[00:00] SturmFlut, yeah a mir disconnect usually causes a abort [00:01] IIRC [00:03] bschaefer: Hm. Well, enough for today, it is 1 AM again. See you tomorrow! [00:03] SturmFlut, cya! === duflu_ is now known as duflu === K1773R_ is now known as K1773R === clopez_ is now known as clopez === duflu_ is now known as duflu === marcusto_ is now known as marcustomlinson_ === marcustomlinson_ is now known as marcustomlinson === ecloud is now known as ecloud_wfh === chihchun_afk is now known as chihchun === chriadam is now known as chriadam|away [07:44] good morning [08:10] dholbach: Good morning! [08:11] hi SturmFlut === Guest54456 is now known as Laney [09:19] sergiusens: \o/ thanks for fixing the emulator === dandrader is now known as dandrader|afk [09:34] :( https://developer.ubuntu.com/en/2014/06/10000-users-of-ubuntu-phone/ [09:34] where can I find ? ^ [09:35] Good morning all; happy Answer Your Cat’s Questions Day! :-D [09:37] l3on: looks like a redirect issue on the new site [09:38] dpm: mhall119 ^^ looks like there's a broken redirect on the new site [09:39] perhaps we didn't port that blog post? [09:41] I don't know... but it's a broken link [09:41] (we link to it from http://voices.canonical.com/user/163/tag/phone/ for example [09:41] popey, http://discourse.ubuntu.com/t/10-000-users-of-ubuntu-phone/1729 === dandrader|afk is now known as dandrader === Elleo_ is now known as Elleo [11:15] popey: goto that link you just posted in chromium do images load for you? [11:16] no, which is what l3on was reporting [11:16] popey: ah cool [11:17] pitti: no problem [11:17] sorry for breaking it [11:17] I never thought changing from GiB to GB would break it :-/ === _salem is now known as salem_ [11:21] sergiusens: heh -- it's rare to see a tool handling disk size which only deals with the power-2 units [11:27] elopio, ping === barry` is now known as barry === dandrader is now known as dandrader|afk === MacSlow is now known as MacSlow|lunch [12:20] oSoMoN, thanks for the review. I've addressed your comments, but I'm not 100% sure where to put the singleton creation callback. [12:23] oSoMoN, the callback is at L549, in plugin.cpp. I'm pretty sure it does not belong in the class itself. [12:34] jgdx, let me check [12:38] jgdx, that looks correct to me [12:39] jgdx, as far as I’m concerned, the MR looks all good now, but if someone who actually knows the code base could take a look that’d make for a more proper review === alan_g is now known as alan_g|lunch === dandrader|afk is now known as dandrader === MacSlow|lunch is now known as MacSlow [13:12] oSoMoN, right. I think Ken has looked it, and will look at it again when time comes. [13:35] Elleo: out of curiosity how much memory does Ubuntu keyboard take up and how much do you think is acceptable? [13:37] saidinesh5: iirc it varies between about 50 and 60MB (depending on the dictionary loaded) [13:38] Ahh hmm... [13:38] saidinesh5: a big chunk of that is memory used by Qt apps for anything (something like 15-20MB as I recall), then another big contributor is hunspell loading whole dictionaries into memory [13:38] Sounds about the same for me [13:38] Yup.. Hunspell was bad... [13:39] we're vaguely considering a branch of hunspell that was developed for chromium, that could improve memory usage, but haven't looked into in detail yet [13:39] Ahh yeah seen that.. Binary format right? [13:39] saidinesh5: yeah [13:40] Yeah but i wasn't satisfied with hunspell suggestions either when it comes to swipes... [13:41] saidinesh5: it has pretty poor performance for suggestions too [13:41] Oh yes, performance is another interesting issue i noticed... It drastically sucks on phone somehow [13:42] Like the same swipes were really fast on desktop [13:42] Compared to the tablet [13:42] Haven't looked in detail though [13:42] saidinesh5: interesting, I seem to remember it being pretty responsive when I tried your plugin on an N4 [13:43] That was brute force :p [13:43] ah, was that with the precomputed swipes? [13:44] Well brute force only for the words starting with the current letter [13:44] Yup [13:44] I haven't tried comparing hunspell on arm to hunspell on x86, would be interesting if there is a big difference [13:45] For me the hunspell matcher on phone performed quite worse on longer words like in your video [13:45] But the same words were fast on desktop [13:45] interesting [13:45] I'll have to have a play with that [13:45] So nevertheless dropped hunspell matcher [13:45] might be some low hanging fruit for optimisation in hunspell [13:46] Mhm .. Also hunspell dictionaries don't account for word frequency right? [13:46] nope [13:46] we use presage for that side of things, which is typically faster too [13:47] but it can be problematic if you feed presage too much training data [13:47] Ahh ... I remember finding words like sqlite in presage , so left it for later [13:47] as it can get quite slow if its database is too big [13:47] Yup presage looks like it was designed for desktop [13:48] N grams and sqlite and stuff [13:48] and it has a few nasty bugs still (like splitting words on apostrophes) [13:48] Ahh the interesting thing android dictionaries did was [13:49] Add the punctuated words as dictionary words [13:49] So you get a lot of things for free [13:49] yeah, that's what most dictionaries do (hunspell does that) [13:50] and we have a patch for presage to make it treat apostrophes more normally [13:50] Ahh i thought hunspell affix files specified that [13:50] Ahh [13:51] saidinesh5: yeah, I'm not certain about hunspell's affix stuff; it might just be for handling less common cases [13:51] So i was thinking of just optimizing skeyer's memory usage and ship the brute force matcher for the first version [13:51] saidinesh5: but certainly commonly punctuated words are in the normal dictionary file [13:51] Ah [13:53] Hmm.. Are there any other fossy dictionary matching algos out there? [13:53] Libraries i mean [13:53] there's libaspell, not sure how its performance compares though [13:54] * saidinesh5 googles [13:55] saidinesh5: http://aspell.net/ === alan_g|lunch is now known as alan_g [13:55] Mhm [13:56] saidinesh5: their own metrics suggest its faster than hunspell, but they look rather out of date (comparing against hunspell 1.1) [13:56] Looks quite like hunspell 0_0 [13:56] The api i mean [13:56] Ah [13:59] Also any updates on merging the mir backend into Maliit? [14:00] saidinesh5: nope, I need to ping the mir folks again; I think the two people I need were on holiday last time I tried; but that was all put on hold a bit while we synced maliit and keyboard stuff between vivid and RTM [14:01] saidinesh5: but we just finished that sync, so we should be clear to start doing more experimental stuff with maliit again now [14:02] Ahh cool... I need to patch Ubuntu keyboard too so that it would be merged? [14:02] Also how does one ship a 3rd party keyboard then? [14:03] In the Ubuntu store... Click packages are a no go right? [14:03] saidinesh5: I've already got a branch that patches the ubuntu-keyboard [14:04] Ahh nice [14:04] Less work for me :p [14:04] saidinesh5: at the moment there's no way to ship a 3rd party keyboard to end users easily, you can make deb packages for advanced users to install (they have to have their device setup read-write) [14:05] saidinesh5: we're going to be working on 3rd party stuff this cycle, but not much has started in that respect yet (we're still mostly busy bug fixing for the upcoming release) [14:05] Was mostly testing stuff with the skeyer demo but i think I'll stop working on matchers and start fixing maliit plugin so was [14:05] saidinesh5: but the first step for 3rd party stuff will be 3rd party things from OEMs (i.e. that get shipped with the device), rather than that keyboard that can be installed from the store [14:06] Also can't i just specify the plugin file in server.conf? [14:06] Ahh [14:06] saidinesh5: I'd like to get us to a stage where keyboards can be shipped as clicks, but that might be a way off for now [14:06] Btw upcoming release = 2weeks right? [14:07] saidinesh5: something in that ballpark, yeah [14:07] Hmm... Like if you have a roadmap, i don't mind pitching in some help [14:09] saidinesh5: not much of a roadmap yet, I think it'd be useful to have a meeting with all the interested parties (you, me, security team, mir folks, etc.) to work things out a bit more publicly [14:09] saidinesh5: but that's not likely to be easy to organise for a little while, as everyone's attention is pretty full currently [14:10] Mhm, maybe include the sailfish folks too ? Even they maintain a private fork of Maliit apparently [14:10] Ah [14:10] saidinesh5: yeah, wouldn't hurt to chat with them; I suspect there wouldn't be a massive amount of overlap for this side of things though, since it's likely to relate mostly to click stuff and app confinement which is pretty Ubuntu specific currently [14:10] Hmm.. Yeah [14:11] Afaik that other guy's swipe keyboard is also quite close to release [14:11] But that's pretty tied to sailfish atm [14:12] Okboard or something it was.. [14:12] saidinesh5: it'd also be interesting to have a conversation about maybe using libskeyer directly in the existing keyboard as a default option; but I think that'd require a chat with legal people too (I have no clue about that side of things) [14:12] Ah sure, I'll need time for that though [14:13] Right now it's really a stupid brute force matcher :p [14:13] heh [14:14] saidinesh5: yeah, I'm thinking for the future really, my own time is pretty tied up the next couple of months with bug fixes for existing keyboard stuff and new features/improvements for our chinese keyboard stuff (which will obviously be pretty important for us in the near future too) [14:15] By tonight i think it would at least be fast enough though... And by this weekend hopefully support for frequently mistyped words.... (This works *very* nicely during prototype) [14:15] saidinesh5: cool [14:16] But yeah even i hate the fact that at the end of all this, if things go weird pirate bay release seems like the only option lol [14:16] Right now I'm almost working full time on this... [14:17] saidinesh5: yeah, software patents make everything painful :( [14:17] Mhm [14:17] on which tablets/SoC can I use Ubuntu? Is it Ubuntu-touch? [14:18] on nexus4 and 10 [14:18] err [14:18] nexus 7 and 10 [14:18] does it work on any quad-core A15 CPU? [14:18] (on 4 too, but thats not a tablet) [14:18] if you port it to it it will owrk on any SoC [14:18] Also for libskeyer to be usable on Ubuntu keyboard wouldn't it need a lot of rewriting? [14:19] Ui wise i mean [14:19] saidinesh5: not sure, I haven't looked at libskeyer in much detail from that perspective; I'd expect us to probably want to integrate it into the existing UI though [14:20] saidinesh5: so we add some swipe detection to the current stuff and pass it through to libskeyer in someway [14:20] Because currently swipe detection assumes keys are laid out in rows of 4 [14:20] ah yeah, good point [14:20] Of 10-11 i mean [14:21] Also then the long press part needs to be rewritten [14:21] To allow for swipes [14:22] saidinesh5: not massively, it can just switch to detecting long press swipes *after* a long press, otherwise look for skeyer style swipes [14:22] Technically it's still doing dictionary matches of input with swipeHints though [14:22] saidinesh5: but it's probably not worth spending too much time considering this sort of thing without some sort of legal goahead first though [14:22] Yeah [14:23] I've heard Firefox OS guys had to cancel a similar project [14:23] ah, that doesn't bode well :/ [14:23] That could also be cuz of the speed of js though :p [14:23] heh [14:24] Speed/memory constraints [14:24] Of keeping the whole dictionary in memory [14:25] The thing with hunspell was it wasn't using keyboard layouts for matching either... So the results were extremely noisy [14:31] yeah, presage doesn't do that either, which is something I'd like to implement at some point [14:35] Ah that works quite well in skeyer now [14:35] Simple really [14:35] cool [14:36] But again you'd need some kind of assumption on keyboard layout then [14:36] Now i remember why i started skeyer from scratch lol [14:36] The layouts were qml files in Ubuntu keyboard right? [14:37] saidinesh5: yep [14:37] Yeah ... Lol that's the reason.. Then it was hard to let the backend know about keyboard layouts... [14:37] That's why i had to start from scratch [14:37] saidinesh5: yeah, I figure we'll need to basically have a duplicated spatial map in a more parsable format [14:37] Instead of forking Ubuntu keyboard [14:37] Yup [14:38] saidinesh5: might be possible to rewrite the current layout stuff to autogenerate from those though [14:38] saidinesh5: whilst still leaving flexibilty for the funkier layouts like emoji, japanese, etc. to do their own thing [14:39] Hmm... [14:40] Yeah qml's children elements could help maybe [14:40] Even I'm not sure how to cleanly implement telephonic layout in skeyer [14:42] jgdx, did you figure anything out about that rtm ap test failure? [14:43] kenvandine, yes, AP generates a pressAndHold as well as a click event, sometimes. That makes it fail. [14:44] ugh [14:44] but why only on rtm? [14:44] and why AFAIK only mako? [14:45] the failure noise is driving me nuts... [14:45] * kenvandine grumbles [14:46] i have a theory that it tries to scroll (pressAndHold) and then clicks, because we use scroll_to_and_click [14:47] I haven't gotten around to changing that though (school) [14:49] who updates the universe copy of phablet-tools? [14:49] if I understand this output correctly: [14:49] http://packages.ubuntu.com/search?keywords=phablet-tools&searchon=names&suite=trusty§ion=all [14:50] it looks like it's got a stale copy of the tools, compared to the PPA [14:50] https://launchpad.net/~phablet-team/+archive/ubuntu/tools?field.series_filter=trusty [14:50] ? [14:50] pmcgowan, ^ ? [14:50] john-mcaleely, all tools should come from the ppa [14:51] hrm, so should we tell trusty people to avoid the universe copy? [14:51] https://developer.ubuntu.com/en/start/ubuntu-for-devices/installing-ubuntu-for-devices/ [14:51] pmcgowan, ^ [14:51] kenvandine, so we might need to stop abusing scroll_to_and_click [14:51] hmm, the instructions should probably be the same for all releases [14:52] is updating the trusty/universe copy an easier path? (not my call) [14:53] * pmcgowan has stale tools too [14:53] what happened to my ppas [14:53] phablet-screenshot may be broken then? [14:53] I've seen reports that is the canary for stale tools [14:53] john-mcaleely, I just used that the other day using universe version [14:54] pmcgowan, fair enough then [14:54] john-mcaleely, but, the instructions do seem wrong to me as we are not suring this back into the archive [14:54] sru [14:54] pmcgowan, ah, right [14:55] mhall119, ^^ who manages that installation page and do you agree its incorrect? [15:11] sergiusens, ^^ trusty universe vs ppa versions [15:12] john-mcaleely: no, updating the trusty archives is not easy or straight forward [15:12] sergiusens, fair enough :-) [15:12] john-mcaleely: I would just remove phablet-tools from the archive and just keep them in the ppa [15:13] sergiusens, is that something I can help with (removing from the archive?) [15:14] I'd be +1 on that [15:15] john-mcaleely: if you want to drive that great [15:15] john-mcaleely: it's a lot of paper work [15:15] sergiusens, I don't know where to start, but happy to [15:16] john-mcaleely: on #ubuntu-release [15:16] john-mcaleely: https://wiki.ubuntu.com/StableReleaseUpdates [15:17] sergiusens, thanks. I'll get reading [15:18] davmor2, the problem you found in silo 4, can you confirm that's a regression? i'm not sure who well manual selection has worked in the past, i know i've never been able to test it [15:21] kenvandine: I can check latter running regression tests now [15:22] davmor2, thanks [15:22] ogra_, do you know where the check for low disk is done and what the threshold is currently? [15:22] i seem to recall something in ofono, where if manual registration failed it fell back to automatic [15:23] davmor2, and it's likely that our UI doesn't handle that well, but that might not be new in this branch === mvo_ is now known as mvo === chihchun is now known as chihchun_afk [16:10] kenvandine: Hi! Do you have time for a couple of questions on the ContentHub (it's for a personal project)? [16:14] mardy, sure [16:15] kenvandine: what is the difference between share and export? [16:15] mardy, from a technical pov, nothing :) [16:15] it's really just difference in the user journey [16:16] so we can have apps that identify themselves as share handlers, meaning they share the content externally [16:16] like facebook, gmail, etc [16:16] kenvandine: right, I guessed so, but I wanted to double check [16:17] kenvandine: so, when I download a pic from the browser, and then I choose to open it in my app, that's "import", right? [16:17] yeah [16:17] your app is an import handler [16:18] kenvandine: can my app know that the pic came from the browser? That is, is it possible to get the app ID of the source app? [16:18] mardy, i think so... ContentTransfer.Source [16:18] i think [16:20] kenvandine: it's not in the docs... I'll check the source [16:20] kenvandine: a couple of questions more :-) [16:21] it might not be exposed [16:21] kenvandine: if I save a pic from the browser, but I don't open my app, then save another pic, and then open my app, I see only the last picture [16:22] kenvandine: is it the expected behaviour, or am I doing something wrong? [16:22] mardy, it is ContentTransfer.source [16:22] kenvandine: thanks [16:22] you are doing something wrong :) [16:23] the gallery handles that case [16:23] kenvandine: ah, true [16:23] kenvandine: I'll check [16:23] i think it just deals with multiple onImportRequested signals [16:25] kenvandine: and about those files in ~/.cache//HubIncoming/, is my app responsible for clearing them? [16:25] no [16:25] well, those should get purged by the service [16:25] you are responsible for copying it if you need it to persist [16:26] kenvandine: OK; but how does the service know that I'm done copying them? [16:26] kenvandine: do I need to set some variable in the transfer object to close it? [16:27] mardy, set the state to Finalized [16:27] or [16:28] the hub should purge it when it shuts down [16:28] even if it's not finalized [16:28] i think :) [16:28] i'm getting a little rusty on the hub... looking forward to hacking on it soon :) [16:28] kenvandine: OK, thanks, I think that's all; tonight I'll try using it more and will let you know [16:29] davmor2, cc me after you've run the tests? :) [16:29] mardy, cool [16:29] kenvandine: but so far it's been a pleasure, it's very easy to use [16:29] great! [16:29] mardy, i love hearing that :) [16:29] :-) [16:29] Elleo, ^^ [16:30] Elleo, another happen content-hub customer :) [16:30] s/happen/happy/ [16:30] :q [16:30] ops [16:33] Elleo, pmcgowan here you have => https://code.launchpad.net/~mandel/ubuntu-download-manager/support-data-uri/+merge/247317 [16:33] Elleo, pmcgowan will do the rmt mr in a second [16:34] mandel, nice [16:37] mandel: cool [16:52] Elleo: PodBird needs a "Downloaded episodes" view, so that when I'm away from my home network I can easily see what I've got available to listen to [16:52] also, does it use DownloadManager? [16:53] mhall119: yeah, good idea; and yep it uses download manager [16:53] ah, cool, so I can switch away from it while it's downloading [16:53] yep [16:54] Elleo, can you please test that with the browser? Or do I just have to open google images and try to get an img? [16:55] pmcgowan, Elleo I realized that we need to use trnnk in udm for rtm, I'm not back porting this to utopic because is not a security issue, agreed? [16:55] mandel, any progress on bug 1330770 [16:56] bug 1330770 in ubuntu-system-settings (Ubuntu RTM) "click packages rely upon tls for integrity and authenticity" [Critical,Confirmed] https://launchpad.net/bugs/1330770 [16:56] mandel: yeah, downloading any thumbnail from images.google.com should replicate it [16:56] kenvandine, I got moved to do a more important once first, will get back to you later tonight with a fix for that one [16:56] Elleo, ack [16:56] mandel, thx [16:57] mandel, just need it in rtm, not 14.10 [16:57] mandel: I'll also test it myself once the debs are built, might not get to it until the morning though, as I'll be out this evening [16:58] pmcgowan, yep, so that mr is the only one we need, once we have the deb I'll retest with the browser and will do all the paperwork to get a silo etc.. [16:59] mandel, ok [16:59] Somebody here has some doc about how software is managed in ubuntu phone? - I meant, how updates come ?... apps and core-apps are the same? The apt paradigm, with repositories and sources file, sofotware center and whatever ... is still valid in u-touch? [17:08] bfiller, pmcgowan: either of you have an rtm krillin device you can test something on? [17:08] kenvandine, indeed I do [17:09] https://code.launchpad.net/~cimi/ubuntu-system-settings/fix-1363400/+merge/247131 [17:09] pmcgowan, can you test the debs from that? [17:09] after connecting to wifi in the wizard, you should see the skip button change to continue [17:09] it didn't work for me on mako [17:10] Cimi, ^^ pmcgowan is going to test it [17:12] I am reflashing to rtm on krillin too [17:12] kenvandine, why are the versions using 20141010? [17:13] pmcgowan, where? [17:14] kenvandine, the deb versions seem odd [17:14] that can't be right [17:14] the debs i see have 15.04 in them [17:14] maybe I got the wrong zip, one sec [17:15] 0.3+15.04.20150114.1~rtmbzr966pkg014.0917+autopilot0 [17:15] pmcgowan, ^^ [17:16] must be wrong zip :) [17:16] pmcgowan, also... before you reboot go into wifi and forget the wifi network [17:16] under previous networks [17:16] just so it doesn't automatically connect in the wizard [17:18] kenvandine, so we are testing the wizard, need to enable it [17:18] yup [17:18] specific the wifi page in the wizard [17:18] make sure the button at the bottom changes from "Skip" to "Continue" when the wifi connects [17:18] after 2 secs likely [17:19] pmcgowan, we've confirmed it doesn't work on mako rtm [17:19] I will re-test on krillin in less than one hour, I have to flash back [17:20] rebooting [17:22] kenvandine, Cimi worked [17:22] pmcgowan, cool [17:23] cool [17:23] pmcgowan, thx for testing [17:23] np [17:23] * kenvandine wonders why it doesn't work on mako [17:23] kenvandine, qt or driver bug [17:23] pmcgowan, ok... so should we let this land for rtm then? [17:24] kenvandine, I am testing again now... [17:24] be good to know why it didnt work [17:24] it flashed quickly [17:24] maybe intermittent or something [17:24] pmcgowan, yeah... i tried several times on mako [17:24] well the driver tells qt "I am connected to this access point!" [17:24] qt sends this with a qml plugin [17:24] Cimi, and you're sure it worked on mako with vivid right? [17:24] it might be something wrong in the driver [17:25] kenvandine, sure sure [17:25] that's what i was thinking [17:25] kenvandine, just try the deb too [17:25] kenvandine, so there is an open bug on that not being reliable [17:25] pmcgowan, there is a bug? [17:26] kenvandine, check this one, scopes buys hit it https://bugs.launchpad.net/ubuntu/+source/qtbase-opensource-src/+bug/1413269 [17:26] Launchpad bug 1413269 in Canonical System Image "Qt network being unreliable establishing a connection" [Undecided,Confirmed] [17:27] kenvandine, it doesn't work on krillin rtm to me now [17:27] kenvandine, don't approve [17:27] :/ [17:27] kenvandine, I will dig more tomorrow [17:27] Cimi, thx! [17:27] Cimi, see that bug also [17:28] that is QNAM [17:28] it worked now [17:28] might not be the same [17:28] but it didn't 10 sec ago [17:28] unreliable... [17:28] ok, not good [17:29] pmcgowan, although the root cause of the QNAM problem could be the same as this [17:30] kenvandine, we may have a good test case [17:31] Cimi, ok, on mako i can make it work [17:31] after it connects, if i hit back [17:32] and then go in again, and it's already connected to wifi [17:32] i see continue [17:32] but it never changes for me on connection [17:32] Cimi, maybe we're not getting notified of the change? [17:32] accessPointName not getting set? [17:37] Cimi, or a race... if the networkNameChanged is emitted before the networkStatus changes? [17:38] Cimi, if networkStatus isn't HomeNetwork, accessPointName gets set to "" [17:39] Cimi, just a thought, maybe the order of the signals isn't reliably [17:39] Cimi, perhaps if you called getAccessPointName in onNetworkStatusChanged if the status was HomeNetwork? [17:39] instead of using 2 signals [17:41] Cimi, that should work, and prevent a possible race there [17:44] of course only if it's not a driver bug :) [17:53] is there a list of devices ubuntu touch may work on and probability [17:56] !devices [17:56] You can find the full list of devices, official images, community images, and works in progress at https://wiki.ubuntu.com/Touch/Devices [17:58] ok i see there is some nexus 5 support [17:59] nexus 4 and 7 still the primary focus? [17:59] the phone that goes on sale soon is the primary focus :) [18:00] ah, interesting, will that have hardware support for convergance eg dock [18:01] there is no support for cenvergence anywhere yet [18:01] *con [18:01] so even if the HW would support it ... [18:01] ok fair enough. Convergance is my interest [18:02] well, convergence of a hone OS requires that you have a phone OS first :) [18:02] heh ok. [18:02] if that is 100% rock solid the next step is to introduce desktop apps and external displays ... [18:03] ... and input devices [18:03] i guess what I really want is ubuntu on the phone but I guess the focus here is a new phone OS, which is a big UI task [18:03] convergence is a target for 16.04, you might see bits and pieces of it before but by then it should be good for production devices out in the field [18:04] not only UI [18:04] the phone OS will also be the base for the future dekstop (and perhaps server) OS [18:05] snappy is the next step after ubuntu touch ... and snappy is supposed to take over the ubuntu world as underlying system [18:06] does 'ubuntu for android' still exist (a seperate project, right?) [18:06] Cimi, ok... i figured it out... it's the currentNetworkMode [18:06] that has been put on hold [18:06] when we check that, it's NetworkInfo.UnknownMode, not NetworkInfo.WlanMode [18:06] u4a needs adjustments to the system you can only do together with a manufacturer ... [18:07] maybe there wasn't enough demand [18:07] shame, I think ubuntu for android is exactly what I'd have wanted [18:09] well, now you will get a real ubuntu on the phone *and* converged :) [18:09]