=== JanC is now known as Guest50625 === JanC_ is now known as JanC === plutoniix is now known as xiinotulp === chihchun_afk is now known as chihchun [04:47] hi [04:48] I am having trouble installing Ubuntu touch on a Lenovo A8-50 or samsung galaxy tablet [04:49] On the Lenovo I have unlocked the OEM, on the Samsung I cannot unlock it === ecloud_wfh is now known as ecloud [08:32] hi! [08:35] Hello kaisoz [08:35] greyback: I suspect (actually can measure) much of our remaining dash performance issues are CPU bound. Anyone looking into that side yet? [08:36] duflu: I believe that is on tsdgeos' plate [08:36] Cool. I know it's impractical to collect CPU profile data on a phone, but it is possible. [08:36] I haven't done it in a while [08:37] Kind of requires a binary that will run without needing interaction [08:37] it's on my plate yeah, i'm first still working on the metrics reporting [08:37] but eventually will get there [08:38] tsdgeos: I've had luck with valgrind --tool=callgrind in the past. Collect data on a phone and then visualize on desktop with kcachegrind [08:39] Also try Google Profiler (the only free one that can accurately measure time wasted in the sleeping state) [08:44] Although you can start using just desktop [08:46] duflu: greyback had some tools he demo'd at the desktop sprint in prague. dunno if you're already aware of them. SoC specific ones IIRC [08:47] duflu: for Qt apps, we can use a power profiler tool in QtCreator [08:48] popey refers to other tools I found, which profile the entire device CPU & GPU, and help you see what's going on in the whole system [08:48] I'm trying to avoid mentioning Q* things as it's entirely possible our problems are simpler than that [08:48] duflu: sure, but you were referring to the Dash [08:49] duflu: I wrote up guide on a nice profiling tool for Mali, DS-5, here: https://wiki.ubuntu.com/Touch/GPU%20profiling/Mali [08:57] popey, how is your battery life on the pro5 ? (i have the feeling mine doesnt properly go to sleep, i lose about 50% over night (rc-proposed)) [09:27] ogra_: mine's charged a lot of the time [09:27] or on charge [09:28] ah, i only charge mine when it tells me to [09:28] (and unplug when it is full) [09:28] last full charge 24 hours ago, now at 69% [09:28] oh, it was off overnight 😃 === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun [10:14] I'm wondering... why stuff like snaps when you can confine with groups and reading permissions? [10:14] there's more to it, isn't there? [10:19] brunch875, try #snappy ;) [10:20] I'm afraid I might get lit up from asking these sort of questions :Þ [10:20] nah [10:21] nobody bites over there (at least everyone pretends not to :) ) [10:23] they only tend to nibble (especially that ogra_ guy) [10:23] haha [11:22] I just used phone to flirt huehue [11:22] utouch++ === mterry_ is now known as mterry [13:00] popey: hi! I'm going through the reviewers queue and clearing out things/getting questions answered. would you mind looking at permy though? its security policy did not change [13:36] oSoMoN, heyo! What's the plan for building webbrowser-app on arm64? [13:36] hmm, so my miracast adapters arrived ... seems my session crashes when i disable the external display [13:36] mterry, the plan is to first have oxide build on arm64 [13:36] ogra_, why would you ever want to disable such a beautiful thing [13:36] oSoMoN, is there a rough ETA on that? [13:36] and the fonts are super blurry (my moniotor says it gets 1080p60 ... though that m,ight only be from the adapter) [13:36] oSoMoN, like next month, next week, etc [13:36] mterry, yeah, when making a call for example ;) [13:37] Are there any event that's propagate to Android container when booting is done? [13:37] bahm and two-finger-scrolling is completely broken ... goes into the wrong direction [13:37] peat-psuwit, only the other way round IIRc [13:37] popey: thanks! [13:38] (teh container listens to ueventd and sends an event to upstart once it is done ) [13:41] ogra_: No property set? Like, when unity finish starting up? [13:42] i dont think we seed anything into the container, no, we need it completely running for most of the graphical bits [13:42] so it has to be done even before lightdm starts the UI session [13:43] (and lightdm waits fro the event from the container ... see its upstart override job) [13:44] ogra_: Ok. I'm trying to mimic this event: "on property:init.svc.bootanim=stopped" in Android init. I guess I have to create an Upstart job to do that. [13:45] err [13:46] that would be long before anything in ubuntu started up [13:46] mterry, that’s tracked by bug #1582638 , see the comments in there [13:46] bug 1582638 in Oxide "Fails to build from source on arm64" [Critical,Triaged] https://launchpad.net/bugs/1582638 [13:46] peat-psuwit, but yeah, you can easily put that into a unity8.override post-start script snippet [13:46] mterry, more in the "next month" ballpark [13:46] peat-psuwit, just make sure the android boot finishes fully ... [13:47] oSoMoN, ok thanks! [13:47] peat-psuwit, i.e. you cant really block the container boot on this else you produce a deadlock [13:51] ogra_: I'm not sure I understand you correctly. [13:52] peat-psuwit, do youo want to make the android container boot block until this property is set ? [13:52] (this wouldnt work since the container must be done booting for lightdm to start) [13:53] ogra_: No, not at all. It's kind of post boot adjustment for CPU governor and such. [13:53] ah, well, then you should just be able to use setprop from either the lightdm or unity8 upstart jobs [13:54] (i think lightdm would be easier since that doesnt run as the user ... though a bit early perhaps) [14:04] ogra_: Ok. Thank you, then. [14:48] dobey: hi! About your comment on https://code.launchpad.net/~mardy/ubuntuone-credentials/signon-plugin-part2/+merge/293217 [14:49] dobey: SSOService::login() is (in the core currently on the images) calling Keyring::storeToken() after the authentication succeeds, [14:49] dobey: and the latter creates the account (if not already present) [14:49] dobey: the signon-plugin takes care of storing the token only, but it has no knowledge of accounts [14:50] dobey: so that dance with the Accounts::Manager to create the account must still be done somewhere [14:50] yes i know how it works currently [14:51] the account should already be created in the qml plug-in afaik [14:52] mardy: doesn't libsignon or libaccounts already have some way to flag the credentials as being invalid for an account? or is the oauth plug-in just hitting the respective server every time the account is requested by an app? [14:53] dobey: ideally, the right place for doing that is the account plugin, but it may be that for this first iteration it's simpler to keep this code in the lib [14:53] dobey: in a next iteration, I'd like to rework the account plugin so that it's QML only, using the Ubuntu.OnlineAccounts module directly (should be doable) [14:56] dobey: no, it's plugin specific; the OAuth plugin also has a ForceTokenRefresh option [14:56] ick [14:58] dobey: I've just checked, the U1 account plugin does not create the account, it relies on SSOService::login [14:58] mardy: so it just shoves another flag into PluginData and then calls authenticate()? [14:58] dobey: yes [14:59] dobey: ah, now I remember why I didn't convert the account plugin to use the signon-plugin directly: the signon-plugin does not have the functionality to create an account on the server (yet) [15:00] well i'm not worried about creating new accounts on the server yet [15:14] dobey: found a little bug with default-token-name, please don't hurry up with landing https://requests.ci-train.ubuntu.com/#/ticket/1461 [15:16] mardy: i haven't even assigned a silo to build it in yet. wanted to get the large part2 branch reviewed enough to be close enough to what we want to land, before building [15:16] and was trying to find a couple other easy things to fix along with [15:17] since our TZ overlap is not great, makes the turnaround for changes a bit annoying :) [15:17] dobey: true [15:18] mardy: why would respondWithStoredToken() always fail? [15:21] dobey: because the stored data it expects must be in the form { "token-name": { "Token": ..., "TokenSecret": ..., ... }, "other-token-name": {...}, ...} [15:21] dobey: unless you tell me that we will never ever support more than one single token name, it's better to store data in this way [15:22] mardy: it still stores data in that way [15:22] dobey: yes, but now process() can be called with an empty token name, so the lookup in the stored data will fail [15:23] dobey: I think that at the beginning of process() we should do tokenName = m_data.TokenName().isEmpty() ? buildTokenName() : m_data.tokenName(), or something like that [15:26] mardy: process() could have been called with an empty token name before too. this doesn't seem like a problem with my branch (if it is indeed a problem) [15:27] dobey: no, tokenName was a required param before [15:28] popey: is your microsoft dongle the one on offer on ebuyer today? [15:28] mardy: but it could have been an empty string [15:29] oh i guess validateInput checked that it wasn't empty [15:30] dobey: well, probably, but than that was a fault with my code; my intention was clearly to have it as a mandatory non empty param [15:30] mcphail, you want the v1 version ... that works for sure [15:30] mardy: anyway, yes, currently we only have the one token name for any one device [15:30] not sure if/how well the v2 works ... i just got two v1 ones ... both work OOTB [15:30] dobey: ok; but do you agree that we could set it right at the beginning of process, if it's empty? [15:31] ogra_: ta. Thought it was a bit pricey anyway [15:31] it is ... 40€ or so [15:33] Hello, so I am thinking of switching android to ubuntu touch [15:33] mcphail: link? [15:33] can you install glibc.i686 for example on the device after you have ubuntu touch as your os? [15:33] mcphail: https://www.amazon.co.uk/gp/product/B00J9C2JDG is the one I have [15:34] mardy: probably not in process(); though i think process() probably needs to be a bit more complex, in order to actually use the signon plug-in [15:34] popey: hold on - just deleted the email after ogra_ told me to look at the other version... :) [15:34] hah [15:35] well, that pic shows v1 [15:35] afaik the "big part" is smaller in v2 [15:35] mardy: if i set it in responseWithStoredData(), it should satisfy your complaint here, no? [15:35] http://www.ebuyer.com/743240-wireless-display-adapter-v2-hdmi-usb-in-p3q-00003 is the v2 [15:35] yes, smaller "big part" [15:36] dobey: yes [15:36] popey: http://m.ebuyer.com/743240?utm_source=2016-05-27&utm_medium=campaign_email&utm_campaign=B2C_%28Friday%29_Openers [15:36] dobey: please apply this to test: http://paste.ubuntu.com/16733924/ [15:36] mcphail, thats a v2 [15:38] can someone please tell me if having ubuntu touch on your device makes it possible in installing 32 and 64 bit libraries? [15:38] you can tell it's a v2 by the price :) [15:38] ogra_: yep. v1 much cheaper anyway [15:38] 34vs50 quid [15:38] grr, pastebin is broken [15:39] Laris, that culd theoretically work in a libertine container if you would have a 64bit rootfs (which we do not have yet) ... then you could enable 32bit execution [15:39] popey: yep - 20 quid more and you could get the little asus nettop [15:39] aha. [15:39] thanks ogra_ [15:40] Laris, note though that the only x86 rootfs/image we currently have is for the emulator [15:40] al supported devices are ARM [15:40] this is what I wanted to find out [15:41] I made the mistake into unrooting my device and getting the latest kernel [15:41] you can do a normal ubuntu install and install the unity8 mir session though [15:41] in case your device is PC compatible [15:41] I can't root it again as there is nothing available for this kernel and the other way to have uid-0 is to lose my knox certificate and other [15:42] it's a samsung a3 [15:42] ah [15:42] I want to try some pentesting tools and I guess I'll have to wait some more or get an older patch of my actual os [15:43] hmm, so casting a 1080p youtube movie from my pro5 to the TV still gets me quite some frame dropping ... [15:43] anyway, thanks, it helped [15:43] popey, ^^ did you try that ? [15:43] also we only have an i386 image for emulator right, no amd64? [15:43] dobey, yeah, i think i mentioned that above [15:43] ogra_: you just said "x86" :) [15:43] yeah :) [15:44] (i could have said "intel" but x86 makes me sound so much more professional :P ) [15:44] "intel" wouldn't have made it more clear than "x86" does, about whether it's 32 or 64 bit [15:45] dobey: this is my proposal, feel free to modify it at will: http://bazaar.launchpad.net/~mardy/ubuntuone-credentials/default-token-name/revision/243 [15:45] its "all x86 sstems" [15:45] dobey: but the test case is fine, on that I'm sure :-) [15:46] i don't know why we don't build an amd64 emulator image [15:46] * mardy is EOD [15:47] cheers [15:52] ogra_: no [15:52] ogra_: I think the dongle scales down to 720p [15:53] well, the one attached to my monitor tells me 1080p60 ... on the monitor side though ... could be it scales internally [15:53] fonts are definitely very hard to read here [15:53] yeah, same [15:53] on both, monitor and TV [15:53] looks awful [15:53] queschiun: Can I put a reminder on calendar so that it reminds me with an alarm? [15:54] or should I use the tasks app? [15:54] brunch875: calendar events can have alarms, yes. [15:54] hoorraaay [15:55] since forever :) [15:56] Now how about I add it on my laptop's calendar program. Will it sync? [15:56] cause that'd be bombastic [15:58] meh desktop calendar doesn't sync with anything [15:58] ? [15:59] depends which one you use [15:59] default, didn't install anything. That's gnome's, isn't it? [15:59] * ogra_ uses evolution and syncs through google ... [15:59] google works "fine" in evolution or gnome-calendar [15:59] dunno, never used it ... most of my machines were installed like 5-6 years ago and since upgraded [15:59] I'm really tempted to boot thunderbird and get evolution [16:00] only 5-6 years? n00b :P [16:00] telegram is accumulating badges, now it shows 3. cant make them go away (talking about the left dock) [16:01] anyone got this issue? [16:01] dobey, well, i have a warty laptop, but that hasnt been booted in years :) [16:01] The green badges? I have those too. I never quite understood what they meant [16:01] s`, file a bug [16:01] brunch875: it's red [16:01] * ogra_ has seen that before as well [16:01] I believe telegram doesn't mark messages as red if you read them someplace else [16:02] brunch875, well, it should remove the badge when you open the app locally ... but it doesnt do that every time [16:02] i mean, the badge color is red, not green :P [16:02] (sometimes it does, dometimes it doesnt [16:02] ) [16:02] never worked for me ogra_ [16:02] like NEVER [16:02] :D [16:03] oh wait [16:03] works for me about 50% of the time ... perhaps i'm lucky [16:03] oh wait, I actually had 4 unread messages [16:03] ok i understood === chihchun is now known as chihchun_afk [16:04] s`, are you using the latest version? [16:04] yes [16:04] OTA-10.2 [16:04] hmph maybe it's different for me because sitting on rc-proposed [16:04] and latest tg version [16:05] how do you check ota version, again? [16:05] system settings [16:05] about [16:05] something like that [16:05] not finding it there [16:05] only on stable [16:06] ogra_ spills knowledge everywhere [16:06] thanks! [16:06] (OTA version for rc-proposed would be nonsense ... rc-proposed is daily builds) === chihchun_afk is now known as chihchun [16:52] I got age of wonders 3 yesterday. Such a fantastic game. Being turn based it would be simply the best game for utouch. By far! [16:53] Someone will have to nag the developers to recompile for arm [16:53] if mc-phail asks, I didn't say anything :Þ [16:54] * brunch875 clings to hope [16:55] Actually, it could be a good place for them to monetize. The ubuntu store === chihchun is now known as chihchun_afk [18:00] * ogra_ LoLs [18:00] popey, seen that http://withinrafael.com/building-a-microsoft-wireless-display-adapter-base-image/ ? seems there is linux inside [18:49] Help/ [18:50] need help [18:50] installing touch os on samsung g note [19:44] Hi. I just got a Meizu Pro 5 ubuntu edition, and found some things which apear strange to me. First of all, I want to use an ext4 formatted SD Card, but the Phone can't authomatically mount it. I tried to modify /etc/fstab, but the file is mounted on a tmpfs, and any change vanishes after a reboot. How is the /etc/fstab file generated on an ubuntu phone, and how can I change it? Secondly, I read that [19:44] the ubuntu phones aren't updated using apt-get, is this really true? If so, why? And lastly, I found an LXC Container named 'android' on my phone, what purpose does it serve? [19:47] DPA: with regards to the latter, some android drivers are required for the hardware. There aren't open equivalents [19:48] DPA: the devs haven't decided on the final strategy for the sd card. You can format as ext4, but you'll be swimming against the tide [19:49] mcphail: Many thanks, this axplains a lot. [19:49] DPA: yes, apt-get is not how the phones are updated. the root fs is read-only. system updates are applied via image diffs [19:50] DPA: the core images are update as a whole, rather than piecemeal with apt get. If you need apt, safest way is with a chroot [19:50] or a container via libertine [19:51] brunch875: am I getting a bad reputation on here? ;) [19:52] Impossible! you're one of the coolguis, mcphail! [19:53] ogra_: oooh! Didn't know it had linux inside, maybe we can make an image for it which has an ubuntu splash :) [19:54] popey: the one good thing about windows is it doesn't have plymouth! [19:55] hah [19:56] seems no documentation for flashing the thing though [19:56] If image diffs are used for system updates, how does this look like from a security perspective? How can I check that no one manipulated the images with the Images? === curtness is now known as strahtw [20:01] DPA: the images are cryptographically signed on our server [20:03] also delivered over HTTPS [20:57] bluetooth is not working on my pro5 with rc-proposed. is that normal? [21:25] do someone have some basic knowledge about sony devices ? [21:33] mcphail: On the SD card front, why are there issues using ext4 or any other filesystem Linux supports?