=== chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk [07:19] tsdgeos: I need your advice. A QML scrolling bug that still exists under Unity7 is which component (not QtMir I assume)? [07:20] duflu: it depends i guess, what kind of bug? === chihchun_afk is now known as chihchun [07:20] tsdgeos: Speed/accuracy. If I'm running Unity7 then it's not QtMir, surely. What component is it? [07:22] duflu: i guess you could start with qtdeclarative-opensource-src [07:22] tsdgeos: That's it, thanks [07:23] i don't understand exactly what you mean with accuracy [07:23] tsdgeos: I mean webbrowser app scrolls well but dash and systems settings do not [07:23] Even under unity7 [07:41] sup all [07:41] the scroll is too fast now with Xmir [07:42] like when you scroll down on g+ it scrolls like 3 screens or such [07:42] worked fine before [07:43] xmir/yakkety-proposed,now 2:1.18.4-1ubuntu4 amd64 [installed] [07:49] * duflu looks [07:49] ohNoes_: Yeah are you using a laptop touchpad? [07:52] ohNoes_: Please let me know what device you're using for scrolling. I'm working on triaging those issues right now [07:52] Hmm is G+ a native or web app? [07:52] I've never looked [07:53] duflu: mouse wheel [07:53] ohNoes_: That's weird. Mouse wheels should not have changed. Any other basic apps affected? [07:53] duflu: run firefox on xmir, open g+ or any other longish page [07:53] OK, I'll do that now. [07:53] scroll up and down with mouse wheel [07:54] tried with geany, firefox [07:54] I'm presently doing an Xmir and scrolling bug overhaul [07:55] nice :D [07:55] would be awesome.. like totally awesome to have a way to configure the scroll rate [07:56] maybe in system settings or something [07:59] ohNoes_: System Settings suffers from bug 1605513 which is different. Thanks for pointing out the regression with Firefox! [07:59] bug 1605513 in qtdeclarative-opensource-src (Ubuntu) "Touchpad scrolling in dash and system settings is disproportionately faster than mouse wheel scrolling" [Undecided,New] https://launchpad.net/bugs/1605513 [07:59] np :D === chihchun is now known as chihchun_afk [08:03] ohNoes_: I can't reproduce any regression. Can you suggest another test case? [08:36] duflu: run firefox on Xmir, go to reddit.com, scroll down once (with the scroll wheel) and then count how many lines down. tap the arrow key UP until you are at the top of the page [08:36] ohNoes_: Yep I figured it out thanks... https://bugs.launchpad.net/ubuntu/+source/xorg-server/+bug/1607240 [08:36] Launchpad bug 1607240 in unity8 (Ubuntu) "Xmir Firefox scrolling is too fast (mouse wheel) when run under Unity8 (since Xmir 2:1.18.4-1ubuntu4)" [Undecided,Confirmed] [08:36] duflu: if i scroll down once it goes 8 lines down [08:36] Yeah [08:36] yay :d === chihchun_afk is now known as chihchun [08:45] Saviq: Can you tell me what code in Unity8 synthesises the vscroll events from Qt wheel angles? [08:45] What source package? [08:47] duflu: i guess you need: http://bazaar.launchpad.net/~mir-team/qtmir/trunk/view/head:/src/modules/Unity/Application/mirsurface.cpp#L106 [08:47] and http://bazaar.launchpad.net/~mir-team/qtmir/trunk/view/head:/src/platforms/mirserver/qteventfeeder.cpp [08:48] anpok: Thanks. But I found the answer (to multiple questions) [08:48] https://bugs.launchpad.net/ubuntu/+source/xorg-server/+bug/1607240/comments/6 [08:48] Launchpad bug 1607240 in unity8 (Ubuntu) "Xmir Firefox scrolling is too fast (mouse wheel) when run under Unity8 (since Xmir 2:1.18.4-1ubuntu4)" [Undecided,Confirmed] [08:51] ohNoes_: https://bugs.launchpad.net/ubuntu/+source/ubuntu-system-settings/+bug/1607240/comments/6 [08:51] Launchpad bug 1607240 in unity8 (Ubuntu) "Unity8 default mouse wheel speed is 7.0 (it should be 1.0)" [Undecided,Confirmed] [08:52] That will break non-Xmir apps but it's a start. We need to fix scrolling in multiple components still [09:08] oh wow! that worked just fine! thanks duflu :D [09:08] Kind of. Except it's broken elsewhere now [09:09] yeah but it fixed muh firefox :D [09:09] for now [09:09] sort of [09:12] duflu: omg! this is super awesome! right? i mean the fact that you can change the scroll rate and it works. you can't do the same with unity7 right? i remember opening a such a bug like years ago [09:13] ohNoes_: Yeah but it's not labelled. Just magic [09:13] awesome black magic :D [09:20] anpok: Weird. QtMir multiplies vscroll by 15 to make angleDelta. And then on the way out just sets client vscroll = angleDelta. No division [09:55] duflu: smooth scrolling in gedit/Xmir only works with the keyboard or with the mouse wheel but only if the mouse pointer is over the scroll bar (when the scroll bar gets wider). if you try to scroll with the mouse wheel and the pointer is not hovering the scroll bar you get jerky scrolling, totally not smooth. maybe iz related [09:56] ohNoes_: I know. It's a bit messy right now. Unity8 multiples, decimates, and then only partially unmultiples scroll values. We need to clean it all up... but multiple parts all at once [09:57] "multiplies" I mean [09:57] ohNoes_: By the way the keyboard is unrelated. Different code [10:02] even so, i find it amazing that i can change the scroll speed! :D [10:02] +1 will use unity8 again [10:03] Hah. Yeah. It would appear more weird. Unity8 converts QtMir's scroll tick of 1 to 15. And then system settings' default scale is 0.5. So apps get +7/-7 instead of +1/-1 [10:03] The left hand doesn't know what the right hand is doing [10:04] X-) [10:04] Anyway, that's progress. My only goal was to fix Xmir if required and it's not required. Dinner time [10:05] duflu: yes, just tried in miral+Xmir. Seems fine [10:05] Yeah I would expect so [10:05] Good night === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === hikiko is now known as hikiko|ln === hikiko|ln is now known as hikiko === chihchun is now known as chihchun_afk === dandrader is now known as dandrader|afk [15:58] alan_g: should I avoid referring to mir::scene::Surface and try use miral::Window as much as possible? Will Window grow to encapsulate more than a single surface? [16:00] greyback_: try to use Window. A Window will always correspond to a Mir Surface (but that can have multiple bufferstreams) [16:00] alan_g: ack. I've no objection, just wanted to be sure which direction your api is going [16:02] Really I'd like Window to go the way that Application did and for it just to be an identifier. But the APIs to allow that are not there yet. [16:02] Once your stuff is working I will do another pass at making that work. [16:03] ok [16:20] greyback_: any time you can't use the miral APIs is a time we need to enhance miral. [16:21] But don't let that stop you getting things working [16:21] alan_g: true [16:21] but I know you've plans for them too [16:21] just want to avoid jmping on something you're about to kill === chihchun_afk is now known as chihchun [16:42] alan_g: I could do with your perspective. Please have a look at lp:~gerboland/miral/add-interactivity-to-WM [16:42] just het last commit. [16:42] I'm trying to expose functionality provided in the WMPolicy [16:42] * alan_g looks.. [16:43] but getting access to the created WMPolicy isn't obvious to me, as it's wrapped with a builder which is only run inside the BasicWM [16:46] Can you pass the constructor somewhere to register the interface you want it to support? The the constructor just does "connector->register(this)" ... and the destructor can deregister [16:59] Yep. That was an option I was considering. [17:00] I can't really think of anything else [17:00] thanks === dandrader|afk is now known as dandrader === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader