[08:18] Mornin' === dot-tobias is now known as dot-tobias[away] === dot-tobias[away] is now known as dot-tobias === dot-tobias is now known as dot-tobias[away] === dot-tobias[away] is now known as dot-tobias [10:50] Saviq I'm experimenting with Frame and multiple displays. I *thought* we'd landed changes to "round robin" which output a client appears on. But, so far, they all appear on the first output. Do you have examples that work as expected? [10:52] Well. We did: https://github.com/MirServer/ubuntu-frame/blob/main/src/frame_window_manager.cpp#L178 [10:54] I've not dug yet, but it is possible the examples I tried actually specify the output. Just hoped you had working examples [10:54] I did, yes [10:54] Coming up [10:54] Do you remember what they were? [10:55] glmark2-es2-wayland, for example, just coming up with a full setuop [10:55] -o [10:57] * Saviq sent a code block: https://libera.ems.host/_matrix/media/v3/download/libera.chat/20ca6ae5273031d3ad09b917cc89dac8e4c1cb78 [10:57] Then, run multiple clients and they get placed on display 1 and 3 in turns [10:59] * Saviq uploaded an image: (3434KiB) < https://libera.ems.host/_matrix/media/v3/download/matrix.org/FviJLxckSMvWGypFWDHoskfP/image.png > [10:59] We need eglmark2 with the Kudu ;D [10:59] OK, I see what is happening but... [11:00] Can use waybg to confirm that when you do ask for an output, you get placed where requested [11:00] Try running glmark2-es2-wayland --fullscreen [11:01] Sounds like it's requesting output 1? [11:02] Yes: `[ 502315.342] -> xdg_toplevel@18.set_fullscreen(wl_output@10)` [11:03] So that's as designed [11:03] Is output mandatory when requesting fullscreen? [11:04] https://wayland.app/protocols/xdg-shell#xdg_toplevel:request:set_fullscreen [11:04] It is [11:04] Actually no - allow null [11:06] The trouble is that all the examples snapsped up are specifying it. (No it is optional) [11:06] WPE doesn't, glmark without --fullscreen doesn't… [11:06] *example snaps I tried [11:08] I don't think we can change Frame's behaviour. [11:09] Another example is waybg, where you can specify the output [11:09] Do you have that snapped? [11:10] Don't think so, no [11:10] I think the Frame behaviour is right. Just need some "good" examples for the HOWTO [11:11] Thanks for being a sounding board. Am unblocked now === dot-tobias is now known as dot-tobias[away] === dot-tobias[away] is now known as dot-tobias [12:02] Huh?! I have photographic evidence that it worked. But now i can't get it to do so again. === dot-tobias is now known as dot-tobias[away] [12:18] OK, this is fairly rough (especially, I need to do some better pictures) but I'll share now: [12:18] https://discourse.ubuntu.com/t/how-to-configure-ubuntu-frame-for-multiple-outputs/33012 === dot-tobias[away] is now known as dot-tobias [12:35] Noice. Will review. I'm thinking we could take screenshots instead? [12:36] Yeah, but that doesn't show what is on each output [12:36] Will come up with something [12:36] Will you hook up to https://discourse.ubuntu.com/t/mir-documentation/27559? [12:36] https://mir-server.io/docs/how-to-maintain-mir-documentation#heading--navigation-pane [12:37] I will, once polished === dot-tobias is now known as dot-tobias[away] === dot-tobias[away] is now known as dot-tobias === dot-tobias is now known as dot-tobias[away] === dot-tobias[away] is now known as dot-tobias [14:38] -GitHub[m]:#mir-server- **[MirServer/mir]** AlanGriffiths opened [issue #2758](https://github.com/MirServer/mir/issues/2758): Scaling doesn't work on overlapping screens... (full message at ) [14:39] -GitHub[m]:#mir-server- **[MirServer/mir]** AlanGriffiths edited [issue #2758](https://github.com/MirServer/mir/issues/2758): Scaling doesn't work on overlapping outputs [14:58] -GitHub[m]:#mir-server- **[MirServer/mir]** bors[bot] merged [pull request #2741](https://github.com/MirServer/mir/pull/2741): ci: fix working with jammy in GitHub Actions [15:27] -GitHub[m]:#mir-server- **[MirServer/mir]** mariogrip opened [pull request #2759](https://github.com/MirServer/mir/pull/2759): Cleanup miroil mirbuffer and align with upstream qtmir [15:37] -GitHub[m]:#mir-server- **[MirServer/mir]** mariogrip edited [pull request #2759](https://github.com/MirServer/mir/pull/2759): Cleanup miroil mirbuffer and align with upstream qtmir [15:49] alan_g it should be `position: [2560, 0]`, no? Otherwise it overlaps by half? [15:51] Also, how did you get glxgears to work on Frame? :D [16:09] * Saviq uploaded an image: (54KiB) < https://libera.ems.host/_matrix/media/v3/download/matrix.org/PLxOrCLBWwssuSmAfgKWgxNN/shot.png > [16:10] I was thinking something along those lines for the screenshots [16:12] * Saviq uploaded an image: (54KiB) < https://libera.ems.host/_matrix/media/v3/download/matrix.org/XVFYfoYpjqnIWdkuBFWsPesv/shot.png > [16:18] * Saviq uploaded an image: (251KiB) < https://libera.ems.host/_matrix/media/v3/download/matrix.org/fAWIeanAqBJPtPXvlBMrujYV/shot.png > [16:22] I tweaked the document very slightly — I would probably strip the "Here's the full file" of all the commented out bits [16:27] "Also, how did you get glxgears..." <- It is the iot-example-graphical-snap branch I had installed [16:29] Looks like screenshots get confused by scale, btw… [16:30] Another scale bug to log then. [16:56] -GitHub[m]:#mir-server- **[MirServer/mir]** Saviq added bug to [issue #2760](https://github.com/MirServer/mir/issues/2760): Screenshotting not scale aware [16:56] -GitHub[m]:#mir-server- **[MirServer/mir]** Saviq opened [issue #2760](https://github.com/MirServer/mir/issues/2760): Screenshotting not scale aware... (full message at ) [17:33] G'night o/