[07:23] <Saviq> mounin'
[08:00] <tsdgeos> Saviq: landing stuff today?
[08:02] <tsdgeos> Cimi: can you re-review this oneliner? https://code.launchpad.net/~aacid/unity8/dashOverviewFromTempScopePreview/+merge/232378
[08:09] <Saviq> tsdgeos, yeah, later today
[08:09] <tsdgeos> :)
[08:09] <Saviq> I mean landing
[08:09] <Saviq> will review now ;)
[08:49] <tsdgeos> how does one run the shell in desktop nowadays?
[08:50] <tsdgeos> nor run.sh nor the binary start here
[08:52] <tsdgeos> ok -f "worked"
[08:54] <mzanetti> tsdgeos: for me it hangs with a wait_condition on the Audio {} element in the notifications
[08:54] <mzanetti> had to comment that away
[09:02] <Saviq> mzanetti, yeah, it's some Gst fookup
[09:02] <Saviq> seems to be a problem for unity8 desktop session, too
[09:02] <anpok_> https://bugs.launchpad.net/ubuntu/+source/camera-app/+bug/1364258
[09:04] <Cimi> I had same issue in the wizard
[09:04] <Cimi> since uses notifications
[09:04] <anpok_> if either of the two libraries that pull in somehow conflicting gst versions is removed the problem goes away but fails later
[09:12] <tsdgeos> Saviq: you wanted me to kill the dash comunicator
[09:13] <tsdgeos> i actually have to extend it ^_^
[09:13] <tsdgeos> need to know from the shell if the overview is shown or not
[09:13] <Saviq> tsdgeos, nooooooo
[09:14] <Saviq> tsdgeos, why?
[09:14] <tsdgeos> so that i know if the left swipe has to chnge the scope or not
[09:14] <tsdgeos> the code is right now
[09:14] <Saviq> tsdgeos, why does *shell* have to know that
[09:14] <tsdgeos> shell code: if (dashSwipe && ApplicationManager.focusedApplicationId !== "unity8-dash") {
[09:15] <tsdgeos> i can either move the ApplicationManger back to the dash
[09:15] <tsdgeos> or the overview status to the shell
[09:15] <Saviq> tsdgeos, doesn't dash know whether it's focused or not itself?
[09:15] <Saviq> tsdgeos, shell should just say "hey, dash, go to apps scope"
[09:16] <Saviq> tsdgeos, then dash should make the decision what do to (animate if focused, no animate when not)
[09:16] <tsdgeos> Saviq: i don't know, how would it know if it's focused or not?
[09:16] <Saviq> tsdgeos, how does any QQuickView know if it's focused or not?
[09:16] <Saviq> it's a standard window focus thing
[09:17] <Saviq> it should *actually* be visible or not, but later
[09:18] <tsdgeos> didn't get that last sentence
[09:18] <Saviq> tsdgeos, later we'll actually need "am I visible at all" and not "am I focused" (on desktop)
[09:18] <Saviq> tsdgeos, and only animate if visible
[09:19] <Saviq> tsdgeos, but for now, I really don't think shell should ever know state of dash, that's why I wanted to rid us of dash communicator, so that we don't get tempted :P
[09:19] <Saviq> it's just an app, tell it to go to scope foo (ideally via url) and be done with it
[09:20] <Saviq> although we might not be able to go through url atm 'cause url-dispatcher will request focus of dash, not sure how that will play out
[09:20] <Saviq> or at least not through url dispatcher, we could go through the uri handler just fine
[09:20] <tsdgeos> ok...
[09:41] <zbenjamin> Saviq: hey, still working on scope support in QtCreator ;). After the new scope exited i want to automatically show the clickscope, but the url dispatching does not work if a scope is shown that is not in favourites
[09:42] <zbenjamin> Saviq: and the scope i currently develop on is never in the favourites
[09:43] <zbenjamin> Saviq: you can easily try by opeing a scope that is not in your favs and just invoke url-dispatcher scope://clickscope from the shell
[09:49] <Saviq> zbenjamin, sure it works
[09:49] <zbenjamin> Saviq: not for me :(
[09:50] <Saviq> zbenjamin, ah you mean it doesn't *close* the non-favourite scope
[09:50] <zbenjamin> Saviq: yeah the clickscope is not shown
[09:50] <Saviq> zbenjamin, yeah, that'd be a bug in unity8, can you please log one?
[09:50] <zbenjamin> Saviq: lets say i have Amazone scope open and want to go back to clickscope
[09:50] <Saviq> zbenjamin, yeah, I even know what's happening
[09:51] <Saviq> zbenjamin, mention non-favourite scope and overview, as expect that to not work well either...
[09:51]  * Saviq hates how many weird cases we have there :|
[09:52] <zbenjamin> Saviq: yeah in a non fav scope you also cannot pull up the overview
[09:52] <zbenjamin> Saviq: no you can
[09:52] <Saviq> zbenjamin, you can, that's per design
[09:52] <Saviq> zbenjamin, not in a preview though
[09:52] <Saviq> zbenjamin, and then not in a non-fav opened from overview
[09:53]  * tsdgeos is very confused
[09:53] <zbenjamin> Saviq: exactly , non fav openend from overview has no pull up
[09:53] <Saviq> zbenjamin, again, by design
[09:53] <tsdgeos> i've edited /usr/share/unity8/Shell.qml to be syntactically incorrect
[09:54] <tsdgeos> yet the shell still shows up
[09:54] <tsdgeos> why?¿
[09:54] <Saviq> tsdgeos, sure it's not running your local one?
[09:54] <Saviq> zbenjamin, because the overview is already open behind the scope opened
[09:54] <tsdgeos> Saviq: i'm rebooting the phone, how would it be running my local one?
[09:54] <Saviq> tsdgeos, just asking ;)
[09:54] <Saviq> tsdgeos, you didn't mention it's on the phone
[09:55] <Saviq> tsdgeos, so yeah, no idea ;P
[09:58] <tsdgeos> Saviq: so yes
[09:58] <tsdgeos> for some reason beyond my understanding
[09:58] <tsdgeos> it was using file:///home/phablet/lvwph_non_interactive_header_animation/qml/Shell.qml
[09:58] <tsdgeos> now that i've removed the file it doesn't start
[09:58] <Saviq> tsdgeos, whoaps, rebooted?!
[09:58] <tsdgeos> i've rebooted like 10 times
[09:58] <Saviq> wth
[09:58] <tsdgeos> making the syntax worse every time :D
[09:59] <tsdgeos> anyway that's fine, i've been being bad with the phone
[09:59] <zbenjamin> Saviq: https://bugs.launchpad.net/unity8/+bug/1364306
[10:02] <Saviq> zbenjamin, thanks
[10:23] <Saviq> might be indeed
[10:24] <Saviq> I think I was only getting bubbles at event time, not reminder
[10:24] <Saviq> and that was the datetime indicator
[10:24] <Saviq> </typing in wrong channel>
[10:32] <dednick> Saviq: ping
[10:32] <Saviq> dednick, hey
[10:32] <Saviq> dednick, https://code.launchpad.net/~nick-dedekind/unity8/prompts-in-prompts/+merge/232362/comments/568103
[10:33] <dednick> Saviq: howdy. I'm having some trouble with some unicode encoded strings in u8 (time formatter). string is: "%a %d %b\u2003%H:%M"
[10:33] <Saviq> dednick, that in javascript or c++?
[10:33] <dednick> Saviq: c++
[10:34] <dednick> something funky is happening in u8 passing it to g_date_time_format that doesn't happen in indicator-datetime.
[10:35] <Saviq> dednick, I assume that's priv->format ?
[10:36] <dednick> Saviq: yeah, but the problem isn't actually qstring related. just doing: g_date_time_format(g_date_time_new_local(2014, 9, 9, 9, 9, 9), "%a %d %b\u2003%H:%M") in u8 fails, while in indicator-datetime it works
[10:36] <Saviq> dednick, iinteresting, wonder if we need to say somewhere that's utf-8 encoded?
[10:37] <dednick> Saviq: ya. that's what i'm wondering. if we specificlly forbid it or something
[10:37] <Saviq> dednick, that's compiler error maybe?
[10:39] <dednick> Saviq: guessing so
[10:39] <Saviq> dednick, well, if you're not getting a compiler error, there's nothing to guess ;)
[10:39] <dednick> omg. someone is renevating on the roof directly under my feet!
[10:40] <Saviq> ticklish?
[10:40] <greyback> skylight?
[10:40] <Saviq> :D
[10:40] <dednick> yeah. hammering away, and it's causing my NAS to beep complain about the shock!
[10:41] <dednick> Saviq: oh right, i thought you meant compiling problem :)
[10:44] <dednick> Saviq: hm. indicator-datetime is compiling that code with std=c99
[10:45] <dednick> and its a c file
[10:48] <Saviq> dednick, yeah, I'm thinking we need to convert it to pass to g
[10:48] <dednick> Saviq: convert to what?
[10:49] <Saviq> dednick, like *gchar or something? don't we get std::string for literals by default?
[10:49] <dednick> Saviq: hm. maybe.
[10:50] <Saviq> dednick, like https://developer.gnome.org/gtkmm-tutorial/3.9/sec-basics-ustring.html.en maybe
[10:50] <dednick> Saviq: or http://en.cppreference.com/w/cpp/language/string_literal
[10:51] <Saviq> dednick, right, yeah
[10:51] <Saviq> so u8""
[10:51] <Saviq> dednick, and then you can use the actual character, no need for escaping
[10:52] <dednick> Saviq: this is to do with that relative time formatter. not sure if it would work in indicator-datetime domain if not the same
[10:52] <Saviq> dednick, mhm
[10:52] <Saviq> dednick, so, any idea about prompts-in-prompts and hw keyboard input?
[10:53] <dednick> Saviq: ah. haven't looked yet
[10:53] <Saviq> dednick, that's blocking prompts from landing, fwiw, so whenever you can
[10:53] <dednick> Saviq: yeah, sorry. will do now
[10:53] <Saviq> dednick, no need to be sorry :)
[10:53] <Saviq> dednick, just letting you know, I know you wanted to get it done, and that people are waiting
[11:04] <facundobatista> Holas
[11:18] <Cimi> dednick, is now possible to know when a connection is established?
[11:18] <Cimi> Wellark, ^
[11:18] <dednick> Cimi: for unitymenumodel?
[11:18] <Cimi> yes
[11:18] <dednick> Cimi: nameowner is valid string i think
[11:19] <Cimi> dednick, camelCase?
[11:19] <dednick> Cimi: most likely. i cant remember
[11:20] <dednick> Cimi: yes
[11:30] <Cimi> dednick, nah
[11:33] <Cimi> dednick, nameOwner is 1.5 here
[11:33] <Cimi> and doesn't change
[11:34] <dednick> Cimi: ? that means it's connected i think
[11:34] <dednick> Cimi: if it isn't empty, it's connected to the bus
[11:34] <Cimi> dednick, but if i connect to a broken ap
[11:34] <Cimi> dednick, is still says 1.5
[11:35] <dednick> Cimi: ... i asked "for unitymenumodel?", not "for access point?"
[11:35] <Cimi> dednick, ah crap :D
[11:35] <Cimi> dednick, I use unitymenumodel for the access point
[11:35] <dednick> Cimi: you can't. it's just active/inactive.
[11:36] <Cimi> dednick, there is active property then?
[11:36] <dednick> Cimi: https://bugs.launchpad.net/indicator-network/+bug/1349371
[11:40] <dednick> Saviq: the datetime is the LC_ALL=C flag btw. not sure why though
[11:41] <Saviq> dednick, aah well, maybe because it doesn't know it's UTF8 then?
[11:43] <dednick> Saviq: C.UTF-8 ?
[11:44] <Saviq> dednick, don't think such a thing exists...
[11:44] <Saviq> dednick, but if it does, sure :)
[11:51] <Cimi> do you guys have wifi strength icons in system settings
[11:51] <Cimi> ?
[11:51] <dednick> i need to find another place to work... my head is exploding from hammering!
[11:51] <Cimi> dednick, wanna come here?
[11:52] <dednick> Cimi: heh. can't need to tidy up today for parent's visit tomorrow :)
[11:53] <Cimi> seb128, hi! do you have wifi icons in system settings?
[11:53] <seb128> Cimi, no, icons are in the theme
[11:53] <seb128> Cimi, oh, and hey ;-)
[11:54] <dednick> Saviq: which tests failed without LC_ALL=C ?
[11:55] <Cimi> seb128, inside the wifi page, all access points are without signal strength to me
[11:57] <seb128> Cimi, check with Wellark or Satoris
[11:58] <Cimi> seb128, yours!
[11:58] <seb128> Cimi, ?
[11:58] <Cimi> seb128, can you run system settings and see if you have the icons?
[11:58] <seb128> I don't
[11:58] <seb128> I only have names and checkboxes
[11:58] <Cimi> mpt, ping
[11:59] <seb128> Cimi, they should have icons, see https://wiki.ubuntu.com/Networking
[11:59] <seb128> Wellark should have a look I guess
[12:51] <mpt> Cimi, did you report a bug about it?
[12:51] <Cimi> mpt, nope, is a bug though
[12:51] <Cimi> mpt, I was wondering if was a design change
[12:51] <Cimi> quite a crazy one :D
[12:52] <mpt> A network should be presented the same way regardless of whether it’s in the indicator, in System Settings, or in the first-run setup <https://wiki.ubuntu.com/Networking#wi-fi-network>
[12:54] <mpt> using the same code, ideally
[12:59] <mpt> Cimi, reported bug 1364366
[13:19] <Saviq> dednick, testPreviewPayments
[13:21] <dednick> Saviq: hm. that fails even with LC_ALL=C
[13:22] <Saviq> dednick, note that the env in QmlTests.cmake is broken
[13:22] <Saviq> dednick, export as you call the test
[13:22] <dednick> Saviq: ah
[13:23] <Saviq> dednick, I never got around to redoing the test macros to gain some sanity in them
[13:23] <dednick> Saviq: mk
[13:23] <dednick> Saviq: C.UTF-8 don't work anyhow
[13:23] <Saviq> dednick, yeah, we never understood what the � there is ;)
[13:31] <dednick> why can't we dismiss notifications?!
[13:31] <dednick> or can we and i'm stupid?
[13:33] <Saviq> dednick, MacSlow is working on that
[13:35] <dednick> Saviq: where can i find the AP tests for silo qa?
[13:38] <dednick> my keyboard is working fine on phone, but under AP the test fails. but i think my failure is due to the "Failed to add storage device" notification being constantly open
[13:38] <Saviq> dednick, right, reboot your phone
[13:38] <dednick> since i think it's trying to do a search, which button is covered by the notification
[13:38] <Saviq> dednick, yeah, should go away after reboot
[13:39] <Saviq> adb shell gdbus introspect --system -p -d org.freedesktop.UDisks2 -o /org/freeedesktop/UDisks2/block_devices/mmcblk0 /org/freedesktop/UDisks2/block_devices/mmcblk0p2 | grep System
[13:39] <Saviq> dednick, ↑ to verify your phone's in a good state, that should say true
[13:39] <dednick> Saviq: doesn't say anything
[13:39] <dednick> after i rebooted
[13:41] <Saviq> dednick, krillin is it?
[13:41] <dednick> Saviq: no. but test was fine (failed due to lack of input)
[13:42] <dednick> it stopped doing the notification
[13:42] <Saviq> dednick, k, so yeah, it must be we don't have focus on the dash any more or some such
[13:42] <Saviq> dednick, and vkb works because it works over dbus
[13:42] <Saviq> or well, maybe we have focus but the keyboard filter doesn't work properly
[13:43] <Saviq> dednick, WindowKeyFilter thing in Shell.qml
[13:43] <dednick> Saviq: is it keyboard input rather than touch input?
[13:43] <Saviq> dednick, yes, ap goes through keyboard input directly
[13:43] <dednick> ah
[13:43] <dednick> would explain why the touch is working then
[13:43] <Saviq> yes
[13:43] <Saviq> which, btw, is a good indication we'd break desktop session again...
[13:44] <dednick> meh.
[13:44] <dednick> :)
[13:44] <jgdx> Saviq, do you know if there's a dbus call uss can make to reboot the phone?
[13:44] <dednick> i'll take a look
[13:44] <Saviq> jgdx, yeah
[13:45] <jgdx> Saviq, great. Care to share? :)
[13:45] <Saviq> jgdx, trying to get you a link ;)
[13:45] <jgdx> Saviq, woo!
[13:45] <Saviq> jgdx, http://bazaar.launchpad.net/~unity-team/unity8/trunk/view/head:/plugins/Unity/Session/dbusunitysessionservice.cpp
[13:46] <jgdx> Saviq, golden. Thank you
[13:46] <Saviq> jgdx, the login1 calls are probably the interesting ones to you
[13:46] <jgdx> Saviq, seems so:)
[13:55] <Saviq> mterry, hey, you know your rtm silo got NACK from QA on the two settings app bugs mentioned?
[13:55] <mterry> Saviq, yeah...  I'm trying to test / see what's up
[13:57] <Saviq> mterry, here's the report from ToyKeeper https://trello.com/c/9B6QLvDd/37-rtm-landing-003-mterry-unity8-ubuntu-system-settings-gsettings-ubuntu-touch-schemas
[13:57] <Saviq> mterry, not sure what to answer on the "makes it possible to launch apps" front
[13:58] <mterry> Saviq, that's a whole separate thing
[13:58] <mterry> Saviq, that at one point was how design wanted it, so the app was ready to go, but I can certainly imagine it being a problem
[13:58] <MacSlow> dednick, snap-decisions are currently getting the "swipe-to-dismiss" feature
[13:59] <dednick> MacSlow: cool.
[13:59] <Saviq> mterry, right, we should maybe start it and kill it if unlock failed / cancelled...
[13:59] <Saviq> mterry, but yeah, separate issue
[13:59] <mterry> Saviq, I don't think that helps -- I figured the problem was that an app might do something on startup?
[14:00] <Saviq> mterry, ok, let's have a chat about this at some point with security folk
[14:00] <Saviq> we should file a bug
[14:21] <mterry> Saviq, I'm trying to schedule a meeting between design and security for an unrelated thing, I can ask about the launched-apps issue then too
[14:21] <Saviq> mterry, sure, please do
[14:39] <Cimi> Saviq, better Qt.rgba(0, 0, 0, 0.5) or color: "black"; opacity: 0.5 ?
[14:39] <Saviq> Cimi, that's different behavior
[14:42] <Saviq> Cimi, #7F should be 0.5
[14:44] <Cimi> Saviq, yup, using the calculator :)
[14:45] <Saviq> Cimi, yeah, I was just unsure between 7F and 80 ;)
[14:46] <Cimi> Saviq, is actually 80
[14:46] <Cimi> Saviq, I just double checked
[14:47] <Saviq> Cimi, well, that's 128, more than 0.5 * 255 ;)
[14:47] <Saviq> Cimi, but yeah, it's .5 either way
[14:48] <Cimi> Saviq, it's 7F.8
[14:48] <Cimi> xD
[14:52] <Saviq> Cimi, ;)
[14:54] <Cimi> mterry, https://code.launchpad.net/~cimi/ubuntu-system-settings/wizard.wifi-fixes/+merge/233071
[14:55] <mterry> Cimi, can you link to a bug and/or expand the description?
[14:58] <Cimi> mterry, done
[14:58] <Cimi> mterry, the changes in Page.qml are cosmetics
[14:59] <Cimi> mterry, the margin changed to 3 times when we added the pagination, but that was removed and the padding not restored
[14:59] <Cimi> mterry, wordwrap was added to avoid wrapping Wi-
[14:59] <Cimi> fi
[14:59] <mterry> Cimi, why switch to WordWrap?
[14:59] <Cimi> in italian
[14:59] <mterry> Cimi, ah... hmm.  But then doesn't Wi-Fi overlap the screen?
[15:00] <Cimi> mterry, it shouldn't...
[15:00] <Cimi> is still there though
[15:00] <Cimi> weird
[15:00] <Cimi> mterry, wi-fi is split here
[15:01] <Cimi> mterry, you can try italian
[15:01] <mterry> Cimi, we should just switch to a unicode non-breaking hyphen
[15:01] <mterry> Cimi, "‑"
[15:03] <Saviq> —
[15:06] <Cimi> mterry, how?
[15:06] <Cimi> Saviq,
[15:06] <mterry> Saviq, that's for parentheticals, right?
[15:06] <Cimi> can I paste that?
[15:06] <Saviq> mterry, no idea, that's just compose + ---
[15:06] <Cimi> Saviq, basically I have
[15:07] <Cimi> "Connetti alla rete Wi-
[15:07] <Cimi> fi"
[15:07] <Cimi> translations should be updated too
[15:07] <mterry> Saviq, yeah, that's an em dash, which is for parentheticals.  The en dash is for ranges and such.  The hyphen is for word joins
[15:08] <mterry> Cimi, just copy paste the hyphen I gave you above
[15:08] <mterry> http://english.stackexchange.com/questions/2116/when-should-i-use-an-em-dash-an-en-dash-and-a-hyphen
[15:08] <mterry> Cimi, I *assume* that Qt's word breaking logic takes the unicode semantics into account, but worth a test
[15:12] <Cimi> mterry, works
[15:12] <Cimi> mterry, but all potfiles will require update
[15:13] <mterry> Cimi, what else is new, we change those strings all the time.  And design isn't done with the wizard
[15:14] <Cimi> mterry, pushed
[15:14] <Cimi> mterry, ok we need to fix button visuals now
[15:14] <Cimi> mterry, they look shite
[15:14] <Cimi> mterry, either in sdk or locally
[15:14] <Cimi> I'd go locally
[15:14] <Cimi> then also push in sdk
[15:15] <Cimi> wizard is a custom themed app, I don't mind having a static theme for a button
[15:17] <mterry> Cimi, I'd really rather just fix it in one place
[15:18] <Cimi> mterry, so when they change the visuals there we will have a broken wizard?
[15:18] <mterry> Cimi, ?
[15:19] <Cimi> mterry, we want monochromatic theme here
[15:19] <Cimi> mterry, the sdk is not now
[15:19] <Cimi> and I don't think we can go and say "let's change all buttons to white"
[15:20] <Cimi> because it will affects all apps using surugradient
[15:21] <mterry> Cimi, well that's my question -- I was under the impression that the visuals for the wizard were made using an updated version of suru.  If design intended those buttons to be wholly-custom to the wizard, that should be clarified -- I've got a meeting with them tomorrow to talk about a security thing too, I can ask then
[15:21] <Cimi> mterry, ok
[15:54] <Cimi> Saviq, infographics, worth me looking or not?
[16:06] <Saviq> dandrader|afk, could you have a look at the changes to dednick's unity8 prompts branch, there was a focus issue
[16:06] <Saviq> Cimi, looking at what? merging? sure, I hope to land this week
[16:14] <dandrader> Saviq, ok
[16:16] <dandrader> dednick_, could we have a qmltest covering that "issue with session focus"?
[16:22] <greyback> Saviq: fyi we've a few other qtmir we'd like to land after prompts-in-prompts does, can you let me know when it lands so I can get a silo for it
[16:23] <Saviq> greyback, sure
[16:24] <Saviq> greyback, unless it gets ACKed overnight and kgunn lands it, won't be earlier than tomorrow noon-ish
[16:25] <greyback> Saviq: that's ok, I just want to avoid parallel silos of qtmir, as it just causes pain
[17:17] <dednick_> dandrader: i can add a test, but i'm having some problems reproducing in qmltests
[17:17] <dednick_> but off for the day now
[17:17] <dednick_> well, at least until later
[17:26] <Cimi> mterry, commented inline
[17:26] <Cimi> mterry, replaced ListView with Repeater + column + flickable
[17:26] <kgunn> greyback: so, dednick's prompt in prompt doesn't look like it'll land until tomorrow earliest (still got some needs fixin' going on)
[17:26] <kgunn> is it ok to hold you off
[17:27] <kgunn> for maybe a 2 days ?
[17:27] <kgunn> or that too much ?
[17:28] <Cimi> Saviq, in regards of this review https://code.launchpad.net/~cimi/ubuntu-system-settings/wizard.wifi-fixes/+merge/233071 better to have hiddenComponent or setting sourceComponent to null?
[17:42] <Saviq> Cimi, of course better to null, why would you have a blank component?
[17:42] <Cimi> Saviq, residue of the past
[17:42] <Cimi> Saviq, C
[17:42] <Cimi> Saviq, avoiding null pointers as much as possible :)
[17:43] <Cimi> mterry, pushed with null component
[17:43] <Cimi> should be better now, works fine
[17:47] <Cimi> the default wallpaper on the phone has 1px bright line on right edge of the screen
[17:47] <Cimi> can you guys confirm that?
[17:54] <Cimi> https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1364537
[17:54] <Cimi> (it was annoying me)
[18:06] <Cimi> mterry, those fixes are fine for now, but I believe we'll be asked to add hidden network too...
[19:31] <Cimi> Saviq, mzanetti do we have a bugreport for this gst audio issue stopping apps to work on the desktop?
[19:32] <Saviq> Cimi, yeah there is a slew of 'em actually
[19:32] <Saviq> Cimi, bug #1363314
[19:33] <Cimi> Saviq, thx
[21:01] <bregma> r7b...
[21:02] <bregma> brzbup
[21:02]  * bregma is having keyboard trouble
[21:18] <dednick> Saviq: you still about?
[21:18] <Saviq> dednick, kinda
[21:18] <dednick> Saviq: :) having some really straing behaviour in a test case.
[21:18] <Saviq> dednick, try me
[21:20] <dednick> Saviq: http://paste.ubuntu.com/8218316/
[21:21] <dednick> the test down below SessionManager.createSession returns a qobject pointer with c++ ownership.
[21:21] <dednick> in testXXX the test function creates 2 copies of Session, in tryXXX it only creates ones like it should
[21:21] <dednick> Saviq: when assigning "sessionContainerLoader.item.session = fakeSession;" it creates another "fakeSession" object
[21:22] <dednick> Saviq: if i don't use a loader, it seems to work ok...
[21:23] <Saviq> dednick, what type is SessionContainer.session?
[21:23] <dednick> Saviq: QtObject
[21:23] <Saviq> ok so nothing special
[21:24] <dednick> Saviq: yeah, same happens if it's a var.
[21:24] <Saviq> dednick, just FYI, if a prop is of a complex type, even if you go : null, it will create the object on creation and replace it with null just after :|
[21:25] <Saviq> but not the issue here of course
[21:25] <dednick> Saviq: ya
[21:25] <Saviq> dednick, so is it calling .createSession twice?
[21:26] <Saviq> or does it copy fakeSession or something?
[21:27] <Saviq> dednick, so the symptom is that your .createSurface is called on a different object than the fakeSession you just got from .createSession()?
[21:27] <dednick> Saviq: it's calling createSession twice.
[21:27] <Saviq> dednick, did you try breaking in createSession() to see the trace?
[21:27] <dednick> Saviq: hm. that's odd
[21:28] <dednick> it's like it's creating a function pointer from createSession
[21:29] <dednick> Saviq: no, not yet
[21:32] <Saviq> dednick, that'd be really odd, check the type of fakeSession?
[21:33] <Saviq> dednick, but even if that was the case, then the object would be of wrong type in item.session (not even QtObject IIRC)
[21:33] <Saviq> dednick, and there's nothing that would actually call it
[21:34] <dednick> QDEBUG : qmltestrunner::SessionContainer::test_interative() qml: Session(0x2478800)
[21:34] <dednick> trying to debug now
[21:41] <dednick> dangit. how the hell do you add a custom run to qtcreator?!
[21:50] <dandrader> dednick, are you by any chance creating a session on the test init/cleanup funtions?
[21:51] <dednick> dandrader: nope
[21:56] <dednick> dandrader, Saviq: looks like something funky is going on though... i removed all the other code from the test and it stopped doing it :/
[21:57] <Saviq> dednick, stupid question: is the test function only called once?
[21:57] <dandrader> when I'm haunted by mysterious issues like that, usually the best thing is approaching it again later, with a fresh mind, next morning.
[21:58] <dednick> Saviq, dandrader: sorry, it looks like i'm a dumbass. the createSurface function is being called by something else
[21:59] <dandrader> dednick, np, glad you found it :)
[21:59] <Saviq> having a sounding board is often helpful :)
[23:26] <dednick> kgunn: i've added the test for the problem we were having now
[23:26] <dednick> zzzzzz