[01:06] <karmavil[m]> Me too. What's is going? ..
[01:07] <karmavil[m]> * is going on? ..
[01:11] <karmavil[m]> Jenga ..
[04:52] <karmavil[m]> By the way I wasn't talking about mir, but software in general
[05:21] <karmavil[m]> Like bash: info: command not found 🙃 sorcery
[07:36] <RAOF> Wat?
[07:36] <RAOF> #Wat?
[07:36] <RAOF>  * # _Wat_?
[07:38] <Saviq> Oh the suspense…
[07:39] <RAOF> gdb is complaining about not being able to find source files for various mesa bits.
[07:39] <RAOF> Which is perfectly reasonable, because the mesa tree does not contain any files of that name.
[07:39] <RAOF> But I don't know how references to those filenames could possibly have ended up in the DSOs?
[07:45] <alan_g[m]> That is a worrying question
[07:48] <RAOF> I have just built and installed a set of unstripped, -O0 mesa packages, and those files just don't exist. Which is why gdb can't find them.
[07:48] <RAOF> But at least with the fully-unoptimised build I can blindly single-step through until I do find some source, which happens.
[08:03] <alan_g[m]> 🥳 it is USN day again
[08:04] <Saviq> But why do we have CUPS in the OSK…
[08:13] <Saviq> Either way. I'll get it done
 "But I don't know how references..." <- Maybe a Not symbols found symbol? Kind of no enty point or something
[08:29] <karmavil[m]> I'm just guessing for fun. That's above my knowledge. But if you need it, maybe binutils could help. (If it's important)
[08:29] <RAOF> Today:
[08:29] <RAOF> * (Eventually) Got source debugging working in mesa
[08:29] <RAOF> * Realised why cross-display rendering wasn't always working on my dual-gpu desktop (I need to make sure the texture is uploaded *on the GPU it's going to be rendered on*)
[08:29] <RAOF> * Identified that the new, broken code is allocating exactly the same type of buffer as the old working code and that the mesa driver is interpreting it in exactly the same way - it *says* the buffers have identical layouts and tiling, but then the rendering is clearly mistiled on the new code and correct on the old 🤷‍♀️
 "Maybe a Not symbols found symbol..." <- Nah. It can find the entrypoints (and give the values of parameters, etc), it just can't find the source to list.
[08:35] <karmavil[m]> Jenga! Ok. You don't have to explain but I was wondering guy debug optimized code. I didn't know it was even posible.
[08:35] <karmavil[m]> Again, you don't have to explain. There is still too much to learn for me
[08:36] <karmavil[m]> s/guy/why/
[08:41] <karmavil[m]> Oh I see. 0 is not, and 4 a mith. That explains a lot
[08:43] <karmavil[m]> s/mith/myth/
 "🥳 it is USN day again" <- alan_g if you encounter this while I'm away, things are better now:
[11:14] <Saviq> https://github.com/MirServer/checkbox-mir/commit/8f363088c41e6b6e03911850037042a2702ffef8
[11:15] <Saviq> I'm hoping that by the time I'm away this will be in the store, too, and things autoconnect
[11:15] <Saviq> Just adding a README now
[11:16] <alan_g[m]> Saviq: That will help
 "That will help" <- Done.
[11:26] <Saviq> Gist of it: checkbox-mir.snaps will guide you through testing the different snaps (not all of them, yet - just those I encountered today, and some adjacent ones) that we maintain
[11:34] <alan_g[m]> OK. A baby step towards full automation. 😀
[16:46] <alan_g[m]> Today:
[16:46] <alan_g[m]> * More experimenting with Python/GTK/DnD (and deciding it is a weird abstraction over something that shouldn't be weird) 
[17:21] <Saviq> Today:
[17:21] <Saviq> - was working too long, will report on Monday morning…