39 |
| popey | no, i mean in a log | 19:39 |
| ogra_ | ciborium ships a popup message | 19:39 |
| ogra_ | and that should have kicked in at 95% i think | 19:39 |
| popey | which I wouldn't see | 19:39 |
| ogra_ | well | 19:39 |
| awe_ | pmcgowan, there's session-watchdog and system-watchdog | 19:39 |
| ogra_ | not sure ... its the same UI as the low battery warning | 19:40 |
| ogra_ | does that one time out ? | 19:40 |
| pmcgowan | awe_, I dont see the system one gets installed, where is that one | 19:41 |
| ogra_ | /etc/init/system-watchdog.conf | 19:41 |
| awe_ | check /etc/init, where all the system jobs live | 19:41 |
| awe_ | damn you're quick ogra_ | 19:41 |
| pmcgowan | there is a bug on the full disk warning too I think you can miss it | 19:41 |
| ogra_ | i have a terminal to the phone open and looked for it before ;) | 19:42 |
| ogra_ | pmcgowan, ah, popey probably has then | 19:42 |
| ogra_ | iirc the threshold was at 95% | 19:42 |
| popey | my apport log is entertaining http://termbin.com/3n6h | 19:42 |
| pmcgowan | popey, so is it related to a data file then? | 19:43 |
| popey | I still don't know really. | 19:45 |
| popey | I mean, it's plausible my disk is full, and some random process (in this case mediascanner) dies | 19:45 |
| popey | I'll delete some crap and see if mediascanner still crashes | 19:45 |
| ogra_ | it might try to index something and the db gets to big | 19:45 |
| ogra_ | or might have tried and now you have a corrupt db | 19:46 |
| ogra_ | or some such | 19:46 |
| pmcgowan | sure | 19:46 |
| ogra_ | files from /var/crash get deleted by cron .. so it is entirely possible the disk was at 100% at some point in the last days | 19:46 |
| pmcgowan | yeah 80 MB is certainly too close | 19:49 |
| popey | nah, loads of disk space free and ms is still crashing and still getting a reboot loop | 19:50 |
| popey | gonna copy the ms database off the phone, delete it and reboot | 19:50 |
| popey | so it has a fresh db, maybe it's corrupt | 19:50 |
| ogra_ | yeah | 19:51 |
| ogra_ | i think corrupt db is more likely | 19:51 |
| ogra_ | caused by full disk perhaps :) | 19:51 |
| popey | yup | 19:52 |
| popey | ok, moved the db, mediascanner lives, phone doesn't reboot | 19:53 |
| popey | so, probable cause, disk full, ms crash, phone reboots | 19:53 |
| popey | ms should detect a duff database and delete it :) | 19:53 |
| pmcgowan | popey, sounds like another bug with that bad db | 19:53 |
| popey | yeah | 19:54 |
| pmcgowan | popey, all time record, 5 bugs for one issue | 19:54 |
| pmcgowan | :) | 19:54 |
| popey | haha | 19:54 |
| ogra_ | yay | 19:54 |
| popey | 3.2MB is quite a size for my mediascanner db! | 19:56 |
| pmcgowan | popey, so did we really decide if turningoff apport made any difference? it shouldnt given it skips the crash file one one exists | 19:58 |
| pmcgowan | once | 19:58 |
| ogra_ | well, it makes watchdog function properly | 19:58 |
| ogra_ | :P | 19:59 |
| pmcgowan | ogra_, which bug is that then | 19:59 |
| popey | well the difference for my situation was with apport on, there was a much longer time between reboots | 19:59 |
| pmcgowan | even to do nothing? | 19:59 |
| pmcgowan | hmm | 19:59 |
| popey | with it on, it writes out to /var/crash | 19:59 |
| pmcgowan | popey, not if there is already a crash there | 19:59 |
| popey | so it's not doing nothing | 19:59 |
| popey | hm | 20:00 |
| pmcgowan | thats waht your log is full of | 20:00 |
| popey | it still sat there and chugged | 20:00 |
| popey | i was watching top and apport was right up there with upstart and mediascanner | 20:00 |
| pmcgowan | hmm then I am unclear what it does | 20:00 |
| pmcgowan | crash already exists and unseen, doing nothing to avoid disk usage | 20:01 |
| guest42315 | popey, https://bugs.launchpad.net/canonical-devices-system-image/+bug/1498169 | 20:07 |
| ubot5 | Ubuntu bug 1498169 in Canonical System Image "some logs in .cache/upstart are missing date/timestamps" [Undecided,New] | 20:07 |
| popey | guest42315: ta :) | 20:08 |
| kd__ | i am looking to install ubuntu on handheld device | 20:08 |
| guest42315 | o/ | 20:09 |
| kd__ | how should i prepate my desktop for it? | 20:09 |
| kd__ | can someone please help | 20:09 |
| dobey | kd__: you're porting to a new device? or you're flashing onto a supported device? | 20:10 |
| popey | kd__: what device? | 20:10 |
| kd__ | i have old phone | 20:11 |
| kd__ | which is now being used | 20:11 |
| kd__ | i want to try on it | 20:11 |
| kd__ | its on android 2.2 | 20:12 |
| kd__ | *its not being used | 20:12 |
| popey | Ubuntu won't run on that. | 20:12 |
| popey | Is it a Nexus One by any chance? | 20:12 |
| kd__ | ohh is it... whats requirement for ubuntu on handheld? | 20:12 |
| kd__ | it is samsung galaxy some model i dont remember | 20:13 |
| popey | !devices | kd__ | 20:14 |
| ubot5 | kd__: You can find the full list of devices, official images, community images, and works in progress at https://wiki.ubuntu.com/Touch/Devices | 20:14 |
| atk33 | hello, there is a way to install ubuntu touch on the meizu mx4 international? | 20:46 |
| dobey | atk33: if it came with android i think you need to get the tools from meizu to be able to repartition it and allow flashing it with ubuntu-device-flash | 20:59 |
| atk33 | yes is coming from meizu with flyos | 21:00 |
| atk33 | but the bootloader is locked apparently | 21:00 |
| dobey | yes, you'll need to ask meizu support for how to unlock it, i think | 21:02 |
| jason__ | Is there a way to check what version of an app is on my phone from the phone UI itself? Or is that something that requires ssh-ing into the phone, or some other process? | 21:30 |
| jason__ | I'm trying to report some bugs, and I want to figure out what version of different things I'm running | 21:30 |
| dobey | jason__: if you install the terminal app you can run "click list" in it | 21:35 |
| jason__ | Ah, that brings me to my next question. I can't find the Terminal in the Ubuntu App Store. I search "Terminal" and don't see anything relevant. Has anyone run into this before? | 21:36 |
| jason__ | thanks btw dobey | 21:36 |
| dobey | no, the terminal is definitely in the store | 21:41 |
| jason__ | hmm, I've tried a few times. I'll give it another shot I guess. | 21:47 |
| jason__ | Tried again and it was there. Not sure what my original issue was. But thanks again for the pointer on "click list". | 21:51 |
| popey | or just long press on an app | 21:51 |
| popey | it tells you the version number | 21:51 |
| mcphail | Is there any way to get a changelog from the "Updates" screen yet? | 21:53 |
| popey | no | 21:57 |
| mcphail | Shame. Would be a nice addition | 21:57 |
| === curtness is now known as strahtw | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!