[00:00] Elleo: the question was whether I should include modified copies of the code that generates the database or if I should just include the original text [00:00] (I noticed that the greek layout only included the original text in the MR [00:00] * trickvi_ goes to learn how to push this to launchpad and make an MR [00:02] trickvi_: cool, send me the link when you're done and I'll review it tomorrow [00:02] trickvi_: if you haven't already could you sign the CLA? http://www.ubuntu.com/legal/contributors [00:14] Elleo: done (except I noticed my email was wrong on the commits) [00:15] Elleo: I put you as the reviewer so you should have gotten an email [00:20] trickvi_: great, thanks; I'll include it with the others tomorrow [00:27] Elleo: awesome thanks, I'm resubmitting it to update my email === chriadam|away is now known as chriadam === aaron is now known as ahoneybun === tmpRAOF is now known as RAOF [02:15] Hi, I've got a hold of a Meizu M1 note, after googling a little i found that it's a close cousin of the one that runs ubuntu. [02:15] Obviously i came here. [02:17] Has anyone had any expirience with those? any tips, linke etc, would be much appreciated. [02:17] of maybe i should just flash it with cyanogen and come back in a few months when other people bricked their phones for me? [02:30] anyonebutme: the ubuntu one isn't even public yet [02:31] ok... so i should be patient? [02:32] is there a way to try it with less then 90% chance of bricking anything? i'm looking at the debian pages too (https://wiki.debian.org/Mobile) this looks even more theoretical then ubuntu atm. === danilos` is now known as danilos [04:14] morphis: did you have to apply any extra patch to that tree? if not I guess we can try to update the tree for mako/flo so we can at least test the userspace [06:30] good morning === VDVsx_ is now known as VDVsx [07:38] jamesh: FYI the QQuickAsyncImageProvider is now in the overlay PPA so you can start using it for a landing too [07:38] Mirv: I noticed. Thanks! [07:39] you're welcome! [08:23] hmm...my Bq 4.5 has suddenly started to switch itself off when there's about 29% of battery left and it won't boot unless I charge it for a while [08:24] either the battery status is horribly off or this is somekind of a hardware failure.. [08:53] juzzlin_: i think you're right, the battery gauge is inaccurate. I have seen similar this weekend [09:06] Good morning all; happy Monday, and happy Twilight Zone Day! 😃 === davidcalle_ is now known as davidcalle === chriadam is now known as chriadam|away [10:05] sergiusens, rsalveti: do you know who could take a look at https://code.launchpad.net/~phablet-team/urfkill/monitor-script/+merge/257592? [10:07] so far, i've had no feedback on http://paste.ubuntu.com/10696198/ [10:14] DonkeyHotei: what you expecting? [10:14] if anyone can say whether it gives any clues as to why there is no video playback [10:15] maybe poke jhodapp when he wakes [10:15] he's our expert in this area I think. [10:15] well, did you follow the resent ML discussions about video playback in vivid ? [10:26] been a couple weeks since i looked at the ml [10:46] ogra_: i can't find a relevant thread; what was the title? === MacSlow is now known as MacSlow|lunch [11:12] alecu, hey, you up already? [11:13] please ping me when you have some time to help me finding an issue with the app store scope [11:13] Argh, some guy managed to flash Android to his Aquaris E4.5 Ubuntu Edition using the Android firmware tools supplied by bq for the Android version -> https://askubuntu.com/questions/621971/ [11:13] mzanetti: pong! [11:14] sturmflut2, isn't he allowed? [11:14] alecu, hey ho. so... my wife's BQ keeps on hiding the review part for apps [11:15] sergiusens, hey, could you take a look at the email? Thanks [11:15] she only can write reviews for like 5 mins after rebooting, then the input box disappears from the store [11:15] alecu, ^ [11:15] mzanetti: yup, we've seen something like that. Let me find the bug... [11:17] jgdx: Technically yes, but in reality no. He used a special tool supplied by the SoC vendor to reconfigure the whole internal structure of his phone, it will no longer run Ubuntu in this state and bq hasn't released the necessary configurations to change everything back for Ubuntu. [11:18] jgdx: If they had, it would have been possible to buy the Android version of the E4.5 and flash Ubuntu on it. [11:18] sturmflut2, hmm, i thought they had flashtool and images for that on their page ... [11:18] john-mcaleely might know where to find that [11:19] ogra_: For the Ubuntu Edition they only have the Manuals, no flashtool. At least AFAICS. [11:19] sturmflut2: other people have done that === dandrader is now known as dandrader|afk [11:19] right [11:19] sturmflut2: "studio_" who hangs around here asked them directly for the tool I believe. [11:20] sturmflut2, ogra_ I understand they are still working on publishing them [11:20] sturmflut2, how do you knoe from the askubuntu text that he did that btw ... [11:20] mzanetti: hmm... I can't find the bug, but I'm pretty sure I saw that bug in the week before the sprint [11:20] mzanetti: I'll ask dobey when he starts his day. [11:20] ogra_: He told me on G+ [11:20] ah [11:20] good day, any idea when 15.04 update will reach bqphone? [11:20] well, there is definitely a way to get the tool and image ... [11:20] bqphone: some weeks. [11:21] it is some weeks ago :P [11:21] ogra_: I'll refer him to the bq customer support then? [11:22] sturmflut2, oh, makes sense [11:22] yeah, i assume thats the best you can do [11:22] bqphone: its not ready yet [11:23] lets hope it comes soon, thx for the answer [11:23] bqphone, well, better hope it comes stable :P [11:23] mzanetti: I think it's this: https://bugs.launchpad.net/ubuntu/+source/unity-scope-click/+bug/1421710 [11:23] Ubuntu bug 1421710 in unity-scope-click (Ubuntu) "Rating entry form does not appear sometimes" [High,Triaged] [11:23] ogra_ i do :) [11:24] * ogra_ prefers stable over soon ;) [11:24] another question, any idea when telegram will support audio messages? [11:24] alecu, do you need any logs or something? [11:41] Hi all! Is there any news about today's OTA-3.5 update? [11:49] mzanetti: no need for logs. I can still reproduce it. I know pstolowski has been working on a some changes to the review code of the click scope to support editing of reviews, so hopefully this will be fixed by it. [11:50] alecu, ack. thanks a lot. === dandrader|afk is now known as dandrader [11:51] mardy, hey, is it intentional that OA deletes all accounts when I reinstall the click package? [11:55] alecu, I just got review editing working :) [11:55] brb [11:59] mzanetti just curious: are you possibly referring to the U1 account being deleted? [12:00] matv1, no, in my case the evernote account [12:01] oh ok. a different issue then. thnks === alan_g is now known as alan_g|lunch === MacSlow|lunch is now known as MacSlow [12:29] mzanetti: Ping [12:29] sturmflut2, hey [12:30] mzanetti: I'm giving the BayTrail tablet a try again. Do I need any PPAs on top of 15.04 to get the most recent Unity8? Or has development switched to Wily already? [12:31] sturmflut2, you want vivid + ppa:ci-train-ppa-service/stable-phone-overlay for now [12:31] alecu, mzanetti: yes, that's the bug [12:31] what's the systemd equivalent for initctl --set-env ? If any [12:31] sturmflut2, although we will probably move to wily soon [12:32] s/--set-env/set-env [12:33] dobey, my wife's bq seems to reproduce that all the time. if I can help with logs or something, let me know [12:33] mzanetti: Okay great, thanks. Let's hope I even get to that point and the internal flash doesn't act up again :/ [12:33] heh [12:35] mzanetti: i can as well, on my nexus4, but i still can't tell *why* it's happening exactly. i don't expect that pstolowski's current work will resolve it though. [12:36] I never saw it on any of my devices, happens 100% for her... really odd [12:44] is there a way to reset the music app, so it reads everything from scratch? [12:44] mzanetti: yeah, it's weird. if it's not the preview immediately after submitting the review, and the package is removable, then the widget should be shown. [12:45] NIN101: yeah, you can get rid of the mediascanner database and music app cache [12:45] mzanetti: Okay, the tablet crashed again while accessing the internal storage. Hooray. [12:45] willcooke: Which Intel tablet do you use for your demonstrations? [12:46] sturmflut2, erm.... [12:46] * willcooke thinks [12:46] sturmflut2, bear with me, I'll find it [12:46] Haha [12:47] * sturmflut2 can see willcooke running around in his appartment, digging through heaps of mobile devices [12:47] sturmflut2, found the order in email :) [12:47] NIN101: http://paste.ubuntu.com/11078767/ something like that [12:48] mzanetti: definitely not, it's not normal; which click package? [12:48] sturmflut2, Acer Aspire P3-171 Core i5 4GB 120GB SSD 11.6 inch Convertible Folding Keyboard Ultrabook [12:48] NIN101: then reboot :) [12:48] mzanetti: ah, is it an account plugin? if so, yes, it's normal [12:48] popey: ok, will try. thanks [12:48] sturmflut2, I bought a reconditioned one: http://www.acerdirect.co.uk/Acer_Aspire_P3-171_Core_i5_4GB_120GB_SSD_11.6_inch_Convertible_Folding_Keyb_a1-NX.M8NEK.001/version.asp === _salem is now known as salem_ [12:49] willcooke: was it a faff to install on? [12:49] willcooke: Aaaaah, you have a Core i5! I assumed it was BayTrail and always wondered how you got it to work, because BayTrail is a horrible mess [12:51] popey, sturmflut2 - it was easy to install.. Switched the BIOS to legacy mode and booted of a USB stick [12:52] k === alan_g|lunch is now known as alan_g [13:13] dholbach: liuxg: sure i'm here [13:13] ok cool [13:13] liuxg: best would be if you ask your question here again [13:14] so yeah... I don't know which paths can be accessed by apps, I'd say that content-hub is the way to exchange docs between apps [13:14] willcooke: out of interest why did you need to switch to legacy mode? [13:14] dholbach, zbenjamin, I am here as well. a developer is developing a media player for media files. he wants to make use of the "Documents" to save all of the files there. [13:14] dholbach: but doesn't the contenthub need at least one app that is registered for that type of document? [13:14] liuxg: didn't you say a pdf reader? [13:14] davmor2, Not sure that I did, and indeed probably didn't need to - I just did. :) [13:15] zbenjamin, if fact, it is the same for the two apps. [13:15] liuxg: yeah because Documents is not the right path for media [13:15] willcooke: fair enough I just wondered if it was strictly necessary :) so now we'll never know :D [13:16] zbenjamin, he wants to use MTP to copy over files so that the users may start to play the files directly. [13:16] . [13:16] davmor2, if I get time I'll try a normal install [13:17] zbenjamin, dholbach, packaging all of the files are not a doable solution. [13:17] jdstrand: kenvandine: could you help with liuxg's request? He needs a way to access files for a) a pfd reader and b) a media player [13:17] willcooke: nice [13:18] dobey, my review editing changes have nothing to do with that problem [13:18] pstolowski: i know [13:18] zbenjamin, jdstrand kenvandine, for android apps, it seems that the app can search for the files and make use of them. Do we have some "PUBLIC" directories which are accessible to all of the apps? [13:19] hi, could anyone help me?, i have been searching how to define packages dependencies, i.e libpurple-dev, in app installation process but i found nothing, i can define this package dependencies in app creation using .pro file (as in gradle or maven) or i must do this from the c++ application code, thank you [13:19] pstolowski: though looking at the code again, depending on what changes you've made, i suppose it's possible that your changes could result in the problem just magically going away [13:20] liuxg, which directories are you looking for? [13:20] frobs: click packages do not have dependencies [13:20] app confinement limits what apps have access to, which is why we have content-hub [13:21] kenvandine, the developer is developing a PDF reader, and another developer is developing a media players. May I know whether our phone has any directories are accessible to the apps without using contenthub? [13:21] frobs: if packages are not part of the SDK, then you basically need to include additional libraries in your package and build them along with your app, with appropriate build configuration for how they are installed and the confinement rules your app must follow [13:22] frobs: libpurple will be problematic without changes [13:22] kenvandine, how about the "http://bazaar.launchpad.net/~ubuntu-docviewer-dev/ubuntu-docviewer-app/trunk/view/head:/docviewer.apparmor" docviewer app? it uses "read_path" and "write_path", is the app a special app? [13:22] liuxg, i not sure, but i know the docviewer uses content-hub now [13:23] kenvandine, I used to work on Symbian, it had some public directories which are accessible to other apps. But each app cannot access other's private directories for sure. [13:24] kenvandine, looking at the .apparmor there, it has the fields. does it violate the security? it is a special case, and we should never promote it to the developers? [13:24] dobey: then....i can't install a package from repositories and use it in my app???for example if i want use libpurple i need include in my app the headers and sources of libpurple and call them from c++? [13:24] frobs: no [13:24] we certainly shouldn't promote it to developers [13:25] frobs: libpurple will need patching and you will need to build it in your app, if you want to use it, yes [13:25] jdstrand, can you comment? is docviewer allowed as an exception? [13:25] frobs: if you are trying to write an IM application, the best thing to do would probably be to work with the messaging-app devs and design team to get it working in there with the telepathy back-ends which already exist on the phone [13:27] kenvandine, is there any C++ APIs for the contenthub? the developer's backend is C++ instead of QML. He is currently developing a common architecture across android, ubuntu and iOS. [13:28] liuxg, there is a c++ api, but last i heard we didn't want to encourage app devs to use it [13:28] dobey: the idea is make a simple IM interfaz and use libpurple and yowsup, and write a whatsapp client [13:28] not part of our documented API in the platform [13:28] and much harder to maintain compatibility [13:28] kenvandine, oh. yeah, that is my understanding. yes, I can only see the Qml API for it. [13:29] mzanetti: I enabled all the quirks for the flash storage, now the Ubuntu 15.04 installer crashed while installing GRUB O_o [13:29] liuxg, perhaps at some point we'll decide to support a c++ API [13:29] but not sure [13:30] frobs: using any client other than the official whatsapp app will result in that client being blocked, and any users of the software getting banned. whatsapp have been quite obvious about that in their actions so far [13:30] Not WhatsApp again [13:30] kenvandine, so, in your understanding, there is no sort of PUBLIC diretories for third party apps? [13:30] frobs: if you really want whatsapp on ubuntu, your best options is to be vocal about it to whatsapp to port it to ubuntu, and get all your friends who use it to be vocal to whatsapp as well [13:30] frobs: in other words, show them that they will have a userbase on ubuntu if they port their app [13:30] kenvandine, all of the directories are private and owned by a certain app? [13:31] liuxg: yes, applciations are confined [13:31] liuxg, nope, and app confinement is a strong benefit provided by our platform [13:31] kenvandine, from end user point of view, it is easy for them to copy over the files to the phone via USB cable. [13:31] but it can make things more work for developers [13:32] so that use case we need to support [13:32] but nothing yet [13:32] we need content-hub integration for mtp, sdcard, etc [13:32] kenvandine, yeah, this is my understanding too. The docviewer makes me a little confused :) anyway, I think it is an exception. [13:33] liuxg, i'm not convinced what it's doing even works [13:34] kenvandine, is that a security beak? [13:35] jdstrand, does that really work? [13:35] what docviewer does? [13:35] what does docviewer do? [13:35] is it just "unconfined" ? [13:35] "http://bazaar.launchpad.net/~ubuntu-docviewer-dev/ubuntu-docviewer-app/trunk/view/head:/docviewer.apparmor" [13:35] it adds read_path and write_path in the profile [13:36] oh [13:36] hmm [13:36] but i didn't think apps could do that there, they have to use the policy groups [13:36] dobey: that's probably the smartest choice...but....i am developer :P if a company don't want contribute with my free os i will do for them, I'm not making them lose money, when they do a client for ubuntu phone i remove my app from the store..but...while... [13:36] kenvandine, jdstrand, even it can access it, but "Documents" directory is only accessible by an app, and no other apps can access it unless it is a system app. [13:37] dobey, it is not an unconfined app. it yes, there is no need to define read_path there [13:37] we have a special case for music with ~/Music and gallery with ~/Pictures [13:37] frobs: well, they will ban your app and anyone using it. you will actually be doing a disservice to users by telling them to use it [13:38] hi, i use usb keyboard on ubuntu touch. and I wanto swap caps to ctrl. how can I swap these key ? [13:39] I found keymap file on /android/sytem. but this directory is read-only ... [13:39] i don't know if that's possible [13:40] unless maybe there's some way to do it with udev config or something [13:41] uhm. thanks. i will search 'udev' keyword. [13:42] dobey: I will not supplicate to a company ... I do nothing better, thank you for the information :) [13:43] frobs: better yet, get people to switch to telegram :) [13:49] kenvandine, if we use contenthub, normally, it copies over the files, right? [13:51] dobey, if we use content hub to select files, normally the files are copied over to the private directories, right? === elopio_ is now known as elopio [13:52] * jdstrand reads backscroll [13:52] liuxg: that's a question for kenvandine [13:53] dobey, jdstrand, kenvandine, if the file is copied over, for the case a big video, basically, it duplicate the space. [13:53] liuxg, usually not [13:53] dholbach: I will take a look at the urfkill MR later today [13:53] it's a link [13:53] dholbach: need to upload that to wily and vivid (overlay ppa) [13:53] if the link fails, then it's a copy [13:54] which should only happen across file systems [13:54] so rare [13:54] kenvandine, I once used contenthub to get a picture, I got a local copy of the file. [13:54] are you sure it wasn't a hardlink? [13:54] it would have only been a copy if it crossed between file systems [13:55] kenvandine, I just found that there was a file there. the was long time ago. I need to double check this. [13:55] i bet it was a hardlink [13:55] rsalveti, thanks a bunch! [13:55] kenvandine, what do you mean by crossing between file systems [13:55] i test that often [13:56] if the content is transfered from a different file system [13:56] liuxg: opening something from an SD card for example [13:56] kenvandine, what if you use a gallery image as bg and then delete it from the image collection? [13:56] kenvandine: why doesn't it use soft links instead? [13:56] it deletes the link [13:56] liuxg, kenvandine: as mentioned, we have special cases for the music-app (~/Music), gallery (~/Pictures) and, yes, document viewer (~/Documents) [13:56] kenvandine, dobey , oh, if we select a file from SD card, it will make a copy of it? [13:56] kenvandine: yes, it does 'work' from a technological POV [13:56] liuxg, yes [13:56] liuxg: and no, we should not promote that to developers [13:57] this is what content-hub is for [13:57] kenvandine, what happens if the video file is very big, duplicating it will waste a lot of space for it. [13:57] dobey, i think a hardlink, apparmor reasons [13:57] popey: Is it April 1st again or is this ChillHub thing real [13:57] ya [13:57] liuxg, eventually we'd want the app to have access to it's own space on the sd card [13:57] so it doesn't have to copy it [13:58] that said, music-app in particular has other exceptions and we don't have a good way to allow a 3rd party music player (well, it can access its siloed data of course, but not ~/Music for example) === marcusto_ is now known as marcustomlinson [13:58] kenvandine, ok. so, eventually, only hardlinks, right? [13:58] yes [13:58] well it'll be supported, depending on what the app wants to do [13:59] like often the app decides where to put the imported content [13:59] if it needs to persist [13:59] kenvandine, currently, between different systems, we copy files, right? [14:00] liuxg, yes [14:00] jdstrand, so, there is no PUBLIC directories accessible to any other apps, right? all directories are private? [14:00] liuxg: yes, by design. it was a security nightmare on android and they've been trying to move away for it for years [14:00] s/for/from/ [14:01] Mirv, hi! Another Checkbox-Touch release done :) [14:01] liuxg, in our model all content is owned by an app, it isn't about public vs. private really [14:01] popey: Can we do anything to prevent the logical next step, the Ubuntu Toilet? [14:01] just the content belongs to an app, and you have to use that app to gain access to it [14:01] that's a better way to put it [14:02] apps own their data [14:02] where do i open the bug for http://i.imgur.com/SDToNJK.png ? [14:02] and user driven interactions permit data sharing [14:02] which app owns the "Music" and "Documents" for example [14:02] Mirv, or should I bother popey instead :) [14:03] those aren't great examples, because we've allowed them to use the old paradigm [14:03] tsdgeos: that's on current vivid-proposed? [14:03] dobey: it's what i got this morning yeah [14:03] account-plugin-ubuntuone it seems [14:03] liuxg: those are imperfect examples, but basically, music-app owns ~/Music and the doc viewer ~/Documents [14:03] liuxg: those are special cases [14:03] tsdgeos: ubuntuone-credentials is the source package, but it's probably a toolkit bug [14:04] dobey: not really [14:04] jdstrand, then Docviewer does not need to use "read_path" and "write_path", right? [14:04] dobey: i just had a look at the code and is clearly a bug in the code [14:04] tsdgeos: how is it a bug in the code? [14:04] jdstrand, this concept is a little bit difficult to developers. I will try to explain this to the developers. [14:04] tsdgeos: what device is that on? [14:04] kissiel: wassup? [14:04] liuxg: it does, because the apparmor template policy doesn't allow that exception [14:05] dobey: the label is anchored to the left, not to the right, so it will just take as much width as needed [14:05] popey, another release of checkbox-touch [14:05] super [14:05] popey, url to click: https://launchpad.net/checkbox-touch/trunk/1.1.3/+download/com.ubuntu.checkbox_1.1.3_multi.click notes: http://paste.ubuntu.com/11079664/ [14:05] liuxg: it is much easier to not hold up the exception as the example for others :) [14:05] tsdgeos: hmm, ok; well ubuntuone-credentials is the package to file the bug against [14:05] jdstrand, yes, I agree with you [14:06] liuxg: there is quite a bit of documentation on all of this: [14:06] https://developer.ubuntu.com/en/publish/security-policy-groups/ [14:06] https://developer.ubuntu.com/en/start/platform/guides/content-hub-guide/ [14:06] https://developer.ubuntu.com/en/start/platform/guides/app-confinement/ [14:06] jdstrand, yes, I have read all of them :) [14:07] sure, I meant for the other guys [14:07] jdstrand, it would be good to have an online training for this. a video would be very good for the purpose. [14:08] dpm: fyi ^ [14:08] jdstrand, especially, a lot of new developers are very hard to follow it. they need to change their mind to take our ways to handle something which is different on other platforms. [14:09] dobey: https://bugs.launchpad.net/ubuntu/+source/ubuntuone-credentials/+bug/1453830 thanks :) [14:09] Ubuntu bug 1453830 in ubuntuone-credentials (Ubuntu) "Sign in to ubuntu one text runs outside the screen" [Undecided,New] [14:09] liuxg: it isn't drastically different from Apple [14:10] it is different from android, but android wasn't designed with these things in mind [14:10] jdstrand, yes, that is the way I tell developers so that it can be easily understood. [14:10] well, I would point them at https://developer.ubuntu.com/en/start/platform/guides/app-confinement/ [14:11] if they read the trust model section, everything else falls into place [14:11] jdstrand, we are facing some desktop and android developers. we need to highlight this to them. [14:11] oh yes, this is very different from traditional linux desktop [14:11] (by design) [14:12] hi [14:12] but, like I said, a few key concepts like our trust model should really go a long way on why we do things and how to work with the system [14:12] jdstrand, in fact, I designed a sample to output all of the variables at http://blog.csdn.net/ubuntutouch/article/details/41281971 [14:13] kissiel: done [14:13] popey, tyvm! :) [14:13] liuxg: you know, our hello-world click should do that. in snappy we have that. after installing on a snappy system, you run 'hello-world.env' and can see everything [14:14] liuxg: or just upload your app :) [14:14] jdstrand, it is interesting. I do not know that. yeah, you can upload my app for sure. It is there bzr branch lp:~liu-xiao-guo/debiantrial/runtimeevn [14:15] liuxg: I'll let you upload it if you want-- it is quite satisfying getting one's app in the store (it is quite easy) [14:16] jdstrand, ok, I will try to do that for sure :) [14:16] cool [14:17] jdstrand, I will let you know it. [14:17] popey, if someone is asking again for the flash-tools for the bq e4.5, they can find them on http://www.bq.com/gb/products/aquaris-e4-5.html [14:17] studio_: thanks === dandrader is now known as dandrader|lunch [14:46] popey, if the peoples want to flash back, they can download the firmware from her: http://www.ubuntugear.com/2015/02/first-ubuntu-phone-official-firmware.html or use their own scatter.txt after they built the kernel from the git. [14:49] hi! Is there anyi info about today's OTA update? sil2100 [14:51] robin-hero: it's not today, it was monday/tuesday and that is for feedback from bq then we release and it will be staggered updates again [14:52] robin-hero: hey! Still waiting for feedback :) [14:52] davmor2: sil2100: Thanks. :) I'm really looking for it. :) === dholbach_ is now known as dholbach [15:17] hi Benno-007 :) === dandrader|lunch is now known as dandrader [16:00] * kissiel is away: gym === chihchun is now known as chihchun_afk [16:17] !seen jhodapp [16:17] I have no seen command [16:18] DonkeyHotei, what's up? [16:18] [Mon 2015-05-11 03:15:04 AM PDT] maybe poke jhodapp when he wakes [16:18] [Mon 2015-05-11 03:15:10 AM PDT] he's our expert in this area I think. [16:19] DonkeyHotei, have a specific issue? [16:19] video playback on hammerhead [16:19] DonkeyHotei, it's the nexus5 right? [16:19] yes [16:20] this happens: http://paste.ubuntu.com/10696198/ [16:20] ogra_ said it was mentioned on the mailing list but i did not see that [16:21] DonkeyHotei, what codec are you trying? [16:21] for video [16:21] just a sec [16:23] h264 [16:23] DonkeyHotei, it looks like it's not configuring the hardware codec properly [16:23] exactly where is the bad code? [16:24] DonkeyHotei, I don't have a N5 so of course this is hard for me to reproduce [16:25] i can reproduce it at will [16:25] DonkeyHotei, well tough to say, but it would most likely be either in our code that lives on the Android side that calls MediaCodec::configure() or the gstamchybris plugin [16:25] DonkeyHotei, gstamchybris code: https://github.com/jhodapp/gst-plugins-bad [16:26] DonkeyHotei, take a look in sys/androidmedia [16:26] considering the latter works on other hardware, you think the former is more likely? [16:26] DonkeyHotei, other hardware issues have usually been in gstamchybris [16:27] DonkeyHotei, things like making sure the right codec config data is passed in or making sure the right number of input buffers or their sizes are used, etc [16:31] jhodapp: where exactly in plugins-bad is gstamchybris? [16:32] sys/androidmedia [16:36] DonkeyHotei, let me know if I can point things out, have to run for now [16:36] jhodapp: how do i play from the cmdline so i can see the gst msgs? [16:36] DonkeyHotei, bzr branch lp:media-hub and check out the README in there [16:36] DonkeyHotei, make sure to "stop media-hub" first [16:37] DonkeyHotei, then add GST_DEBUG=*:3 or whatever else you want to that line from the README to get the gstreamer messages [16:37] * jhodapp bbiab [16:48] 0:00:17.809332025 13802 0xb2528100 WARN basesrc gstbasesrc.c:3470:gst_base_src_start_complete: pad not activated yet [16:48] 0:00:17.828363015 13802 0xb2e02d20 WARN qtdemux qtdemux_types.c:202:qtdemux_type_get: unknown QuickTime node type iods [16:48] 0:00:20.919490356 13802 0xb241f100 FIXME bin gstbin.c:4023:gst_bin_query: implement duration caching in GstBin again [16:48] 0:02:01.026177819 13802 0xb241f100 WARN playbin gstplaybin2.c:5426:setup_next_source: error: No URI set [16:49] bool gstreamer::Playbin::set_state_and_wait(GstState): requested state change. [16:49] jhodapp: ^ [17:04] awe: ping === alan_g is now known as alan_g|EOD [17:15] peat-psuwit, how can I help you? [17:20] awe: I'm thinking of the way to implement multi-sim on my device in ofono. (You remembed qcom-msim plugin, right?) [17:21] Could you give me some hint about where should I do that? [17:21] yes, I remember. We merged your changes, but haven't yet released a new ofono to our overlay PPA yet. That should happen later this month [17:21] peat-psuwit, it really depends on how your device implemented dual-SIM [17:22] my suggestion would be to look at the mtk plugin, and associated mtkmodem driver [17:22] there are many device-specific assumptions made in that code [17:22] ( ie. the SIM can't be accessed with the radio of ) [17:23] is your device the same as BQ's dual-standby? [17:23] can it have a phone call using one SIM, and data call using the other SIM? [17:24] also, the big question is how does it expose dual-SIM from ril? [17:24] is everything done via a single socket, or are there two sockets available like BQ? [17:24] awe: At least, there is 2 ril socket in /dev/socket [17:25] ok, that's good news [17:25] I would suggest starting a wiki page, or a google doc, and start documenting some of this [17:26] and honestly, as the BQ phone is our only dual-SIM implementation to date, it's probably the best place to start [17:26] you already had to create a device plugin, and new driver type [17:26] so you've got some of the basic structure already in place [17:28] awe: Using Android to observe it's behavior? [17:32] DonkeyHotei, that's all benign stuff [17:32] peat-psuwit, yes.. that's how you should be able to confirm messaging flows, quirks, ... [17:33] jhodapp: anything in particular i should look for? [17:34] DonkeyHotei, I'd debug around the call to *configure* [17:35] awe: Thank you for your sugestion! [17:37] your welcome [17:37] awe: By the way, my changes will eventually land in image after they land in overlay PPA, right? [17:38] jhodapp: when is that called? [17:38] DonkeyHotei, right when a video is getting setup to be played [17:38] peat-psuwit, yes [17:39] jhodapp: GST_DEBUG=*:3 isn't enough? [17:39] OK. [17:40] DonkeyHotei, line 130 of http://paste.ubuntu.com/10696198/ is really the thing to figure out why it's failing to initialize/configure [17:40] * peat-psuwit plans to remove quirks from device tarball after that [17:42] jhodapp: i did GST_DEBUG=*:3 and those were the only debug msgs [17:43] DonkeyHotei, http://docs.gstreamer.com/display/GstSDK/Basic+tutorial+11%3A+Debugging+tools [17:56] jhodapp: how do i stop media-hub-server? [17:56] DonkeyHotei, stop media-hub [18:04] jhodapp: http://danielg4.drivehq.com/typescript.xz [18:04] DonkeyHotei, ? [18:04] the debug log [18:05] DonkeyHotei, can you generate that again removing the color from the output? [18:06] DonkeyHotei, google for the flag to remove that, I forget how offhand [18:08]