[00:08] nhaines: of course. I was merely stating that the work is planned and current being designed/implemented [00:40] 8Hello, everybody! If you have anything about ovation, let me know! [00:53] Any word from wesnoth? I'm curious about the ovation image. [01:02] Did anybody look at the links I left last night? The Russian one is the one I want to know about [01:09] nhaines: You available? I have a question - What do you use to download Ubuntu apps under Ubuntu Touch? Is it Software Manager? [01:11] If it is, then the ovation image has a major drawback - I couldn't download anything! [01:12] I mean, it doesn'c have Software Manager. [01:12] doesn't [01:13] UbUntUToUchLover: you use the Ubuntu App Store. It's a Unity lens. [01:13] It works just like the Application lens in Unity 7 except better. [01:14] Hmmm. The ovation image didn't have anything like that. Wesnoth forgot to put that in there. [01:14] It may not have existed at the time. [01:15] If the main interface has a purple background, then you just search right from the apps lens. If it's white, you should have a large orange Ubuntu App Store icon available. [01:15] I was able to use the "apt-get" command under the terminal app, but it develped initramfs-tools errors. [01:17] Yes, the root filesystem is read-only. [01:23] So, how do I modify that? Can I change a read-only root filesystem, or is that only for the Ubuntu developers? [01:28] nhaines: I just read what you said a while ago about backgrounds. I remember that wesnoth used the Saucy Salamander 13.10 arm image, specifically saucy-preinstalled-touch-armhf.zip. According to the link, it directs to an image from October 18th, 2013, but the page doesn't exist. [01:31] UbUntUToUchLover: probably the old one, then. No app store. [01:32] You can make the root filesystem read/write, and it is only for developers. I think there's a special file you touch. I don't remember if there's a phablet-* command for your computer to automate it. It might have changed in 14.04 or 14.10. [01:32] What can I do, then? I'm positive that I can get Ubuntu Touch to work well under my Nook; I'm just not sure how. [01:33] UbUntUToUchLover: if you make the system read/write, you can no longer apply updates and have to reflash again if you want updates. [01:33] UbUntUToUchLover: unless you're a hardware developer there's not really a lot you can do. [01:36] Hmm, what a predicament! Still, I believe I can do it, even though it's still very much in the beta stage. Since some of the Ubuntu Desktop command also work under Ubuntu Touch, is there a command or list of commands I can run from Touch that might fix the initrwamfs-tools problem? [01:37] I meant, of course, initramfs-tools. [01:45] How about changing it from a laptop? I also have Ubuntu Desktop, so if I hook my Nook with Ubuntu Touch to my laptop that has Ubuntu Desktop, can I be able to fix it using a terminal or GUI program [01:45] ? [01:46] You can use it as a CLI-only system but you can never run apt-get upgrade because apt-get can't handle the various paritioning and breaks. [01:46] What about adb? [01:47] That's how you connect to it. But you're still stuck with Ubuntu 13.10 ARM stuff. [01:47] We [01:47] Oops. [01:49] Well, is there another program I can use? Maybe Android has something that I can use to fix the problem. After all, isn't Ubuntu Touch still Android-based? [01:49] Ubuntu Touch has never, ever been Android-based. [01:51] A stripped Android kernel with graphics and sound drivers runs in an LXC container that gets booted after Ubuntu boots. [01:51] UbUntUToUchLover: you can make the image writable by running "phablet-config writable-image" from your desktop (with the phablet-tools package installed) [01:51] UbUntUToUchLover: not 100% certain if it'll work with such an old image though [01:52] UbUntUToUchLover: that command creates an empty file on the filesystem. So if it doesn't work, it's harmless. [01:52] Wow! I didn't know that! Thanks, Elleo! === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === duflu_ is now known as duflu [03:51] Ubuntu Touch Lover. I've switched over to my laptop running Ubuntu Desktop. [04:11] Say, I know that Linux Mint is based on Ubuntu, somewhat. Is it possible to put Linux Mint on a tablet? [04:12] I would ask under the Linux Mint channel, but I'm not getting a response. [04:12] Presumably, if you had an x86 tablet. [04:13] Aww, nuts! I was hoping there was an arm port! [04:13] There *might* be, but then it would have to be ported specifically to each individual ARM platform. [04:16] Yeah, as I figured. However, on the Linux Mint Community website, there's a link to a YouTube video where somebody has it on a tablet, but it's a specific type of tablet, and probably not arm-based. [04:19] Even though Linux Mint is not totally Ubuntu (in fact, I read that they're becoming Debian-only), I thought that anybody here might know if it can be put on a tablet. [04:22] For anyone interested, here's a link, but I'm still a Ubuntu supporter: http://community.linuxmint.com/idea/view/814 [07:52] good morning === chihchun is now known as chihchun_afk [09:21] popey, hey, do you still maintain this regular checkout of all apps in the store? [09:26] oSoMoN: yes. http://popey.mooo.com/mirror/clicks/2014/12/2014-12-04-050001/ [09:29] popey, thanks! [09:34] is there going to be a party when there are 1000 clicks [09:36] Good morning all; happy Cookie Day! :-D [09:58] I ran across an idea: Someone managed to flash Ubuntu Touch to a phablet very similar to a Nexus by editing some of the properties of the build.prop file, thus tricking it to flash the mako image. Even though the Nook is far different from a Nexus, is it possible to do this to the Nook, as well? [10:00] sounds likely to fail to me [10:04] Yes, but figure this: The Nexus 10 is a ten-inch tablet. The Nook HD+ is usually a nine-inch. I figure the only problems may be that the buttons become unresponsive to some apps. [10:05] that build prop thing was most likely the two nexus 7 tablets ... one is called deb, the other is flo and they only differ b an added 3G modem [10:06] I'm gonna have to find that link... [10:06] in such cases you can indeed just fiddle a bit, all the drivers and the kernel would be the same aynway ... but that wont work if you dont have identical architectures underneath [10:07] (like ... 100% identical ...) [10:11] Here. It's not the link I remember, but it's similar: http://forum.xda-developers.com/showthread.php?p=48884484#post48884484 [10:14] Nuts! I closed my IRC app again! Sorry if I missed your responses. [10:14] you didnt [10:14] (miss anything) [10:23] Unfortunately, I've tried the Ubuntu Dual-Boot app, but no matter what image I chose for the Nook, it always halted with the same error. [10:40] Well, this is interesting, but it doesn't solve my problem: http://www.makeuseof.com/tag/try-ubuntu-touch-right-now-ubuntu-computer/ [10:41] I'm trying it now. [11:10] ugh, someone shjould have told him that this is outdated by a year ... http://davidplanella.org/ubuntu-emulator-quickstart-guide/ would have been the better guide [11:11] what would it take to make a kubuntu phone image using the same base as ubuntu touch but different UI? is it just like another flavour of ubuntu or is there anything more? [11:12] Riddell, you would have to support mir [11:13] (or come up with an Xorg layer that can talk to hybris/android HAL) [11:14] or use wayland? [11:14] isn't that what sailfish does? wayland + hybris & HAL? [11:14] dunno, perhaps ... oour hybris layer has never been tested with wayland or Xorg i think [11:15] sure, it'd need a fair amount of work I imagine. Fun project if you have time Riddell ☻ [11:17] hmm, sounds a bit scary [11:18] and you would indeed need a lot of work to talk to all the sensors etc i guess [11:18] there are Qt modules for most of them, but i bet you need a lot new code to make use of the features [11:26] ogra: You obviously didn't know that this is also UbUntUToUchLover. I am using UTL under Ubuntu desktop. I will check that link you put out here. [11:28] Seems like I ran across that once before, but since it had nothing to do with my current problem, I ignored it. [11:28] Thanks anyway, ogra. [12:03] Riddell, well, if you have a supported device you can surely try it out === _salem is now known as salem_ [12:04] ogra_: like a nexus 5? === Tassadar_ is now known as Tassadar|nym [12:05] heh, well, thats not supported, but runs good enough for such fiddling i guess [12:05] what is supported? [12:06] mako (N4), flo (the later N7 model) and partially manta (N10) [12:06] but all you need for a start is to have a workinng base, a working lxc container and you need to know the UI comes upp with ubuntu ... that should be sufficient for a start [12:07] then you need to start replacing bits and pieces of it === alan_g is now known as alan_g|lunch [12:41] Ugh! I can't get out on the internet under the Touch Emulator, and the background is messed up. I was hoping to get a better idea of a working Ubuntu Touch image. [12:47] bzoltan, mvo, hey, is that a known issue http://people.canonical.com/~seb128/click.png (trying to create a touch project on qtcreator on current vivid amd64) [12:48] seb128: yes, it is a known issue and it is being fxed [12:48] any workaround meanwhile? [12:48] or ppa I can opt in? [12:49] seb128: but the end of that logs is more interesting [12:49] seb128: use https://code.launchpad.net/~mvo/click/chroot-15.04-multiarch [12:49] mvo, danke [12:49] bzoltan: hm, the error looks unfamiliar though [12:50] seb128: it might be a different error, but the ppa should fix at least one known issue [12:51] mvo, let me try that [12:51] mvo, in what ppa is that? [12:51] http://people.canonical.com/~platform/citrain_dashboard/#?distro=ubuntu&q=click has nothing [12:56] seb128: I don't think there is a ppa just yet [12:56] seb128: I meant to put it it into a silo today [12:59] mvo, k, I'm going to wait then, thanks [12:59] * seb128 wants to play with the sdk but no luck for a week trying to get qtcreator to work :-/ [13:00] mvo: you ain't gonna put that into no silo today :) [13:00] mvo: feel free to push it to the Dev Tools PPA [13:01] bzoltan: heh :) indeed [13:01] mvo: https://launchpad.net/~ubuntu-sdk-team/+archive/ubuntu/tools-development [13:01] bzoltan: I'm currently not in the team (I unsubscribed because I got all the MP spam from your busy team ;) [13:05] mardy, hey [13:06] mardy, valgrind often workaround segfaults but you likely get invalid read or write errors in the log still no? [13:08] seb128: no, under valgrind it was just working [13:08] mardy, well, it might still trigger invalid read/write in the valgrid log [13:08] even if it works [13:08] seb128: but anyway, I hope I found the issue; I'm just not sure how to fix it... vala ownership is quite complicated [13:08] that's often the case [13:09] seb128: nope, valgrind was perfectly happy [13:09] mardy, maybe larsu or desrt can help with vala questions [13:10] * larsu peeks in [13:13] mvo: OK, I will push then your branch [13:14] mvo, bzoltan, same error "failed to write session file: No such file or directory" using the click fix from mvo [13:15] larsu: hi! So, I have a GLib library with an object which has a method which returns a GHashTable, without incrementing its reference count [13:15] mvo: hey, long time no see :) [13:16] larsu: then, in the vala code, I'm instantiating the object, calling the method, then deleting the object, and finally I want to return the GHashtable [13:16] seb128: could you please pastbenin the full logs. Usually at the bottom you see the real reason. [13:16] popey: hey, we're ready for the release, what should we submit for security review? [13:16] mardy: uh oh. Is the method annotated with Returns: (transfer full)? [13:16] larsu: this currently does not work, because the GHashTable gets deleted when the object is destroyed; how can I tell vala to increse the HT reference count, before deleting the object? [13:17] zyga: you just submit the app like any other. if it needs security review we get notified [13:17] bzoltan, mvo, http://paste.ubuntu.com/9367159/ [13:17] larsu: the library method is "transfer null", because it doesn't increment the reference count [13:17] popey: well, wait, I cannot submit anything (com.ubuntu), I can only send it to someone I presume [13:18] larsu: the vapi file lists it as "unowned" [13:18] popey: and I was wondering how to attach the source [13:19] larsu: so, it looks like that everything is working fine if I keep a reference to the object in a private member (that is, if I don't destroy it) [13:19] mardy: hm, vala should do the right thing in this case. Try casting it with (owned) [13:19] popey: or should I actually go to developer.ubuntu.com and try to submit it there? [13:19] seb128: why the armhf at the end? [13:19] mardy: the hash table, I mean [13:19] zyga: yes, we discussed this yesterday remember [13:19] larsu: I tried, I get a compiler error; let me paste it... [13:19] zyga: mirv and balloons can upload with that namespace [13:20] bzoltan, end of what? [13:20] seb128: hm, anything unusual on your system? do you have /var/lib/schroot/session/ at all? [13:20] popey: yep, I'm just confused as to how to do that excatly, just send you an email with link to the click file? [13:20] bzoltan, I just click the "create new kit" in qtcreator, dunno [13:20] ../common/accounts.vala:27.16-27.49: error: Reference transfer not supported for this expression [13:20] return (owned) auth_data.get_parameters(); [13:20] larsu: ^ [13:20] ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ [13:20] mvo, it's a test box, nothing unsual that I can say [13:20] seb128: try `sudo click -a armhf -f ubuntu-sdk-15.04 create` [13:20] zyga: yeah, or ping them on irc [13:20] mardy: bah. Can you assign it to a new variable? [13:20] popey: ok, that's all I needed, thanks :) [13:20] mardy: HashTable<> x = (owned) auth_data.get_parameters(I) [13:20] larsu: if it works, I'm going to kill someone [13:21] mvo, no /var/lib/schroot/session no [13:21] mardy: ya, I hate vala's memory managementt. It's all over the place and not documented well [13:22] larsu: no, I get the same error on that assignment [13:22] seb128: confusing, if you create that dir, does that help? === yorick_ is now known as yorick [13:22] mvo, what is supposed to create it? [13:23] larsu: but I don't want to waste your time -- I got a solution anyway, by not destroying the object, I was just curious to see if I could have done better [13:23] mardy: ya not sure, sorry [13:24] mvo, started again a kit creation, telling you in 15 minutes [13:24] larsu: nm, thanks anyway [13:25] seb128: well, something is supposed to create it [13:26] bzoltan: if you make me a ubuntu-sdk-team member again I will upload to your ppa [13:27] mvo: I will add you, but I can do the upload this time.. I started already :) [13:27] bzoltan: cool, even better [13:28] mardy: so, desrt says this works: http://paste.fedoraproject.org/156570/17699679/ [13:29] larsu: right, that should work; let me try... [13:31] zyga, okay, I've got the package [13:31] kissiel: let's move there [13:31] kissiel: oh [13:31] we're here [13:31] :D [13:31] kissiel: link please :) [13:32] zyga, https://www.dropbox.com/s/sq97yacz3ykat89/com.ubuntu.checkbox_1.0.1_armhf.click?dl=0 [13:32] kissiel: ok, getting and testing [13:32] mardy, you should join #ubuntu-desktop so larsu wouldn't have to proxy the discussion with desrt and you between channels ;-) [13:33] seb128: thanks :) [13:34] oh, right [13:36] Mirv: hey [13:36] Mirv: we're (kissiel and me) from the hardware cert team [13:36] Mirv: we've prepared an application that we'd like to upload to the click store [13:36] Mirv: kissiel sent a link earlier there [13:37] zyga: actually if Mirv uploads it, it'll be "owned" by "Ubuntu Core Apps Developers", right Mirv ? [13:38] popey: I don't think that's a problem really [13:38] popey: as long as we can upload it over and over [13:38] popey: it's all GPL [13:38] ok [13:38] popey: and we welcome contributors [13:39] kissiel: it works on my kryllin [13:39] kissiel: I'll check exportt [13:40] kissiel: it just crashed [13:40] kissiel: on startup (resume propmpt) [13:40] zyga, oh [13:40] zyga, testing [13:40] kissiel: doesn't crash anymore [13:40] kissiel: maybe clear .cache [13:40] kissiel: and try again [13:42] kissiel: I cannot reproduce that anymore [13:42] kissiel: I think it's good [13:43] zyga, w8, I want to test one more thing [13:43] kissiel: reports are empty? [13:44] kissiel: nah just mpt bug [13:44] kissiel: they are ok [13:45] zyga, stable here [13:47] Mirv: ok, this is good [13:47] kissiel: upload that to launchpad [13:47] kissiel: and sign the click package as indicated [13:47] zyga, allrighty [13:51] zyga, are you calling me buggy? [13:51] mpt: no, where? :-) [13:52] ah [13:52] media transfer protocol :) [13:52] I don't know how I wrote that down as mpt, not mtp [13:52] It’s a common mistake ;-) [14:03] I read somewhere that some Ubuntu Touch images are wrapped around Cyanogenmod. Is that true? [14:04] balloons: hey [14:05] balloons: or maybe you [14:05] UTL: i think we switched to AOSP. [14:05] balloons: do you have a moment to upload our app to the store under com.ubuntu? [14:05] UTL, not "wrapped around" no [14:06] AOSP? I've heard that. What is that? [14:07] the free android source [14:07] zyga, let me look and ensure I can [14:07] balloons: thanks a lot [14:08] but we dont "wrap" around it, you boot into normal ubuntu. like on your desktop ... later there gets an lxc container started in which some android services run ... to give you access to the closed source drivers [14:08] balloons: https://launchpad.net/checkbox-touch/trunk/1.0.1 [14:08] balloons: that's the click package [14:08] balloons: it actually works on armhf and x86 [14:08] balloons: so if there's a way to upload it for other arches (emulator) then that's doable as well [14:09] did ubuntu-device-flash stop supporting --revision? I'm not sure if this is an ogra_ question or somebody else [14:09] balloons: sounds like he wants one of your patented fat packages ☻ [14:09] balloons: it's a pure python/qml app with duplicated libs for all architectures [14:09] nuclearbob: it does still support it, I think you need to add "touch" first [14:09] zyga, and it's already in the store right? [14:09] nuclearbob, more a sergiusens one, but no, it didnt stop supporting --revision [14:09] popey: I have touch first [14:09] ogra_, thanks, I'll ask sergiusens [14:09] i still see it in the online help [14:09] zyga, yes we can make it a fat package. [14:09] balloons: nope, that's the first upload [14:09] yeah, me too [14:10] balloons: great! [14:10] would be evil to drop that since QA uses it daily :) [14:10] but I get unknown flag `revision' [14:10] ogra_, that's why I'm concerned :) [14:10] zyga, ahh ok, in that case, let's upload it. I need an icon, description and screenshots as well ideally [14:10] also a support link or email address [14:10] brb [14:10] nuclearbob, i'm sure davmor2 used it this week :) [14:10] (or last at least) [14:11] maybe davmor2 can tell me what I'm doing wrong. I've tried it as the last argument, the first argument after touch, and a couple other places [14:11] kissiel: ^^ [14:11] kissiel: let's make some screenshots please [14:11] balloons: do you need a given form factor? [14:11] zyga, sure [14:11] kissiel: and the icon [14:12] balloons: there's an icon in the click file (svg) [14:12] nuclearbob: ubuntu-device-flash --channel devel-proposed --revision 39 --wipe this works fine for me if that is the kinda of thing you are doing [14:12] balloons: we can extract it for you [14:12] kissiel: I can make screenshots from kryllin [14:13] alan@deep-thought ~> ubuntu-device-flash --channel devel-proposed --revision 40 --wipe [14:13] DEPRECATED: Implicit 'touch' subcommand assumed [14:13] zyga, what is preferred? both? [14:13] unknown flag `revision' [14:13] 0.4+15.04.20141125-0ubuntu1 of u-d-f, what version you have davmor2 ? [14:13] kissiel: don't know [14:13] kissiel: weird, it just crashed on startup [14:13] kissiel: where are the logs? [14:13] davmor2, popey, thanks, I'm using a different channel, but not having any luck with either one. I'll make sure all my stuff is up to date [14:13] nuclearbob: ubuntu-device-flash --revision ## touch --channel ... [14:13] kissiel: now it works again [14:13] zyga, .cache/upstart [14:14] zyga, I had to restart my device one time [14:14] sergiusens: so --revision comes before --touch ? [14:14] that works [14:14] zyga, like previous stuff was not cleared from mem [14:14] nuclearbob: not --touch, touch, a positional command [14:14] er, yeah, touch, not --touch, sorry [14:14] kissiel: I don't have that directory [14:14] nuclearbob: I sent an announcement email to ubuntu phone a month ago with this info ;-) [14:15] popey: I haven't thought about the ownership part or how the stuff looks that I upload.. [14:15] zyga, IDK then [14:15] sergiusens: yeah, I'm sure I read it at the time, but I'm using some ci tools, I'll file an mp with them, thanks [14:15] zyga: ok. [14:15] sergiusens, we're all old men nowadays ... you have to re-send that once a month for two months at least :P [14:15] So, if I flash Cyanogenmod/Android first and then flash Ubuntu Touch, it refers to the previous image - or does it come with Android services? [14:15] Mirv: baloons is helping us now [14:15] zyga: so if you need something, give a direct link and I'll look at the upload. [14:16] zyga: oh, checkbox is coming to the phone? [14:16] UTL, it completely wipes your phone during install [14:16] Mirv: it is ready now :) [14:16] I guess that means that it has some services preinstalled. [14:16] ogra_: is upstart logging disabled on rtm? [14:16] ogra_: I don't have per-app log files [14:16] (somehow() [14:16] zyga, nope [14:17] zyga: cool! :) [14:17] hmm [14:17] * zyga reboots the phone [14:17] Mirv: it's also converged so expect to see it on the desktop soon [14:17] UTL, no, why would using hardware drivers have anything to do with android services ? [14:17] Yeah, I guess you're right. [14:18] zyga, they should live in ~/.cache/upstart/ [14:18] zyga: should you get the click package automatically created at jenkins from some code branch? I also prefer fetching .click:s over VPN instead of downloading from random dropbox links :) [14:18] mvo, where is the list of click chroots stored on disk? [14:18] ogra_: I'll reboot and see, I didn't have that before [14:18] Mirv: oh, don't please [14:18] Mirv: the package is still built manually but we'll get better at it [14:19] balloons: Mirv: can we agree who's doing the upload [14:19] not to get two somehow :) [14:19] Mirv: the signed click package is on https://launchpad.net/checkbox-touch/trunk/1.0.1 [14:19] k, I'm back. [14:20] zyga: the second upload would be rejcted :) [14:20] Mirv: that's good to know, I just don't want to waste anyone's time [14:23] kissiel: I'm making screenshots now, just one more [14:23] Mirv, balloons: so who shall upload it? [14:24] zyga: I can [14:24] ok [14:24] kissiel: ^^ [14:25] k, I'll let it be. Are you making it a fat package? [14:26] ok 4 screenshots ready [14:26] zyga: popey: I haven't seen this error before http://pastebin.ubuntu.com/9367860/ [14:26] zyga, Mirv : https://www.dropbox.com/s/kr4q9jiaakyies7/checkbox-touch-flo-screens.tar.gz?dl=0 [14:26] i suspect datacentre timeouts [14:26] i just had all kinds of timeouts in lp and others [14:27] * Mirv retries [14:27] * zyga tarballs his screenshots [14:28] seb128: /var/lib/schroot/chroots and /etc/schroot [14:28] mvo, thanks [14:29] hey zyga, just saw your earlier message, hello to you as well :) [14:29] mvo, I wiped out /var/lig/schroot/chroots and it was still telling me the target existed ;-) [14:30] balloons: ok, would you like to try from that LP address? I'm just getting that error again and again. === dandrader_ is now known as dandrader|afk === alan_g|lunch is now known as alan_g [14:32] I also have never uploaded such a big .click package, although gallery has been 5MB [14:32] Mirv: uploading screenshots ... [14:32] Mirv: perhaps we can trm it down [14:32] Mirv: to be just armhf [14:32] zyga: I'd think the size shouldn't be any problem, but this'd be something else [14:32] kissiel: ^^ can you try just to see how big it is [14:32] Mirv: ok [14:32] it's 8MB now, not a big leap from gallery's 5MB [14:32] Mirv: https://www.dropbox.com/s/y2igzd90sa0jbp2/checkbox-screenshots-1.0.1.kryllin.tar.gz?dl=0 [14:33] zyga, sure, trimming [14:34] zyga, Mirv: It's 4.6MiB after trimming [14:35] seb128: yeah, kill it i /etc/schroot as well [14:36] seb128: its a bit of a idiosyncrasy of click that it does not make that easier or detect that its only half there [14:37] mvo, worked, danke [14:39] Mirv, which address is failing for you? [14:39] checkbox is breaking the click store :) [14:39] balloons: trying to upload the .click from https://launchpad.net/checkbox-touch/trunk/1.0.1 [14:39] darn :) [14:40] we should add a test for that (head explodes) [14:40] Mirv, right, yes I have the click from that page === zyga is now known as santa [14:40] seb128: cool and chroot creation works now too? after you created that sessio ndir? === santa is now known as zyga [14:41] mvo, yes [14:42] mvo, thanks for that as well [14:42] mvo, still unsure what component is buggy/why that didn't happen automatically though [14:42] Mirv, so I will create a new application for it and try uploading [14:44] Mirv, zyga seems to have uploaded into the store fine for me [14:44] balloons: right, I don't know how to create new apps since I have always just uploaded updates [14:44] balloons: great! [14:44] balloons: great [14:44] balloons: fetch screenshots from the two tarballs me and kissiel sent [14:44] balloons: both on dropbox in the backlog here [14:44] zyga, can I have a tagline and description from you? [14:44] balloons: the icon is in the click itself (svg), tell us if you need something else [14:44] oh [14:45] hmm [14:45] =System testing utility for Ubuntu [14:45] description... [14:45] hmm [14:45] kissiel: ^^ any ideas [14:45] mvo: /build/buildd/click-0.4.35.1~0ubuntu1~0vivid3/click/chroot.py:179:21: W291 trailing whitespace [14:45] zyga, System testing [14:45] seb128: yeah, no idea, pleae file a bug for click so that it does not get lost [14:45] mvo, ok, thanks [14:46] mvo: https://launchpadlibrarian.net/191832332/buildlog_ubuntu-vivid-amd64.click_0.4.35.1~0ubuntu1~0vivid3_FAILEDTOBUILD.txt.gz [14:46] "Checkbox is a system testing application for the Ubuntu platform. It can be used to verify that the device you are using is functioning correctly. Detailed results of the test can be exported to a report which may be of use for post-sales support" [14:47] balloons: I don't know if you need any stuff like license, links etc, the code is on lp:checkbox, bugs can be filed on lp:checkbox or lp:checkbox-touch) [14:47] balloons: the code is GPLv3 [14:47] zyga, so I see you have multi-arch libs in here.. [14:47] balloons: with small bits of LGPLv3 [14:47] balloons: yep [14:48] balloons: it's a qml + python + multiarch .so click file [14:48] and tests.. but I don't see a manifest yet. how did you make this? [14:48] balloons: manifest? [14:48] kissiel: ^^ [14:48] balloons: click build [14:49] balloons, it's there [14:49] balloons: we have a script that gets the libaries from the ubuntu archive, unpacks them and does click build [14:49] balloons: we don't compile anything ourselves [14:49] pitti, hi are you the right person to shepherd the landing to fix https://bugs.launchpad.net/ubuntu/+source/ubuntu-system-settings/+bug/1337200 [14:49] Launchpad bug 1337200 in upower (Ubuntu RTM) "High CPU due to excessive device changed signals from upower" [High,Confirmed] [14:49] balloons: the only manifest I can think of is in lp:checkbox in the directory checkbox-touch/ === dandrader|afk is now known as dandrader [14:49] balloons: manifest.json [14:50] also I see click-review fails due to the maintainer email domain @canonical.com [14:50] bzoltan: sorry for that, please merge again fixed in r552 i my branch [14:50] balloons: oh [14:50] zyga, because of how's its built it seems like it will only be recognized as armhf in the store [14:50] balloons: how should we build it then? [14:50] there's also quite a bit of stuff that isn't needed in the package, hence my asking [14:51] balloons: such as/ [14:51] pmcgowan: quite overloaded ATM -- I thought kenvandine was landing this? [14:51] the readme's, shell and build scripts, tests [14:51] pitti, ok talking to sil2100 it may be handled [14:52] balloons: that's true, we don't separate those out (tests), as for build scripts and readmes we could do that but I'm not sure of a good way to do it [14:52] pitti, i'm not that comfortable with powerd and upower... i can take a swing if needed though [14:52] zyga, does the project build with cmake or qmake? [14:52] I thought it was all already in a silo? [14:52] balloons: neither [14:52] balloons: it's not compiled [14:52] kenvandine: hey! [14:52] kenvandine, seems te silo installs cleanly, what was the concern? [14:52] balloons: it's just a qml + js + python combo [14:53] kenvandine: so, we actually *want* upower and powerd along with this silo ;) [14:53] pitti, i had it in a silo for testing, not necessarily ready to land [14:53] zyga, but you have.. ohh the compiled code is libpython :-) [14:53] balloons: our build is 1) get deps from the repo, unpack, 2) click build [14:53] balloons: it's straight from the archive :) [14:53] kenvandine: some of our guys + QA tested it a little bit before and it helps [14:53] i just pulled the vivid packages for that silo, which might include more stuff than we want [14:53] balloons: no need to build our own fork, the one in the archive is good and has security updates that we can integrate trivially [14:53] ahh I see the build-me [14:53] i know system-settings in that silo includes lots of stuff [14:54] i can prepare a branch for settings [14:54] balloons: the release we're trying to get into the store will be merged by tarmac soon, the tag is checkbox-touch-v1.0.1 [14:54] balloons: kissiel can tell you how it was built exactly (we used released components of the test data and plainbox (internal dependency) [14:55] balloons, it pretty much takes whole working directory with all necessary libs and runs click build on that directory, it does not cherry-pick files to embedd into click... [14:55] ok, well we will need to fix the click-review.. the store won't take it as-is. Also I assume you *have* to unconfined? that will force a manual review [14:55] balloons: if there's a way for us to not include everything (an inclusion manifest) we could do that [14:56] balloons: other than that, we don't have that much needless stuff there [14:56] zyga, the thing to do would be to output to a build dir and click build it there. So just copying over the bits needed to run [14:56] hmm, ok, we can think of doing that [14:57] balloons: is this a blocker for 1.0.1? [14:57] balloons: we want to put that into the store and iterate [14:57] balloons: and catch things like this on the first upload [14:57] no, the only blocker is the click-review complaints [14:57] balloons: to know what to change [14:57] ok [14:57] kissiel, also so where exactly is the manifest file? I don't see it at all [14:57] balloons: so what are those? the @canonical email? [14:58] balloons: checkbox-touch/manifest.json? [14:58] zyga, yes, security of unconfined and the canonical email much match the namespace [14:58] balloons: ahh [14:58] (Your email domain needs to match the reverse package namespace.) [14:58] balloons: ok [14:58] balloons: can we put zkrynicki@ubuntu.com? [14:58] that should work [14:58] balloons: we cannot put an @ubuntu team name, can we? [14:58] balloons: unconfined is how it's meant to be [14:59] balloons, whoa, `click contents com.ubuntu.checkbox_1.0.1_armhf.click |grep manifest` returned none o.O [14:59] balloons: currently we don't use it much but later on some tests will have to [14:59] kissiel: weird? [14:59] zyga, weird indeed [15:00] balloons: http://bazaar.launchpad.net/~checkbox-dev/checkbox/trunk/view/head:/checkbox-touch/manifest.json that's how it is supposed to look like [15:00] kissiel, heh, good to know I'm not crazy [15:00] I want to make sure the manifest is correct [15:00] (we still call it -touch in a few places, mostly not to collide with the desktop-y version [15:01] so we should update architecture in the manifest to say you support i386 and amd64 also [15:01] balloons: how should that look like [15:02] "architecture": ["armhf", "i386", "amd64"], [15:02] balloons: ok [15:02] kissiel: I'll make those changes [15:02] balloons: so the email and that [15:02] * balloons checking on other thing [15:03] zyga, I think it would be useful to also thinking about including the extra x-source entry that lists the bzr source and rev number [15:03] balloons: oh, great [15:03] not something that has to be done now, just thought [15:03] balloons: any examples? [15:03] balloons: sure, I want that [15:06] zyga, well we actually do it a couple ways.. let me get an example [15:06] zyga, http://paste.ubuntu.com/9368397/ [15:06] balloons: fantastic, adding [15:09] balloons: how should vcs-bzr-revno be maintained? [15:09] balloons: can we use tags there? [15:09] balloons: or just ints [15:10] balloons: especially since all our landings to trunk are autmoatic merges so the real revno will be like 123.5.1 [15:12] zyga, also notice we like to include the bzrevno in the version, but not required. "version": "3.2.@BZR_REVNO@", [15:12] oh, nice [15:12] what expands @BZR_REVNO@? [15:12] click build? [15:12] zyga, yes it's built with cmake [15:12] balloons: ah, we don't have cmake [15:12] balloons: ok, we'll do it once we have cmake [15:12] right, so that's up to you.. the x-source data is built the same way [15:13] I think it makes sense at least to denote the click build in the store corresponds to the revision number somewhere. [15:13] balloons: thanks [15:13] so however you wish to keep track of it :-) [15:13] balloons: as long as that data can be a tag, that's okay [15:14] balloons: we're always tagging releases [15:14] balloons: and matching tags to builds is easy [15:16] ok, so let me finish this upload, then we'll need a new click [15:16] kenvandine: ok, just give us a sign once you're happy with the system-settings part in the silo ;) [15:16] kenvandine: since we're most interested in the powerd parts [15:16] every upload has to be a newer version number, so keep that in mind. If you fail review, you need a newer version to upload [15:16] sil2100, i just updated the settings branch for that silo [15:17] kenvandine: I think you won't be able to rebuild it right now though [15:17] the silo includes a sync of indicator-power too, so we should make sure that doesn't include unwanted bits [15:17] Since CI Train is still down [15:17] yeah... [15:17] sil2100, but my MP is updated :) [15:17] \o/ [15:17] Excellent ;) [15:17] Thanks [15:18] balloons: ok [15:18] balloons: I've got a merge request with the changes we've talked about [15:19] balloons: https://code.launchpad.net/~zkrynicki/checkbox/checkbox-touch-1.0.2/+merge/243686 a quick sanity check [15:20] mvo: there is an extra chroot thing ... we need 14.09 chroots for the RTM devices [15:20] balloons: btw, did you somehow manage to get jenkins to build multi-arch clicks or something? [15:21] dobey, I didn't but it should be trivial to do so. sergiusens even has a script that I believe CI was using for this purpose [15:21] hmm [15:22] hello [15:22] bzoltan: ok, what kind of libs do hey get? [15:22] mvo: it is pretty much identical to utopic. Only the source is different. [15:24] ok [15:25] dobey, http://bazaar.launchpad.net/~sergiusens/uci-engine/click/view/head:/click-builder/click_builder/clickbuilder.py [15:26] balloons: does that MR look okay? [15:26] balloons: I wonder what to do with the manifest file [15:28] zyga, yes it looks fine [15:28] balloons: ok [15:28] balloons: anything else or should I hand you another click file :-) ? [15:31] go for it. But you might want to consider trying to pull only what you need as suggested. Not sure how easy it is to modify the build script [15:32] balloons: I won't make that change now, not for 1.0.2 [15:32] balloons: we've considered doing cmake based builds and we will but for 1.1 [15:32] balloons: we want to get this to $customers for testing [15:33] k, then fire away and I'll push the new click [15:33] kissiel: did you set any special test plan? [15:33] kissiel: I'll just do ./build-me --provider=... [15:34] d'oh :) [15:35] zyga, just ./build-me --provider, if there's only one TP it will select that one automatically [15:36] zyga, does it answer your question? [15:37] kissiel: yep [15:37] balloons: ok, tested ok, uploading [15:43] balloons: ok, ready: [15:44] just send a link whenev's [15:44] balloons: https://launchpad.net/checkbox-touch/trunk/1.0.2 [15:44] balloons: just finishing the upload [15:44] balloons: interesting to see the .multi extension [15:44] balloons: k [15:44] it gave you one? excellent [15:44] balloons: https://launchpad.net/checkbox-touch/trunk/1.0.2/+download/com.ubuntu.checkbox_1.0.2_multi.click [15:45] balloons: yep, ready and tested [15:46] this looks better [15:46] \o/ [15:46] \o\ [15:46] /o/ [15:46] !o! [15:50] balloons: so the next thing I expect is the immediate security review [15:50] balloons: immediate as in it's going to be blocked quicky :) === kissiel is now known as kissiel-afk [15:52] it's pending manual review now [15:53] balloons: excellent, thank you very much :) [15:57] sergiusens: any idea when we'll have multi-arch clicks building in jenkins by default? [16:02] mvo: zbenjamin: the Dev Tools PPA is good to go from the point of 15.04 and qmake support. Good job guys, thank you. [16:02] yay [16:02] bzoltan: awesome :) [16:06] dobey: no, I'm not involved in that anymore; it's a pure ci story; you may ask Ursinha as she masters the roadmap [16:06] * balloons notes we should get it on the list [16:06] oh, ok [16:06] haha I wish I "mastered" something [16:06] yes [16:07] Ursinha: the scrums? [16:07] dobey: well, yeah :) [16:07] dobey: there is at least one story to build clicks automatically, we can create another one of that's not enough [16:08] right in this cause, we want to insure the clicks are multi-arch. You could probably just modify the story to include it [16:09] right === chihchun_afk is now known as chihchun [16:13] balloons: no no [16:13] no modifying stories directly :) [16:13] if a story is already there, contact me or ev so we can do that for you [16:13] Ursinha: it's a choose your own adventure story ;) [16:14] otherwise we'll lose track of the information we already gathered, and that changes the scope of the story [16:14] dobey: hehe [16:15] Ursinha: i think the problem here is that they really aren't different stories. the underlying technical requirements to fulfill the story might be slightly broader though. [16:15] Ursinha: from my POV the story is still "i am a developer and want a click package built that is verifiable, repeatable, and to be uploaded to the store" [16:16] given how the packages are defined, on an underlying technical level, that should inherently mean that it would be a multi-arch fat package if the app contains compiled code [16:16] dobey: we can't have underlying technical requirements, they really have to de explicit in a way we can choose to have first a more basic version of the final solution (multi-arch builds) [16:16] but i'm not your PM of course :) [16:17] dobey: if that's the case where the very minimum viable thing is multi-arch, then we adapt and explicit the scope [16:17] dobey: PMs don't exist in scrum, really :) [16:17] Ursinha: the P can be for pseudo if you want :) [16:17] lol [16:18] dobey: what the PM would do? [16:18] i thought "story" was the same as "user stories" are for design, though [16:18] Ursinha: i don't know, you're the scrum master, not me :) === JMulholland_ is now known as JMulholland === Saviq_ is now known as Saviq [16:19] dobey: it's a story that goes into our backlog, it can be a user, a technical or a defect one [16:19] dobey: as I said, a PM doesn't exist on scrum, so I want to understand what kind of input you need [16:19] :) [16:20] Ursinha: i'm just a developer. i need automated builds of multi-arch click packages :) [16:20] well just drop intel [16:21] :'-( [16:21] (everywhere) [16:21] But I gave away my arm laptop . . . [16:21] these ubuntu isos would surely be smaller being armhf only [16:21] balloons: just one more question, how will we get questions / feedback on the secuirty review [16:21] balloons: or notification on the actual upload? [16:22] dobey: noted :) (no, seriously, we're dealing with the citrain outage right now but once that is back I'll reach out to you to get more details, does that sound good?) [16:22] zyga, I will have to let you know since it's under the common namespace [16:22] ogra_: the "con" in "convergence" isn't short for "contra" :) [16:23] dobey: hahaha [16:23] Ursinha: sure [16:23] dobey: cross build your packages [16:23] dobey: or build them in a PPA and combine as we do [16:23] dobey: solves a lot of the problems [16:23] zyga: eh? [16:23] there are armhf tablets, TVs, dishwashers and laptops [16:23] PPAs don't build clicks :) [16:23] * zyga read a bit of the backlogs [16:23] dobey: I know [16:23] ogra_: and x86 too :) [16:24] pfft [16:24] dobey: but you can package your deps as debs and then just unpack them [16:24] obsolete arch [16:24] :) [16:24] dobey: look at checkbox the click package [16:24] dobey: it's all compiled by launchpad [16:24] * dobey builds a powerpc click for ogra [16:24] dobey: just assembled from .debs and manifest by a shell script anywhere [16:24] dobey: I bet we're the first click app that does that too :P [16:24] zyga: sounds like overengineered solution to me :) [16:24] dobey: so we reuse 99% of the lp workflow [16:25] dobey: it's just plain easy, fast iteration time [16:25] dobey: (as the app is python) [16:25] dobey: anyway, just a thought :) [16:25] dobey: until there's a widely available click builder [16:25] dobey, keep in touch on what happens. I'll let you make sure the story gets captured :-) I imagine once click supports it, CI will adopt it [16:25] which is to say, once click makes it automagical [16:26] we'll see [16:26] * zyga wonders if anyone wants a ppc click for power [16:26] * ogra_ fires up unity8 on his G3 iBook [16:26] er a checkbox click for tower [16:26] er [16:26] power [16:26] zyga, jdstrand needs to ack checkbox since it's unconfined [16:26] man, I'm tired [16:26] balloons: yep, we know that [16:27] balloons: it's mostly the limit of the current SDK [16:27] balloons: content hub being unfinished etc [16:27] balloons: we may go down the unconfined path later on [16:27] balloons: but not for RTM [16:27] zyga: well, my stuff only veers slightly outside of using what's in the SDK itself. sadly i don't get to write python [16:27] dobey: well, we really tried not to abandon python :) [16:27] zyga, I'm saying for the review and upload, he needs to sign off on it. So it will be in review until that happens [16:27] balloons: understood, thanks [16:28] jdstrand: please keep me in the loop, if you have any questions I'm here all the time [16:28] anyway, i should get lunch [16:29] balloons: can you point me to the review? [16:29] jdstrand, https://myapps.developer.ubuntu.com/dev/click-apps/1284/ [16:30] * zyga sticks around :( [16:30] :) === chihchun is now known as chihchun_afk [16:37] jdstrand: if you want I can explain why we used unconfined policy === dandrader is now known as dandrader|lunch [17:02] zyga: this is a multiarch package? [17:05] jdstrand: yes [17:05] jdstrand: we take prebuild packages from the ubuntu repository [17:05] jdstrand: python3.4 and pyotherside [17:05] jdstrand: and unpack them [17:05] jdstrand: and libxml IIRC [17:06] jdstrand: the rest is arch independent [17:06] * zyga checks libxml [17:06] jdstrand: yep, libxml2 as well [17:06] jdstrand: the execution starts at qmlscene [17:06] jdstrand: which loads pyotherside, then python and the rest of our code [17:06] zyga: I commented on the unconfined bit, so you are unblocked there. I'd like someone more familiar with fat packages to review that part [17:06] jdstrand: ok [17:07] zyga: perhaps beuno knows who could look at that [17:07] jdstrand: we're not sure if there's any docs really [17:07] * jdstrand hasn't worked with fat packages much yet [17:07] jdstrand: we rely on a hack in how click works and how app runtime is set up [17:07] I can do the fat part [17:07] I've been eating my share of ice cream [17:07] jdstrand: so lib/$arch is setup at runtime and things just work [17:07] beuno: thanks! [17:07] yeah, there's a bug [17:07] I think, in the review scripts [17:07] beuno: if we can get that package into a state that is correct, I can adjust the tools to verify fat packages properly [17:08] for mapping multi to multiple arches [17:08] offtopic, is there any kind of "DLC" supprot planned for the click store? 99% of the changes we'd like to make are to a "DLC" to the app (the actual tests0 [17:08] jdstrand, I think it is, it's just trying to match "multi" to all the arches in the manifest [17:08] cause those checks are pretty ancient and written based on conversations and docs for long ago [17:08] and that could perhaps speed up the review (no need to do separate security review for a bunch of data files) [17:08] which is annoying, mxing dpgkisms and our stuff [17:08] zyga, in ~6 months [17:09] beuno: well, like I said, I'll fix the tests if you verify the package [17:09] beuno: great, I'd love to beta-test that :) [17:09] jdstrand, deal. [17:09] checkbox could download DLC for a particular hardware it runs on for example [17:10] is there a way to check out click apps on the website somehow? (other than having a device to browse through the store? [17:10] zyga, not currently [17:10] zyga, approved and published [17:11] beuno: awesome news, thanks! [17:12] \o/ [17:16] zyga: http://popey.mooo.com/screenshots/device-2014-12-04-171612.png is that git revision really needed? [17:16] would a YYYYMMDD not be more useful? [17:21] popey: oh, good idea [17:21] popey: we're just getting started [17:21] popey: I'll add that for 1.1 [17:21] thanks! [17:26] popey: reporeted as https://bugs.launchpad.net/checkbox-touch/+bug/1399282 [17:26] Launchpad bug 1399282 in Checkbox Touch "Include the build date on the welcome screen." [Undecided,New] [17:26] popey: we'll get it fixed for the next release :) === kissiel-afk is now known as kissiel [17:30] kissiel: re :) [17:30] kissiel: get it while it's hot [17:30] kissiel: it's really here [17:34] zyga, I know! :) cool! \o/ [17:43] popey: what do you have against arcane hexadecimal strings?! [17:44] dobey: nothing a backspace key can't help === dandrader|lunch is now known as dandrader [18:10] stgraber, do you have an idea why the system-image importer runs so long even if it doesnt import anything ? [18:11] ogra_: would have to run in debug mode, but it could be because some servers are not responding and it's hitting http timeout [18:12] i'm just noticing that i have a lock at every copy-image run when i try to promote a set ... (5 arches atm) [18:12] 5697 ? S 9:34 /usr/bin/python /srv/system-image.ubuntu.com/bin/import-images [18:25] mterry, what's the status of the wizard silo? anything blocking that besides the outage? [18:25] kenvandine, no [18:26] great === alan_g_ is now known as alan_g|EOD [18:40] jgdx, can you review this when you get a chance https://code.launchpad.net/~ken-vandine/ubuntu-system-settings/brightness_slider_rtm/+merge/241982 [19:20] stgraber, ok, vivid images promoted, would you mind switching the devel alias to vivid ? [19:20] oh, wait, slangasek was the new alias bender for s-i, wasnt he ? [19:23] stgraber: I still want us to fix it so that devel and devel-proposed are channels, not aliases, and we get rid of the vivid* channels. Does changing the devel alias in the short term make this any easier or harder later? [19:24] it will at least bring devel back in sync with devel-proposed [19:25] slangasek: shouldn't make a difference, we'll need manual surgery in any case. [19:25] but big +1 from dropping all distro named channels [19:25] s/from/for/ === jhodapp_ is now known as jhodapp [19:27] stgraber, ogra_: I'm not in a position to make the alias switch right now; hopefully one of you can do so, given that it's an obviously-correct change [19:29] doing it [19:30] >>> pub.change_channel_alias("ubuntu-touch/devel", "ubuntu-touch/vivid") [19:30] True [19:30] will be live by the end of the current import-images run [19:32] stgraber, ah, cool, i would have forgotten the ubuntu-touch prefix [19:32] * ogra_ makes a note for next time === dandrader is now known as dandrader|afk === IdleOne- is now known as IdleOne [20:28] hi, i've got a Dell TP713 (very similar to Apple Magic Trackpad). Is it possible to enable some of the touchscreen multitouch (3-finger pinch, etc) features via the pad? === dandrader|afk is now known as dandrader === salem_ is now known as _salem [23:15] why would some scopes lose their 'favorite' setting after updating? [23:27] hi [23:27] somebody? [23:28] i want to ask something [23:28] kostis just ask [23:28] you dont have to ask to ask :) [23:29] i have samsung galaxy tab s sm-t805 [23:29] is it possible to install ubunt touch? [23:29] or ubuntu some edition [23:30] kostis not without some serious hacking I guess [23:31] where can i find info about that? [23:31] kostis your device is not one of the reference devices that are used to develop touch [23:31] i know that [23:31] kostis there is a page of ports did you find that? [23:32] i asked if there is a way to do it [23:32] send me the link to the ports please [23:32] kostis: of course. You just have to write drivers for all of the hardware in your phone. [23:33] how to do that? [23:33] i don't know [23:33] If you have to ask, you can't do it. [23:34] kostis https://wiki.ubuntu.com/Touch/Devices [23:34] It's not impossible, but you have to be a hardware engineer. If it were easy, well, people probably would have done it by now. :/ [23:35] It's not impossible, but you have to be a hardware engineer. If it were easy, well, people probably would have done it by now. :/ [23:35] 15:34 < matv1> kostis https://wiki.ubuntu.com/Touch/Devices [23:35] what to do? [23:36] i opened that page [23:36] then? [23:36] kostis you could see if someone else did a port already [23:37] i can't see anyone [23:38] scroll down the page [23:39] but I already looked you version is not there [23:39] i see it too [23:39] so what next? [23:39] kostis learn how to write drivers [23:40] but like nhaines said [23:40] its difficult [23:40] very [23:41] kostis otherwise you could trade your tablet in for a nexus tablet [23:42] no thanks [23:42] kostis ok :) [23:42] kostis do you know how to program in any language? [23:42] is there a way to install ubuntu desktop? [23:43] kostis no not anything serious