[06:15] Scopes question: is it possible to have a Scope add something to the clipboard on Ubuntu Phone? [06:15] I don't see any docs talking about that [06:17] https://developer.ubuntu.com/en/api/apps/qml/sdk-15.04/Ubuntu.Components.Clipboard/ (404) [06:18] Ah, https://developer.ubuntu.com/api/apps/qml/sdk-15.04.1/Ubuntu.Components.Clipboard/ [06:19] Can I call that from a Scope? === chihchun_afk is now known as chihchun === lan3y is now known as Laney === Laney is now known as Guest80303 === Guest80303 is now known as Laney [09:33] Good morning all; happy Thursday, and happy Hugging Day! 😃 === lool- is now known as lool === _salem is now known as salem_ [12:14] mcphail: so I did some more testing with ptrace and the like, I found a way to do more efficient reads using ptrace, by having the tracer and tracee share virtual memory, but speed-wise it didn't benefit much [12:14] but I just remembered seccomp, and that increases speed drastically, as it only pauses the tracee when calling open [12:15] though it seems seccomp filter may not be enabled in the current kernel [12:51] bartbes: Gosh - that is interesting. I had never heard of that before, but a bit of Googling took me to https://pdos.csail.mit.edu/archive/mbox/ which seems to do the same thing. I wonder if we could subvert this mbox routine for our own ends? [12:53] it looks like it does that already, though that still requires the kernel support if it wants to use seccomp filtering, of course [12:53] though it looks like it is largely equivalent to using something like overlayfs, which may have the performance advantage === jdstrand_ is now known as jdstrand [15:38] bartbes: yes, I've asked for overlayfs before, but I don't think it is available for the android 4.whatever kernels [15:38] right, overlayfs may not be, but unionfs, aufs or something of that nature might be [15:42] I'd imagine that wouldplay havoc with the app confinement model [15:43] Although it might be a better model over all [15:47] ahayzen: yo, any udpate? [15:47] faenil, for which? the scrollview stuff ? [15:47] ahayzen: yup [15:47] usertesting is approaching :) [15:47] faenil, not yet, i maybe able to look in a bit/over the weekend [15:47] faenil, how soon ? ;-) [15:48] ahayzen: I think my email said today :D [15:48] ah [15:48] and I asked for an ETA but nobody replied :( [15:49] ok i'll have a go in a minute, faenil is all the code in rc-proposed ? [15:49] i replied ;-) [15:49] DanChapman: you had it done already, no ETA needed :D [15:50] ahayzen: yep...thanks! [15:50] ahayzen: if you really can't do it don't worry, just let me know and I'll relay the info [15:50] ok, i'm on it now [15:52] ahayzen: awesome! thanks a lot! [15:52] ahayzen: if you need help just shout, it should be really quick [15:52] okies [15:52] just wrap the views in ScrollView { } [15:52] and that's it, basically [15:52] hopefully [15:52] heheh === chihchun is now known as chihchun_afk [16:02] faenil, hmmm, so i put a ScrollView { anchors { fill: parent } ... MyListViewThing {} ... } then it mucks up the header [16:03] i shouldn't need to set the flickable of the page to the ScrollView right? [16:03] ahayzen: what do you mean? [16:03] faenil, the content goes over the header and the header doesn't slide in.out [16:03] in/out [16:03] over? :| [16:04] yup [16:04] ahayzen: can you link me to the page you're editing? [16:05] faenil, http://bazaar.launchpad.net/~music-app-dev/music-app/trunk/view/head:/app/ui/Songs.qml [16:05] faenil, i put a ScrollView {} around the MultiSelectListView {} [16:06] ahayzen: try setting the flickable of the page [16:06] but that should be automatic, afaik... [16:06] was just trying that [16:06] faenil, file:///opt/click.ubuntu.com/com.ubuntu.music/2.3.latest/app/ui/Songs.qml:36:16: Unable to assign ScrollView_QMLTYPE_118 to QQuickFlickable [16:07] not the scrollview, the flickable [16:07] scrollview is just a wrapper [16:07] oh [16:07] :D [16:07] t1mp: how many levels does the header go down when looking for a flickable? [16:08] faenil: 1. It must be directly in the Page. [16:08] faenil, yeah that seems to work [16:08] t1mp: ok [16:08] faenil: you can set Page.flickable to override the searching [16:08] faenil, oh except that the scrollbar goes behind the header sometimes :-) [16:08] faenil: or with the new PageHeader, set PageHeader.flickable. PageHeader doesn't do any looking so you have to set it yourself. [16:08] yes, I had the flickable set in my examples [16:08] ahayzen: yes that's a known issue, don't worry [16:08] and yeah can hardly see it as the colour is similar to the background [16:08] no design solution yet [16:09] and about the colour, visual design tested that already afaik [16:09] you can't have a colour that works amazingly on all surfaces unfortunately [16:09] as in on mako, you can hardly see it, it must be like Qt.lighter(background, 1.1) or something [16:10] that means it changes colour based on the background [16:10] which would be best :-) [16:10] and I think we don't want a scrollbar that can get blue or green :P [16:10] well a dark colour for light backgrounds and a light colour for dark backgrounds [16:11] anyway, faenil shall i do that flickable 'hack' for now, so you can at least test ? [16:11] ahayzen: sure, I mean, that's probably what everyone will have to do when moving to scrollview [16:11] it's not a hack [16:11] ugh [16:11] it's just that the header assumes that flickable is directly inside Page [16:11] and that's not the case if you have a scrollview [16:11] why can't the Page thing go, if the child is a ScrollView then look at the child of that? [16:12] ahayzen: sure, of course [16:12] ahayzen: but that could create problems in other usecases [16:13] ahayzen: I just didn't want to lean forward and give the trivial solution [16:13] t1mp: can you have a closer look at that and change it if it doesn't screw anything else? [16:14] ahayzen: checking qml types is not exactly cheap, so you usually try to avoid that [16:14] faenil: are you talking about going down more levels to find a flickable? [16:14] faenil, ahayzen: with the new PageHeader, basically you have to set the flickable that controls the header yourself. [16:14] t1mp: just special handling when the child is a ScrollView [16:15] faenil, ahayzen: auto-detection can be bad for performance. And often it doesn't work the way you want. [16:16] besides checking the flickable, it also needs to check the anchors of the flickable (it only makes sense if the flickable anchors to the top of the page), and whether the flickable can actually flick (which can depend on the height and contentHeight of the flickable) [16:16] ^all this together can some times cause binding loops [16:16] so making the app developer, who knows what (s)he's doing with the flickable, set PageHeader.flickable, avoids those issues [16:17] ahayzen: also, what is "background"? you need to sample whatever is behind the scrollbar, which is not always a "Page", right? [16:18] and it could have different colours along the height/width of the scrollbar [16:18] faenil, yeah, but would we be able to set custom styles for the scrollbar? like with QtQuick.Controls ? [16:19] faenil, picking the MainView.backgroundColor or from the MainView.theme.name would be pretty close though [16:19] ahayzen: not really, you don't know what's behind the scrollbar [16:19] yeah you don't, but that'd work for most cases [16:19] (unless you get the framebuffer before painting the scrollbar and do analysis of that) [16:20] (and even then, it's not trivial at all) [16:23] ahayzen: well, if you want to go that way then you can just sample the current palette [16:23] mm no, that wouldn't work [16:23] the appdev being allowed to set what they won't as an override would probably work :-) [16:24] ahayzen: StyleHint is your friend, I don't see why not :) [16:24] :-) [16:24] right just need to sort out the GridView's having an incorrect topMargin [16:27] ahayzen: the problem is we don't currently promise API stability of Styles (afaik) [16:27] so, using stylehint could break in the future if the style implementation changes [16:27] heh, hence why i haven't transitioned music to proper styling yet :') [16:27] :) [16:27] ahayzen: so, for the moment, that's an issue that visual design has to deal with, not ours :) [16:28] and the one with the header is for UX designers :) [16:28] I already hit my head against the wall enough to find a decent solution :) [16:28] (and if you have a good solution, propose it and I will implement it) [16:32] faenil, lp:~ahayzen/music-app/use-scrollview :-) [16:33] ahayzen: awesome, thanks! [16:34] faenil, let me know if anything is broken, or you have trouble running etc :-) [16:34] ahayzen: will do! waiting for a unity8 fix as well [16:34] no apps run at all on this laptop atm [16:34] that's being looked into by mzanetti :) [16:36] ahayzen: and if you happen to find a solution on how to make scrollbar behave with a sliding header, you know where to find me ;) [16:36] taking into account that the scrollbar is draggable on touch when there are >10 pages to scroll [16:37] (that's actually the tricky bit, otherwise there would be no issue) [16:37] so, basically think what happens when you scroll the flickable by dragging the scrollbar on touch, with a sliding header [16:37] faenil, ensure the top of the scrollbar starts below the bottom of the header [16:37] the key point being we can't afford resizing and repositiong the scrollbar while it's being dragged of course [16:37] ahayzen: see ^ [16:38] if the header didn't move, the fix would be trivial [16:38] but i mean, when the header disappears...don't move the top back up [16:38] just keep it at 6.125GU from the top [16:38] and in fact it works perfectly when in dual panel mode [16:38] ahayzen: i.e. *always* leaving a space at the top of the page? [16:39] even when the header is hidden? [16:39] yeah [16:39] I proposed that [16:39] it wasn't accepted [16:39] heh [16:39] (it's what Whatsapp does on Android, fwiw) [16:39] they have the same issue [16:39] and they just leave a space at the top even when the header is smaller [16:39] but I agree that it looks weird [16:40] ahayzen: let me know if you come up with any other idea :) [16:40] will do :-) [16:40] thanks :) [16:40] ahayzen: btw, you probably know this already, this is not a problem if you use the convergent components and you're in multipanel mode [16:41] because the header should be locked in that situation [16:41] header(s) [16:41] and scrollbar aligns properly in that case [16:41] yeah, we need to get back onto convergent stuff for music :-/ [16:41] :) === BOHverkill_ is now known as BOHverkill [16:50] ahayzen: would you mind reporting a bug about the scrollbar colour so we can track it and not forget it? [16:51] faenil, yup [16:51] ahayzen: cheers [16:52] ahayzen: I didn't even know about the MainView.backgroundColour property [16:52] but that could indeed be a slightly more suitable heuristic [16:52] even though it would still break in an infinite number of usecases [16:53] yup tricky :-/ [16:55] faenil, bug 1536722 [16:55] bug 1536722 in ubuntu-ui-toolkit (Ubuntu) "ScrollView scrollbar on a dark theme can be difficult to see" [Undecided,New] https://launchpad.net/bugs/1536722 [16:55] ahayzen: awesome, cheers === chihchun_afk is now known as chihchun [17:16] DanChapman, hiya, when will you release a 0.6.x version of dekko? [17:19] pmcgowan, hey! as soon as the next OTA is released. Still got a few things to fix and also got to wait on ScrollView component to land in stable [17:19] But it's mostly ready now [17:20] DanChapman, awesome [17:23] DanChapman, have you a click package for testing? [17:35] pmcgowan, you can grab clicks from here http://dekko-ci.cloudapp.net:8080/job/dev-trunk-build/. There's no translations included with those === chihchun is now known as chihchun_afk [17:37] DanChapman, thanks got it [18:26] ahayzen, sorry I didn't get to finishing the experiment on jenkins yesterday [18:26] balloons, no worries :-) [18:32] ahayzen, fiddling with it again, I'll ping when there's something to see [18:46] mcphail, Guess what...? I almost have it working! [19:53] hey all. I'd like to create a scope with JS, but I think this tutorial is really outdated: https://developer.ubuntu.com/en/scopes/tutorials/developing-scopes-javascript/ [19:53] I'm using Ubuntu 16.04 (dev branch), but I only see C++ scope templates [19:54] in the SDK [19:54] And I can't install the mentioned unity-js-scopes-dev package as well [21:17] davidcalle, ^^ === salem_ is now known as _salem [22:42] DS-McGuire: almost? [22:43] mcphail, Yep! Now I just have: [22:43] CMakeLists.txt:1: error: The CMAKE_C_COMPILER: /usr/bin/arm-linux-gnueabihf-gcc is not a full path to an existing compiler tool. Tell CMake where to find the compiler by setting either the environment variable "CC" or the CMake cache entry CMAKE_C_COMPILER to the full path to the compiler, or to the compiler name if it is in the PATH. [22:44] I spent about 12 hours sorting out all the problems on my system and I finally downloaded the updates I needed. [22:44] don't know what is going wrong there. The SDK should set up cmake for cross-compiling automatically [22:44] I think Ubuntu doesn't want me to make a JS scope haha! [22:45] DS-McGuire: someone is sabotaging your entry to the competition ;) [22:45] * DS-McGuire eyes everyone in the IRC [22:45] mcphail, I've almost given up haha [22:46] DS-McGuire, js scopes? Bug davidcalle! [22:47] davidcalle, poke! Stuff doesn't work. === _salem is now known as salem_ [23:20] DS-McGuire: have you ran the unity-js-scopes-tool setup command? [23:21] DS-McGuire: do other app templates work fine? [23:22] bzoltan_: any idea about "CMakeLists.txt:1: error: The CMAKE_C_COMPILER: /usr/bin/arm-linux-gnueabihf-gcc is not a full path to an existing compiler tool. Tell CMake where to find the compiler by setting either the environment variable "CC" or the CMake cache entry CMAKE_C_COMPILER to the full path to the compiler, or to the compiler name if it is in the [23:22] PATH." ? When trying to run/setup a js scope? [23:43] davidcalle, Sorry I was playing CS:GO. I'll run it again. [23:46] davidcalle, I ran it. Same problem. Output of the scopetool command gave: Remounting chroots ...sh: 1: click-chroot-agent: not found [23:46] If that helps at all. [23:49] davidcalle: whilst I can build JS scopes fine on my machine, I also get an error when running "unity-js-scopes-tool setup" as per http://paste.ubuntu.com/14583740/ . Doesn't seem to cause any problems, though