[07:01] <RAOF> Urgh, that's annoying.
[07:02] <RAOF> Okaaaay, I'll do the bit fiddling then...
[07:22] <Saviq> Morn \o
[08:34] <RAOF> Today:... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/a2eb962020c410a764f0007797f1b6db4fcf636d>)
[08:34] <RAOF>  * Today:... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/02dc5ebfc26d28d1b3ab4400348460da53a764dc>)
[08:34] <RAOF>  * Today:... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/69f41b437f56d864808e17cb788016f858509fdc>)
[15:44] <Saviq> sophie hey, so one of the problems I have with the bandwidth tests is that the gtk client app for for the dragged test seems to do nothing on slower hardware, any idea how to debug? I've used `cursor=software` to show that the tests works as expected, but it draggs nothingness :)
[15:44]  * Saviq uploaded a video: (37KiB) < https://libera.ems.host/_matrix/media/v3/download/matrix.org/mssmDvlMWbDLblSuwMVIDxMd/recording.mp4 >
[15:46] <Saviq> The app itself is silent:... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/54519c13c5fdd65e50e983a2f1ab869b59986cd7>)
[15:47] <Saviq>  * The app itself is silent:... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/16f733cd2983bd6fa47498b350c4adb583a77cad>)
[15:48] <Saviq> Looks stuck in some init phase...
[16:46] <Saviq> <Saviq> "Looks stuck in some init phase..." <- Here's `WAYLAND_DEBUG=1`: https://pastebin.ubuntu.com/p/Gkr9wgJM3T/, maybe it sheds a light
[16:49] <sophie-w> Will take a look 
[16:52] <Saviq> sophie-w: I have a system handy that I can give you access to
[16:53] <alan_g[m]> Today:
[16:53] <alan_g[m]> * extending the DnD spike
[16:53] <Saviq> Saviq: `ubuntu@10.102.157.81` has your SSH key, but it's likely to go away in a bit - let me know and I can re-provision it for you for a longer debugging session
[17:04] <sophie-w> <Saviq> "`ubuntu@10.102.157.81` has..." <- I can't get in, also `ping 10.102.157.81` doesn't work
[17:04] <Saviq> sophie-w: VPN?
[17:05] <sophie-w> I'm not on a VPN, do I need to be on the company VPN or something?
[17:05] <Saviq> Yes, company VPN. 10.* are intranet :)
[17:18] <Saviq> sophie if that still didn't work: `ssh 3jPdMxazEjm0rlmJGfbU:MDZjZjdiZjEudm0udXB0ZXJtLmludGVybmFsOjIyMjI=@uptermd.upterm.dev`
[17:18] <Saviq>  * sophie if that still didn't work:
[17:18] <Saviq> ```
[17:18] <Saviq> ssh 3jPdMxazEjm0rlmJGfbU:MDZjZjdiZjEudm0udXB0ZXJtLmludGVybmFsOjIyMjI=@uptermd.upterm.dev
[17:18] <Saviq> ```
[17:19] <sophie-w> Thanks, I was half way through figuring out how to get on the VPN, but that's easier
[17:21] <sophie-w> Do I need to install deps to make it work? Or is there something I'm missing?
[17:23] <Saviq> You're set. Need to run from mir-ci/frame so it picks up conftest.py
[17:24] <Saviq> sophie it only has deps for `mir_test_tools`
[17:24] <sophie-w> Saviq: Ah, thanks, that's the trick
[17:24] <Saviq> Best run -k 'dragged and tools', that gives you the failing test alone
[17:26] <Saviq> sophie  ?
[17:35] <Saviq> Let me try and disconnect, it's probably messing up with your terminal... let me know if that kills the session
[17:38] <Saviq>  * @sophie let me try and disconnect, it's probably messing up with your terminal... let me know if that kills the session
[17:38] <sophie-w> It was working but that killed my session
[17:39] <Saviq> Here's another one:
[17:39] <Saviq> ```
[17:39] <Saviq> ssh 7tSOqDcRiZw9zq5wU0O3:MWNlNmQ0OGEudm0udXB0ZXJtLmludGVybmFsOjIyMjI=@uptermd.upterm.dev
[17:39] <Saviq> ```
[17:39] <Saviq> Not sure how the multiplexing works ?
[17:43] <sophie-w> > std::exception::what: Failed to find any platforms for current system
[17:44] <sophie-w> Do I need to install and use XVFB?
[17:44] <sophie-w> Didn't seem to have this problem before the reset
[17:49] <Saviq> That's because there's a server stuck in the bg
[17:49] <Saviq> It's running on GBM there, no xvfb
[17:50] <sophie-w> Cool, fixed
[18:44] <Saviq> sophie to see what's on screen, I used ubuntu-frame-vnc and mir_demo_server as host, then nest the tests inside. But with that tests you'll see black, unless you `--cursor=software`
[18:44] <Saviq> Did the session die?
[18:45] <sophie-w> Yep, died again.
[18:45] <sophie-w> I suspect the issue was with d-bus again, and I was trying to repro locally 
[18:46] <sophie-w> Saviq: I've been using `WAYLAND_DISPLAY=wayland-88 grim ~/a.png && cacaview ~/a.png`
[18:46] <Saviq> Oh. That could explain indeed
[18:46] <sophie-w> (which displays the contents of the screen within the terminal using text lol)
[19:07] <Saviq> I can't reproduce locally, no session dbus running for my user and it completes fine. Unless a system-wide daemon placates it. But I'd think the GitHub runners don't have DBus either (and that's where we saw the gedit timeout, but not this issue)
[19:08] <sophie-w> Oh, hmm, my hunch might be wrong then
[19:10] <Saviq> (reprovisioning the thing); I'll try extending the timeout in the test, to see if the app ever pops up
[19:13] <Saviq> Didn't succeed in getting a session DBus on the remote machine, snap refused to start the server because of a changed cgroup... and I lost the machine before I was able to try with mir_demo_server from deb...
[19:46] <Saviq> sophie your hunch was right
[19:46] <Saviq> dbus-run-session pytest... makes it go with mir_demo_server, so now just have to solve the snap cgroup complaint
[19:47] <Saviq> Thanks for pointing at that, I'll take it from here