[00:01] <pseudonymous> mcphail & popey: I just wanted to thank you guys profusely. Of course it's impossible to say to which extent the bugs are gone, but I've noticed that some really aggravating ones are seemingly absent. Your help was much appreciated
[00:02] <popey> I did nothing.
[00:02] <popey> But thanks :)
[00:03] <mcphail> pseudonymous: I did even less...
[00:03] <popey> ^5
[00:03] <mcphail> pseudonymous: current version are so much better than the old ones
[00:05] <pseudonymous> So far it seems nice. The small lockups which would happen occasionally during browsing haven't manifested themselves yet and my podcast app (podbird) can actually, finally fetch newer episodes of a certain podcast I've dearly missed. Here's to hoping it won't lock up mid-podcast
[00:56] <rbasak> Ah, Bluez5! I've been looking forward to this.
[02:09] <trelane> If I buy an unloked Meizu MX4 can I flash ubuntu on it?  (are they planning to make more of these with ubuntu on it?)
[02:10] <lotuspsychje> is the r350 also ota9 on nexus7?
[02:34] <popey> trelane, hard to know, as some are flashable, some aren't
[02:34] <popey> and you can't tell if it can be before you buy it
[02:34] <trelane> popey, Excellent.
[02:34] <trelane> so russian roulette
[02:34] <trelane> crap
[02:35] <lotuspsychje> trelane: how about buy a second hand nexus4 and flash it
[02:36] <trelane> lotuspsychje, perhaps it's my irrational fear of poor unity performance... but yeah I'd like more horsepower
[02:36] <lotuspsychje> nexus devices perform very well actually
[02:37] <trelane> yeah but octocore :)
[02:37] <lotuspsychje> :p
[02:37] <lotuspsychje> trelane: maybe one day we will get more of choices
[02:37] <trelane> lotuspsychje, I really want to review this thing
[04:51] <MCMic> How long is the update supposed to take? It says «Installation…» with no further details.
[04:54] <MCMic> Oh, that’s a button… It’s not clear ^^
[04:57] <MCMic> The «…» should be removed, it made me think the installation was taking course
[04:59] <MCMic> Yay Thai characters works now :-)
[05:08] <MCMic> Hum, the GPS switch is not in the position panel anymore?
[08:18] <BlackJohnny> hello devs!
[08:18] <BlackJohnny> anyone here that developed the latest version of qtubuntu-camera?
[08:19] <BlackJohnny> i have an issue with the viewfinder resolution after OTA9
[08:20] <BlackJohnny> the resolution i set within my app gets changed
[08:20] <BlackJohnny> and I cant seem to find why
[08:21] <BlackJohnny> I can change it back later and it remains like that but that "later moment" is not so convenient
[08:21] <BlackJohnny> i need to find a better later moment :)
[08:21] <BlackJohnny> qtubuntu-camera or maybe qtvideo-node
[08:21] <zetheroo> got an OTA update this morning :)
[08:22] <zetheroo> Anything I should be looking out for in particular?
[08:22] <BlackJohnny> bugs :)
[08:22] <zetheroo> doh :(
[08:22] <BlackJohnny> i am kidding
[08:22] <BlackJohnny> sorry about that
[08:22] <BlackJohnny> it is actually faster
[08:22] <BlackJohnny> and the icons look better
[08:22] <BlackJohnny> and the engine is more polished
[08:24] <zetheroo> icons look exactly the same
[08:25] <zetheroo> nothing seems faster ..
[08:25] <zetheroo> as for a polished engine .. will have to take your word for it ;)
[08:25] <BlackJohnny> the outer border/shadow is better on top
[08:26] <BlackJohnny> u want a release notes to read some of the things they changed?
[08:26] <zetheroo> sure
[08:26] <MCMic> zetheroo: It’s definitely better on scope switching, it does not refresh the whole scope as before
[08:26] <BlackJohnny> https://wiki.ubuntu.com/Touch/ReleaseNotes/OTA-9
[08:26] <BlackJohnny> indeed
[08:26] <BlackJohnny> that too
[08:27] <MCMic> zetheroo: And they added thai support which i badly needed ^^
[08:27] <zetheroo> MCMic: ok, I only use 2 scopes so didn't notice that
[08:27] <zetheroo> MCMic: haha :D
[08:27] <BlackJohnny> the apps scope u use it all the time
[08:27] <BlackJohnny> refresh that
[08:27] <BlackJohnny> pull down
[08:27] <MCMic> These are the changes I was waiting for but there are others.
[08:27] <BlackJohnny> and you will see it
[08:28] <BlackJohnny> it does not flicker on replacing the new content
[08:28] <zetheroo> BlackJohnny: It never flickered before for me
[08:29] <BlackJohnny> zetheroo, flicker maybe is the wrong term. But you could have seen that the icons were replaced/redrawn
[08:29] <zetheroo> ok
[08:29] <vayan> The screen wake on notifications is new ? how to disable that ?
[08:30] <tathhu> No it isn't :o
[08:32] <sil2100> BlackJohnny: hey! As for qtubuntu-camera then one of the people responsible for changes there is Kaleo but not sure if he's around right now
[08:33] <BlackJohnny> sil2100, i have compose an email for him but I need some info quite fast
[08:33] <zetheroo> I am guessing that with 16.04 some major fixes are set to be implemented ... ?
[08:33] <BlackJohnny> sil2100, i have a published app that missbehaves badly
[08:33] <sil2100> Ouch ;/
[08:49] <BlackJohnny> sil2100, oOo ... i now realize who answered me :)
[08:50] <sil2100> BlackJohnny: oh, what do you mean?
[08:50] <BlackJohnny> sil2100, I read all your emails and wait for OTAs :)
[08:50] <sil2100> Glad to hear that :)
[08:50] <sil2100> Thanks!
[08:50] <BlackJohnny> sil2100, I thank you ! :)
[08:55] <tathhu> :P
[08:56]  * tathhu waits working bluetooth on nexus 7 
[10:00] <JamesTait> Good morning all!  Happy Thursday, and happy Data Privacy Day! 😃
[10:01] <mzanetti> haha
[10:21] <willl> hi
[10:23] <willl> Is the a chance we see Touch in a WileyFox Swift one day (I can't find if hw/sw prerequisites are met)?
[10:32] <willl> hi
[10:33] <willl> Hi Patricia
[10:43] <popey> willl, if someone ports to it, sure
[10:45] <lotuspsychje> https://www.youtube.com/watch?v=95OdkWYL61k
[10:45] <lotuspsychje> OTA9 video :p
[10:46] <core_t> nice video!
[10:46] <core_t> so many apps installed!
[10:47] <slvn_> lotuspsychje,  hello! quick question ... could you try negative space on OTA-9 to see if it works. thanks
[10:48] <lotuspsychje> slvn_: what you mean space
[10:48] <lotuspsychje> core_t: pretty nice made indeed heh
[10:49] <slvn_> lotuspsychje, you have an app/game (in the video), called "Negative Space" !
[10:49] <lotuspsychje> slvn_: its not my video, just sharing it here :p
[10:50] <slvn_> ok, sorry !
[10:50] <lotuspsychje> slvn_: but ill try the game on my bq
[10:50] <slvn_> if you have OTA-9, please try it, that would be great !
[10:50] <lotuspsychje> slvn_: are you the writer?
[10:51] <slvn_> yes, with an idea and help of a former colleague
[10:52] <lotuspsychje> slvn_: works like a charm :p
[10:52] <slvn_> can you confirm you have OTA-9 ?
[10:52] <lotuspsychje> slvn_: yes sir, updated to OTA9 yesterday on bq 4.5
[10:53] <lotuspsychje> slvn_: sound & gameplay working nice
[10:53] <slvn_> great ! I was worried ! thanks
[10:54] <lotuspsychje> slvn_: you dont have a device yourself?
[10:54] <slvn_> no ... I used to have a Nexus 10 ...
[10:54] <lotuspsychje> kk
[10:55] <slvn_> i tested on ubuntu on my nexus 10, but that was 1 year ago. then ubuntu get obsolete
[10:56] <slvn_> I just updated it a few month ago, before OTA-8
[10:57] <lotuspsychje> ok
[10:58] <lotuspsychje> http://www.omgubuntu.co.uk/2016/01/best-new-features-in-ubuntu-phone-ota-9
[11:28] <zetheroo> show your support ;)  https://productforums.google.com/forum/#!topic/gmail/_1O8YkvZRAQ;context-place=forum/gmail
[11:44] <zetheroo> for google hangouts: https://productforums.google.com/forum/#!topic/hangouts/4hUouBqoYjg;context-place=forum/hangouts
[12:54] <popey> dobey, Saviq what needs doing on bug 1472186 ?
[12:55] <Saviq> popey, releasing it IIUC
[12:55] <Saviq> popey, https://code.launchpad.net/~dobey/indicator-network/lib-recommends/+merge/283213
[13:00] <popey> Saviq, who's that fall to?
[13:00] <Saviq> popey, I'd imagine dobey will land it soon
[13:00] <popey> ok
[13:00] <popey> thanks
[13:02] <matv1> this has been bugging me for a long time: most scopes has a search input. previous search results are shown when you start typing. But these search sugestions are from across all scopes.
[13:02] <matv1> In my experience it is seldom usefull to have search sugestions presented that i did in another scope.
[13:03] <matv1> Or is this purposely designed like that?
[13:03] <matv1> didnt find any ubuntu-ux bug on that
[13:40] <mardy> mpt: hi! I saw your changes to the OA wiki, about the account authorization prompts
[13:41] <mpt> mardy, that was quick. :-) I just asked a question on bug 1522360.
[13:41] <mardy> mpt: could you please clarify, how should the dialog appear? Always on the background, or (if the app is focused) on the foreground?
[13:43] <mpt> mardy, the latter. Otherwise you’d be sitting there wondering why nothing was happening
[13:43] <mardy> mpt: but only if the requesting app is on the foreground, right?
[13:44] <mpt> mardy, or if you’re doing nothing else at the time. Tracking that is the window manager’s job.
[13:46] <mardy> mpt: that sounds good; however, I'm afraid that unity8 does not support that yet; would you mind describing the behaviour explicitly in the wiki, so that I can point it to the unity8 guys in case I need to file a bug?
[13:59] <tvoss> mhall119, running 5 late
[14:04] <mpt> mardy, done. <https://wiki.ubuntu.com/OnlineAccounts?action=diff&rev2=28&rev1=27>
[14:05] <dobey> popey: pete-woods has that in a silo. but really it's not a supported part of the SDK. the QML API is though i think
[14:05] <mhall119> mariogrip: tvoss having to reboot, will join shortly
[14:06] <mariogrip> mhall119: He's already here
[14:11] <mardy> mpt: thanks! That sounds good; just to be absolutely clear: does it mean that if the dialog appears when you are busy with other apps, it should appear in the background (and maybe trigger the launcher's in, to notify you that a new window has appeared)?
[14:12] <mardy> mpt: and I just commented on bug 1522360, it's long but I hope it's clear
[14:12] <mpt> mardy, no, it doesn’t…
[14:12] <mardy> mpt: mmm... so, what should happen in that case?
[14:13] <mpt> oh, I misunderstood your if as iff :-)
[14:13] <mpt> mardy, yes, in that case it should appear in the background.
[14:13] <mardy> mpt: perfect, thanks :-)
[14:16] <mpt> mardy, thanks for the follow-up. Is it ever/often/always the case that an OAuth-token-requiring service will show you the first Web page in the flow diagram, and then immediately show you the second one?
[14:16] <mpt> e.g. “Register or sign in” and then “Confirm access for Dekko”
[14:37] <davidcalle> sil2100: heya, I'm adding some test packages on my phone and I keep hitting No space left on device, do we have cached images, or anything that I can remove? I've already cleaned up apt cache.
[14:37] <davidcalle> deb packages*
[14:39] <sil2100> hmm, hard to say if there's any easy way to free up space
[14:40] <davidcalle> sil2100: ok, I might end up removing libertine things for now
[14:45] <mardy> mpt: sorry for the late reply; no, it can be either "Register or sign in" or "Confirm access for Dekko", or both. Any combination can happen :-)
[14:46] <mardy> mpt: or actually, I believe that "Confirm access for Dekko" should always be there
[14:47] <mardy> mpt: optionally preceded by the login request (depends on whether the cookies have expired or not)
[14:52] <mpt> oh what fun
[14:55] <awe_> mpt, sorry for not ack'ing your bug link.  I'll try and take a look later on...
[14:56] <mardy> mzanetti: hi! Do you have a minute for a unity8 question?
[14:57] <mzanetti> mardy, sure
[14:57] <mardy> can you please read the second paragraph (starting with "Apart...") after the picture in https://wiki.ubuntu.com/OnlineAccounts#App_access ?
[14:58] <mardy> mzanetti: I wonder is unity8 supportsthis kind of spontaneous windows
[14:58] <sil2100> davidcalle: I now learned that in theory you might try removing the mir-test-tools package ;)
[14:58] <sil2100> davidcalle: you could try and see what happens, but it's 35 MB of binaries and the name doesn't seem like it's anything necessary during runtime
[14:59] <davidcalle> sil2100: sudo apt-get purge *-tools *-dev :p
[14:59] <mardy> mzanetti: so, depending on how the user is interacting with the device, we want to have windows spontaneously appearing either in the foreground or in the background
[15:00] <mzanetti> mardy, not sure I completely understand it yet
[15:00] <mzanetti> mardy, what do you mean with appearing in the background?
[15:00] <mzanetti> mardy, so far it seems to me that this is a regular trust prompt
[15:01] <mardy> mzanetti: in most cases (when the requesting process is a regular app) yes, it is
[15:01] <sil2100> ;p
[15:01] <sil2100> davidcalle: ok ok, I wouldn't go as far as that! ;)
[15:01] <mardy> mzanetti: but think of account-polld, when it's checking gmail every 5 minutes but you don't have the gmail webapp or dekko running
[15:02] <mardy> mzanetti: in that case, we have a backround ui-less process which needs to popup a dialog
[15:02] <mardy> mzanetti: and given that you don't have any app window to parent it to, it's parentless
[15:03] <mzanetti> mardy, right... we don't support that atm. and the design guidance was quite clear that there are no parentless dialogs allowed (which we obviously already violate for the wifi password entry)
[15:03] <mzanetti> mardy, but trust prompts can't do that, no :/
[15:04] <mzanetti> mardy, but I would think that the polld stuff should only do polling once it has everything it needs
[15:05] <mardy> mzanetti: true, but at some point the access token will expire
[15:05] <mzanetti> right
[15:05] <mzanetti> odd
[15:05] <mardy> mzanetti: and we don't want account-polld to silently fail
[15:05] <mzanetti> yeah, I totally see that
[15:05] <mzanetti> seems a mismatch in the specs
[15:06] <mardy> mzanetti: how to proceed? Do you want me to file a bug?
[15:06] <mzanetti> mardy, so from a technical point of view, currently all we could do is to trigger a notification that informs the user that stuff expired and he needs to go to settings to re-grant it.
[15:06] <mzanetti> mardy, going forward, I guess we'd need to talk to design what they want
[15:07] <mzanetti> Saviq, you might be interested in this ^^
[15:07] <mardy> mzanetti: we had that, design asked us to remove it :-)
[15:07] <mzanetti> Saviq, in short, if a OA token expires, design wants a trust prompt to pop up and reauth. however, no parentless dialogs etc...
[15:07] <mardy> mzanetti: the rationale was "don't ask the user to do something, do it for him"
[15:08] <Saviq> mzanetti, it looks like we will have parentless dialogs after all
[15:08] <mzanetti> Saviq, for trust prompts too?
[15:08] <Saviq> mzanetti, for whatevers, if that's what design wants
[15:08] <Saviq> mzanetti, it won't be a trust prompt per se
[15:08] <Saviq> as it's not about giving a particular app access to $resource
[15:08] <Saviq> but rather making $resource available to apps that already have access to it
[15:09] <mzanetti> i.e. it should ideally be the exact same dialog. but yeah, I guess we can put it once into a dialog, and the other time in a whatever-popup
[15:09] <mzanetti> s/dialog/trust-prompt/
[15:09] <mzanetti> mardy, ok... so. in short-term, how about "Qt.openUrlExternally("settings://accounts/"). In the long run we'll get you a parentless dialog
[15:12] <mardy> mzanetti: no, I need a bit more context than a link; but I could just start the app myself with the infamous --desktop_file_hint, I guess
[15:13] <tvoss> mardy, just encode it in the url :)
[15:13] <tvoss> mardy, @context, instead of starting it manually
[15:14] <mardy> tvoss: given that it's atemproary solution, I'd rather go for what causes the minimal amount of changes in the code: now we are starting the trusted prompt as a child process, so the easies way is to replace the --socket param with the desktop file
[15:25] <tvoss> mpt, so what happens if authentication fails multiple times?
[15:26] <mpt> tvoss, fails at what step? Inside the Web frame?
[15:27] <tvoss> mpt, in the background, would the prompt open again and again?
[15:31] <mpt> tvoss, {the prompt and the Web dialog that comes after it} don’t go away until either you’ve succeeded, postponed, or denied access
[15:31] <mpt> If you’ve succeeded, no problem
[15:31] <mpt> If you postpone, it comes back later
[15:31] <mpt> If you deny, it doesn’t come back unless/until you go into System Settings and flick the switch
[15:32] <mpt> Hmm, the diagram needs arrows starting from the “Cancel” buttons, doesn’t it :-)
[15:34] <dobey> i don't think we should ever do pop-under for auth
[15:35] <mpt> mardy, tvoss, what do you think should happen if you “Cancel” at the Web frame? Go back to the previous step, or assume you don’t want the app to have access after all?
[15:35] <mpt> The trouble with assuming you don’t want the app to have access is that the service might just be down for a minute
[15:35] <mpt> or your Internet connection might just be too slow to load the Web page
[15:36] <mardy> mpt: I'd treat "Cancel" as "postpone"
[15:37] <mpt> Ok, that works in the (re-)auth case
[15:38] <mpt> And it works in the local access case, if you have the account set up already
[15:40] <mpt> I’m not sure what happens in the no-accounts-of-that-type-yet case
[15:40] <mpt> I guess we need an explicit “Try again later” error for apps to handle?
[15:41] <mpt> as opposed to being denied
[15:42] <dobey> cacnel == didn't log in.
[15:42] <mpt> dobey, when tvoss said “in the background, would the prompt open”, he wasn’t talking about the prompt opening in the background (necessarily), but the auth taking place in the background (e.g. an e-mail checking service)
[15:43] <dobey> mpt: so you think that should cause a pop-up to log in, when the user is doing something else, or nothing at all?
[15:46] <dobey> mpt: imo, we should only ever pop up a log-in dialog in the source of user-directed action where the flow requires an account to continue; if no account is required to proceed further, things should proceed, and provide a button or some other method in the UI to allow the user to direct the system to log into an account
[15:46] <mpt> dobey, the alternatives are worse. Either we silently fail (causing missed messages, appointments etc), or we put up a notification that gives you unavoidably complex instructions of what to do but doesn’t let you do it.
[15:46] <mpt> If you could be more specific than “some other method”, I’d be delighted :-)
[15:47] <dobey> mpt: it is application/dmoain-specific, which is why i said "some other method"
[15:48] <mpt> Give two examples, then?
[15:48] <mpt> (to demonstrate why it needs to be custom)
[15:50] <dobey> well, system-settings update panel for example shows a button (and i think that is appropriate there). background process like calendar sync could show a notification, and the action on the notification could be to proceed through login process, so when user opens the notifications menu and taps on the "You need to sign in to Google for calendar sync" or whatever notification, the user just logs in and goes about their business 
[15:50] <mpt> dobey, this is highly analogous to the situation where you are connected to a wi-fi network, and its admin changes the network password. Either you get a dialog to reauth, unrelated to whatever app you happen to be using; or … you aren’t connected any more and might not realize why.
[15:52] <dobey> imo that shouldn't pop up a dialog either
[15:53] <dobey> i mean, we have cases today where that dialog pops up, and the password wasn't changed or anything.
[15:55] <mpt> dobey, you’re thinking of bug 615239, but even when that bug is fixed the situation I’m describing still needs handling
[15:56] <mpt> And the only difference between “a notification [with] the action on the notification … to proceed through login process” and the wi-fi auth dialog, is that the wi-fi auth dialog is less annoying
[15:56] <mpt> because you can put it behind what you’re doing if you want to, which you can’t with a notification.
[15:57] <dobey> mpt: i'm not saying it doesn't need handling. i'm saying that popping up any dialog as the result of a background thing, is poor UX
[15:57] <mpt> And I’m saying it’s annoying, but much less bad any alternative I’ve seen proposed.
[15:57] <mpt> +than
[15:58] <dobey> and wifi auth isn't the only problem there; what happens when vpn auth fails?
[15:58] <ulrichard> It's great that I can finally use the bluetooth keyboard with my aquaris, but where can I configure the keyboard layout?
[15:58] <dobey> well, the wifi is slightly less annoying than a full screen window, sure
[15:58] <mpt> yep, VPN auth failing is one of the same class of situation
[15:59] <dobey> but if i'm sitting in a browser, logging into some web site, some window popping up in front and stealing focus, is a problem
[15:59] <dobey> regardless of the reason for it
[16:04] <mpt> So people have been saying, for the past 30 years or so, with alternatives only nibbling around the edges
[16:06] <matv1> im 100% with dobey though, fwiw :)
[16:08] <NwS> Heya guys
[16:08] <NwS> Happy new year and all :P
[16:08] <mpt> It’s easy to be against things
[16:08] <NwS> Any updates on any new models yet?
[16:08] <matv1> haha i know but dobey has a fair solution i think
[16:09] <mpt> Really? Do tell me what it is :-)
[16:09] <dobey> i can't. my phone interrupted me and told me it's time to get lunch. so i must go do that.
[16:11] <matv1> i was assuming he was hinting at a notification center signal. at least not something that gets in the way.
[16:11] <nik90_> hey guys, I'm having trouble getting OTA-9. Despite several reboots and constant checks, I haven't got OTA-9 on my bq e4.5 device.
[16:12] <mardy> mpt: I would opt for opening the dialog always in the background, and have a small and touch-unsensitive visual indication that there's something needing your attention
[16:13] <mardy> mpt: a bit like the launcher sliding in and shaking the icon, but even less intrusive, if possible
[16:14] <matv1> maybe like the little red number when theres an update ?
[16:15] <mardy> mpt: maybe the notification indicator could become red or something, without actually showing a notification
[16:15] <mardy> mpt: but if you open it, you'd see the notification inside it; and tapping it would bring the dialog to the foreground
[16:15] <mpt> mardy, that’s the equivalent of going into Compiz Config Settings Manager and setting focus-stealing prevention to “High”. It’s fine if you’re the sort of person who notices every bounce and glow and indicator around the edge of your screen. Unfortunately most people don’t.
[16:16] <mpt> The smaller the screen is, the more likely you’ll notice a change, but on the other hand, the less space there is on screen to present the change.
[16:16] <mardy> mpt: right, I admit I was thinking of the phone only; on a bit screen we might want something more visible
[16:16] <mardy> s/bit/big
[16:17] <matv1> also, considering things like vpn notifications, users might not even care to do something about them at that moment
[16:18] <tathhu> nik90_, you could atleast flash it if you have acces to ubuntu pc and usb-cable :P
[16:27] <mpt> matv1, the badge on System Settings is for a situation that is (almost always) much less urgent. You might have even removed System Settings from the Launcher, so that the badge is visible only in the Dash.
[16:27] <mpt> (At least, once the Unity developers finally get around to implementing badges on apps in the Dash.)
[16:33] <matv1> mpt you are right. But i meant something as unobtrusive as that. I should however always show. Is it possible to turn the notification center icon red? Like along mardys lines
[16:36] <matv1> didnt mean to elbow my way into the conversation btw :) this stuff just feels important to me.
[16:37] <matv1> because if you start adding up all the possible scenarios that might trigger such a popup in the more distant future, we might get to see a lot of it
[16:37] <dobey> mpt: who is determining that urgency?
[16:41] <matv1> dobey right. as i said, sometimes users might not even care to take any action. depening on the scenario
[16:42] <matv1> anyway that me butting out now :)
[16:42] <dobey> matv1: well, my point is that the only one capable of determining urgency, is the user. treating everything as urgent will only make the user experience worse
[16:53] <mpt> dobey, in the case of software updates, it’s engineers in general and the security team in particular. If Ubuntu Touch users were under constant attack from ransomware and botnets and we were constantly issuing unphased updates to stymie them, updates would be more urgent than just a badge.
[16:55] <dobey> mpt: system updates do not require an account. what is the suggested solution for such critical security updates? are we going to pop up an overlay that requires people to update before they can continue using their phones?
[16:56] <dobey> and why is my not getting twitter notifications now because my token expired that important? or why isn't it that important? why does ubuntu get to make that determination, and not me?
[16:57] <mpt> I don’t see how requiring an account or not makes any difference…
[16:57] <dobey> well i thought we were talking about accounts popping up login dialogs, not system updates
[16:58] <dobey> they are two very different problems, and different problem domains
[16:58] <mpt> Whether something requires an account or not is orthogonal to how urgent it is.
[16:59] <mpt> For example, iOS treats system updates more urgently than we do: as well as the badge on the icon, you get a weekly (?) dialog reminding you that the update is available.
[16:59] <dobey> that does not sound especially urgent
[17:00] <mpt> No, it’s just one step up the scale from us
[17:00] <mpt> Going up the scale, you could make that dialog more frequent
[17:01] <dobey> so to protect against ransomware, we may need to make system updates behave as ransomware?
[17:01] <mpt> Even further, you would head into obnoxious Windows territory, where they say something like “Oh hi there. Unless you cancel, this computer will restart to install updates in 59 ... 58 ... 57 ...”
[17:01] <dobey> that is a level of annoyance, not urgency
[17:01] <mpt> Annoyance is the effect, not the rationale
[17:02] <dobey> it is the implementation
[17:02] <dobey> it is trying to install urgency upon a matter via annoyance
[17:03] <mpt> Instill? Perhaps.
[17:03] <dobey> yes
[17:04] <mpt> But no designer gets up in the morning saying “I want to annoy people today”
[17:04] <mpt> I bet there were screaming arguments at Microsoft about the automatic restarting
[17:04] <dobey> well, it's the effect, not the rationale :)
[17:05] <mpt> With designers saying “How dare we” and security engineers pointing at graphs of infection rates
[17:06] <dobey> well, screaming matches are certainly one way to get poor design into a product
[17:07] <dobey> but back to the topic of account logins. no login error is urgent, unless it occurs during the course of the user attempting to log in.
[17:09] <dobey> missing messages or appointments are separate problems
[17:09] <mpt> In 2009 I wrote up a simple matrix of urgent/not vs. interactive/not, and techniques for each combination. <https://wiki.ubuntu.com/NotificationDesignGuidelines>
[17:10] <mpt> The simplest reason that account authorization failure can be urgent is that it might result in you missing calendar appointments.
[17:11] <dobey> the only person capable of making a determination of urgency, is the person using the phone though; not the person writing the online-accounts plug-in or the calendar sync
[17:11] <mpt> E-mail, very urgent for some people, much less so for others (including most in this channel).
[17:11] <mpt> So, you’re saying it should be configurable?
[17:11] <dobey> no
[17:12] <mpt> How else would the user express that determination?
[17:12] <dobey> i'm saying we can't make a determination of urgency on the behalf of all users, for such things
[17:13] <mpt> Designers make determinations of things like that dozens of times a day.
[17:13] <dobey> i'm saying that "missed an appointment" is a separate problem, orthoganal to the auth error problem
[17:14] <mpt> It’s not a separate problem, if one is caused by the other.
[17:15] <dobey> it is still a separate problem, as it can be solved separately
[17:16] <dobey> popping a dialog to request logging in again isn't going to fix missed appointments
[17:16] <mpt> How can it be solved separately?
[17:17] <mpt> It is going to fix missed appointments, if you log in so that you resume getting calendar notifications.
[17:18] <dobey> why did you stop getting calendar notifications in the first place? it wasn't because of a login auth issue; we do sync, and the events from the previous sync should still be in the calendar
[17:19] <dobey> if the events got wiped becuase of the login issue, then that's a bug that needs fixed, because we shouldn't empty the events when an account exists, but there was a login error.
[17:19] <mpt> Because the event was created after the last sync.
[17:19] <mpt> (Or moved up. “Preponed”, as the Indians say.)
[17:20] <dobey> so if your phone had no network, you'd have the same issue
[17:21] <mpt> Yes, but it would usually be more obvious, because it wouldn’t be the only thing that had stopped working.
[17:21] <dobey> should we pop up a full screen window forcing the user to see they must connect to the network, because someone might miss an appointment somewhere, in that case?
[17:23] <mpt> It wouldn’t be full-screen, either in that case, or the re-auth case. But apart from that, maybe: that’s bug 1275761.
[17:24] <dobey> well "maximized" then
[17:24] <mpt> iOS uses a dialog in that case. <https://i.stack.imgur.com/gyPDr.png>
[17:25] <dobey> but the point remains. it seems to me like we're trying to use a canon to kill a mosquito
[17:25] <mpt> (Maybe unless the app specifically says “Don’t worry, I can present offline situations prominently myself”. I haven’t investigated yet.)
[17:25] <dobey> mpt: i expect that dialog is the result of the user expressly attempting to perform some action which requires a data connection?
[17:27] <dobey> as it appears to be on top of an app, and not the home screen
[17:27] <mpt> dobey, “expressly” is fungible. I get that dialog when I switch to Tweetbot and it tries to auto-refresh. I know that it tries to auto-refresh whenever I switch to it; does that mean it was an express instruction? Arguable.
[17:28] <dobey> mpt: you opened an application which requires internet access
[17:28] <dobey> mpt: maybe the message could be better for that app or something, but it is not the result of an actual background process doing things
[17:29] <dobey> ie, you don't get that dialog when calendar sync fails because you're in airplane mode
[17:29] <mpt> That’s true
[17:30] <mpt> But when I’m in airplane mode, it’s much more obvious, because (a) I can’t do anything on the Internet, in any app, and (b) there’s an airplane icon by itself in the corner and I know what it means
[17:31] <dobey> likewise, you probably don't get an auth dialog popping up when on network and calendar sync fails
[17:31] <mpt> I think I have, actually, but it’s only happened once and I didn’t take a screenshot.
[17:31] <dobey> mpt: but in ubuntu, the network indicator is not by itself in the corner, and airplane mode doesn't cover the case where there is no wifi, and you are roaming in an area where you don't have cellular data, but you can still do voice/sms
[17:33] <dobey> then there's the always fun transparent proxy open wifi that screws up dns, and causes all kinds of weird problems
[17:34] <mpt> Captive portals?
[17:34] <dobey> yes
[17:35] <mpt> Yeah, I get the feeling that bug 914507 is “Wishlist” because all the developers understand what’s going on and none of the user testing takes place behind a captive portal :-]
[17:36] <dobey> i don't think that's it at all
[17:36] <mpt> That would be, once implemented, another example of a window appearing out of nowhere, zomg
[17:36] <dobey> well, not really. it would be a window opening as a result of specific user action
[17:37] <mpt> Only in rare cases would it *not* be the result of the specific user action of trying to connect to the wi-fi network
[17:37] <dobey> but the problem is, there's no magic flag for "hey, you're connecting to a captive portal" and they don't all behave the same
[17:38] <mpt> Windows and OS X and even elementary OS manage it.
[17:38] <dobey> windows also uploads all your WPA passwords to a server and shares them with the world, too
[17:38] <mpt> That is entirely unrelated.
[17:39] <dobey> i don't see what any of this has to do with the original discussion :)
[17:39] <mpt> Coincidentally it happens to be another example of a problem that is least-badly solved by a window you weren’t necessarily expecting
[17:40] <dobey> it is a completely different situation and problem
[17:40] <mpt> Yes, it just happens to have the same kind of solution
[17:41] <dobey> not really
[17:41] <dobey> it's just the only "solution" you aren't being sarcastic and dismissive about :)
[17:47] <mpt> dobey, the only specific relevant solution I saw you suggest was a notification, and the reason I gave for a dialog was not sarcastic at all: you can put it behind what you’re doing if you want to, which you can’t with a notification.
[17:48] <mpt> (For example, to Google the problem or to message someone about it.)
[17:49] <dobey> mpt: my solution is to avoid such interruptions unless absolutely necessary. and there are literally no cases where it's absolutely necessary.
[17:52] <dobey> forcing the phone to wake up, and load a web page, is not a good solution; especially for people who may have to pay for data
[17:52] <mpt> dobey, you could say the same (and various extremist designers have said the same) about icons. Or checkboxes. Or menus. Or secondary dialogs. But each of them are the best solution to a class of problem.
[17:54] <dobey> nevermind.
[18:39] <tedg> mpt: network manager can know if it's behind a captive portal, we just don't enable that feature.
[18:44] <tedg> mterry: I pushed a new revision that works for me with launching libertine apps.
[18:45] <mterry> tedg, oh nice -- was it the version thing?
[18:45] <tedg> mterry: Mostly, we also weren't getting the pid list correct, but that was a bug in really old UAL code.
[18:46] <popey> tedg, one day....
[18:46] <tedg> mterry: The version was the weird one, the other was obvious once that was resolved.
[18:46] <popey> we've talked about that at every uds forever, and now we don't even have UDS
[18:46] <tedg> What? Libertine?
[18:46] <tedg> Or PIDs
[18:49] <tedg> Or that we had a summit instead of UDS and then I couldn't go.
[18:51] <popey> tedg, network manager captive portals
[18:51] <popey> one day we'll implement it
[18:52]  * popey notes our downstream elementary os already has
[18:52] <dobey> call your congressman and convince them to ban captive portals
[18:52]  * popey tries that
[18:52] <popey> then realises he lives in the UK so should probably call the queen or something
[18:53] <dobey> popey: call your lord i guess
[18:53] <tedg> The queen is a captive portal. Think about it.
[18:53] <popey> Mind
[18:53] <popey> Blown.
[18:53]  * popey wanders of
[18:53] <popey> *off
[18:53] <tedg> popey: I had it enabled for a while, it was interesting, but needs support at the indicator-network/nm-applet level to make work.
[18:55] <dobey> design is a captive portal.
[20:05] <matv1> I just made a first call after ota 9. after the call ended and i hung up, the call reconnected itself. really weird
[20:05] <matv1> anyone recognize that?
[20:06] <matv1> actually it happened twice.
[20:06] <matv1> only thing i could do was shut off the phone
[20:09] <matv1> would there still be any usefull logs to retrieve now?
[20:25] <matv1> there is nothing in relevant in /var/crash
[20:30] <matv1> but then that wasnt really a crash
[20:31] <matv1> hmm definitely got some funky stuff going on in the dialer app
[20:32] <matv1> I also noticed that the screen blacks out when i start a call
[20:33] <matv1> regardless it being to my face or not. the sensor isnt covered in any way
[23:50] <micael> Hi, i just update my bq to ota9 and i cant access my contacts form the sim card
[23:51] <micael> Is this a bug that a don't know or what?