[01:01] <RAOF> So, people have been talking about wlcs again in #_oftc_#wayland:matrix.org, and Daniel Stone brought up a [licensing point](https://matrix.to/#/!SMVhYgTjRgVBkaCWjA:matrix.org/$FV8Xz7zxYEXF_Y8rGf13hY9rsNTH4hArDQJB8mdqjaw?via=matrix.org&via=gnome.org&via=kde.org) I had not otherwise considered (consultancy being paid to implement a specific extension).
[02:35] <RAOF> Boo! We need a bunch more wlcs work before a test for #2621 can be usefully written. :(
[05:09] -GitHub[m]:#mir-server- **[MirServer/mir]** bors[bot] merged [pull request #2621](https://github.com/MirServer/mir/pull/2621): Bind shm textures when we create the buffers
[05:09] -GitHub[m]:#mir-server- **[MirServer/mir]** bors[bot] closed [issue #2620](https://github.com/MirServer/mir/issues/2620): Latest swaybg (on 22.10) doesn't paint on Mir
[07:38] -GitHub[m]:#mir-server- **[MirServer/mir]** Saviq opened [pull request #2641](https://github.com/MirServer/mir/pull/2641): Revert "Merge #2622"... (full message at <https://libera.ems.host/_matrix/media/r0/download/libera.chat/a3a14fc3f21efbeec1667c0db77f1c18ac3e1847>)
[07:54] <Saviq> If this is another GCC-12 irk, this isn't funny anymore
[07:58] <alan_g[m]> It affects 20.04 as well?
[08:00] <Saviq> Yeah it does. So not a 12 bug, and the fix is likely just:... (full message at <https://libera.ems.host/_matrix/media/r0/download/libera.chat/721891eafc10fa1633f8a7cb594915536403e370>)
[08:00] <Saviq> Here: 
[08:00] <Saviq> https://github.com/MirServer/mir/pull/2622/files#diff-f810b494f93934426a29e31ec7d1a41c50c20979123b4c37b933cfc6462d24a9R818
[08:01] <Saviq> Hmm or not
[08:02] <Saviq> `timestamp` is used below, but indeed never initialized
[08:02] <alan_g[m]> We need better reproduction of the PPA build environments before uploading
[08:02] <alan_g[m]> Or even merging
[08:03] <Saviq> That's https://github.com/MirServer/mir/pull/2530
[08:03] <Saviq> Let me rebase on that to confirm it would catch
[08:04] <Saviq> *rebase that on main
[08:30] -GitHub[m]:#mir-server- **[MirServer/mir]** bors[bot] merged [pull request #2641](https://github.com/MirServer/mir/pull/2641): Revert "Merge #2622"
[09:05] <Saviq> So yeah, #2530 did fail in the same way the PPA did. If only I could get ccache to work there…
[09:27] <alan_g[m]> That would be nice
[10:09] -GitHub[m]:#mir-server- **[MirServer/mir]** bors[bot] merged [pull request #2633](https://github.com/MirServer/mir/pull/2633): Add quirks for false-positive compiler warnings
[10:17] <Saviq> OK so that should mean a 100% working PPA build. I think.
[11:04] <alan_g[m]> 🎆 I have my background back on 22.10+ppa:mir-team/dev+Miriway
[12:23] <Saviq> Hmm mine's gone again…
[12:25] <Saviq> Back with `/dev`, my local build must've been messing with it
[12:36] -GitHub[m]:#mir-server- **[MirServer/mir]** Saviq closed [issue #2639](https://github.com/MirServer/mir/issues/2639): Shutdown crash when OSK is running
[15:36] <Saviq> OK this is looking promising
[15:36] <Saviq> `cache hit rate                     20.47 %`
[15:37] <Saviq> Now I need to work out how to fix precompiled headers with ccache in the mix
[15:37] <Saviq> But that's a tomorrow problem. See you all o/
[16:17] <alan_g[m]> I would guess a good cache hit rate is going to help more than precompiled headers
[16:30] <Saviq> Right, but we want PCH to reproduce PPA builds
[16:31] <Saviq> It's only clang that complained
[17:25] -GitHub[m]:#mir-server- **[MirServer/mir]** wmww opened [pull request #2642](https://github.com/MirServer/mir/pull/2642): Verify serials again
[17:25] -GitHub[m]:#mir-server-  
[17:25] -GitHub[m]:#mir-server- > This reverts commit 71a40e7, reversing changes made to 387705a. I fixed the obvious uninitialized use in xwayland_surface.cpp that seemed to cause the PPA failure. Not sure why neither local builds nor CI picked up on that but PPA did.
[17:46] -GitHub[m]:#mir-server- **[MirServer/mir]** wmww opened [issue #2643](https://github.com/MirServer/mir/issues/2643): Cursor only updates once per enter on XWayland
[17:46] -GitHub[m]:#mir-server-  
[17:46] -GitHub[m]:#mir-server- > When the cursor enters an X11 window it is correctly set. After that as it moves around the window the cursor is not updated until it leaves and re-enters. I suspect this is to do with #2617, but I need to investigate.
[23:57] -GitHub[m]:#mir-server- **[MirServer/mir]** wmww opened [pull request #2644](https://github.com/MirServer/mir/pull/2644): Improve Client class lifetime management
[23:57] -GitHub[m]:#mir-server-  
[23:57] -GitHub[m]:#mir-server- > Libwayland notifies us that clients are destroyed before it cleans up all of the client's resources. The `wl_client` isn't freed until that is done. Previously, our WlClient class being destroyed before the resources, and logic was needed to prevent using using it after. This Makes `Client` an interface in `mir/wayland` and had `Resource` hold a shared pointer to it. This way the client isn't destroyed until it's last resource is. It also
[23:57] -GitHub[m]:#mir-server- makes the client class more readily available.