12:00 <snakedye> @leon-p sick! how did you do it?

16:02 <ifreund> leon-p: the proper way to bind the wlroots API without doing manual libinput bindings would be to make the function return an opaque pointer and then use @ptrCast() to pass it to libinput

16:03 <ifreund> i.e. `const LibinputDevice = opaque{};` in the zig-wlroots file

16:04 <ifreund> novakane: I just got back from some backpacking last night, I'll be updating everything to zig 0.8.0 first and then I'll have a look at the damage tracking thing

16:04 <ifreund> inclement: I think you're 100% right about the subsurfaces, thanks for the tip!

16:05 <novakane> ifreund: no worries :)

16:07 <novakane> btw I pushed an update for zig on void since I saw you looked like busy irl, not sure if this is a thing only the maintener can do

16:10 <ifreund> novakane: looks great, thansk

16:11 <novakane> no problem

16:12 <novakane> also yeah I think inclement is right too, but not sure how to do it

16:29 <inclement> I think it just needs an iteration over subsurfaces when mapping a new surface (in addition to binding to the new_subsurface event) - I made a POC patch to play with zig but I couldn't reproduce the original issue so I didn't test it, it could easily have been wrong

16:30 <inclement> I think you can see the same think in sway's surface handling

16:32 <novakane> allright I'll try

16:37 <inclement> novakane: What I was trying is https://gist.github.com/inclement/3fab077b5df19152cf02ca538cf12795 - but 1) this is probably not the best solution, 2) it might just be wrong, 3) I only put it in xdg-popup, you might need the same in other map events. If you're trying though, maybe it can help

16:41 <novakane> inclement: ok thanks it's a starting point I'll see what I can do

16:47 <novakane> debugging is so annoying since it works great in river in river so I always need to quit river and relaunch to test

16:48 <ifreund> novakane: if you don't get it no worries, it should only take me a minute or two

16:51 <novakane> ifreund: damn that hurt :P

16:52 <ifreund> novakane: it should be 3 while loops calling Subsurface.create()

16:53 <ifreund> actually no, only 2

16:53 <ifreund> one in xdg-toplevel map and one in xdg-popup map

17:58 <leon-p> ifreund: where were you backpacking, if you don't mind me asking?

18:35 <ifreund> leon-p: badlands national park! I decided to spend a bit of time in the US, see my family for the first time in over a year, get my covid vaccine, and spend a lot of time outdoors :)

18:39 <leon-p> that sounds great!

18:42 <leon-p> I really need to get out more as well. Used to have great stamina but in the recent year I got a bit to comfortable :D

18:43 <ifreund> I did a lot of distance running back in highschool, both cross country and track & field so I've got a pretty good base there

18:44 <ifreund> been running a bit more often on my own since the whole pandemic started too since I get restless if I sit around for more than a few days in a row

18:45 <ifreund> hmm, I wonder if we should have a #river-offtopic

18:45 <ifreund> don't really this kind of thing cluttering up the logs as nice as it is xD

18:52 <novakane> in the same time it's not like it's a really busy channel so not sure if off topic is really annoying

19:04 <dedguy21> Getting the following build error:

19:05 <ifreund> dedguy21: before you even paste the build error, what zig version?

19:06 <ifreund> zig 0.8.0 has just been released and river requires it

19:07 <dedguy21> Ah, thanks.

19:10 <novakane> ifreund: oh i just realized when I made the wlr_text_input_v3.h bind I didn't had a const in wlroots.zig

19:11 <ifreund> novakane: ah whoops, I shoulda caught that

19:11 <novakane> do you mind if I'll add it in my other PR?

19:12 <dedguy21> Yup updated zig 0.8, builds like a charm

19:13 <ifreund> novakane: probably easiest to just open a new PR. If you'd rather put it in the same one please put it in a separate commit

19:13 <ifreund> dedguy21: nice :)

19:17 <ifreund> neat, zls completions for packages started working when I updated river to zig 0.8.0

19:17 <ifreund> I guess my zls was trying to use zig master to compile its build runner or something like that

19:18 <novakane> ifreund: no problem I'll open an other PR, is it sorted in wlroots.zig?

19:18 <ifreund> wlroots.zig is very loosly structured

19:18 <ifreund> just put it somewhere that makes sense to you

19:19 <novakane> allright

19:23 <novakane> anyway, I'll always use zls with zig master so didn't see the difference, I just wish when looking for a std function it woudln't open the const in std.zig

19:40 <novakane> wait, there is a typo in the text_input filename too smh

19:46 <ifreund> looks good now :)

19:56 <novakane> sooo how close are we to the first river release now with zig 0.8.0 :P

20:04 <ifreund> probably won't happen before I get back to aachen in ~2 weeks

20:06 <ifreund> other than that, there are a few open bugs that would be nice to fix

20:06 <ifreund> I'd also like to put some time into at least thinking about river-control-v2 so we can keep that minimally breaking

20:07 <novakane> I see, sounds fair

20:08 <novakane> also I ask that whereas I don't even intend to stay on a stable release :P

20:09 <ifreund> yeah, I don't really expect anyone involved in the development to want to stick to a stable release

20:10 <ifreund> the point of doing the release is mostly so distros have something to package so that more people get hooked on river/zig and start contributing

20:12 <novakane> yeah, like an issue open like 2 days ago, also for distro that don't really package from git master it sure is important

20:13 <ifreund> as demonstrated by how many bug reports we've gotten about the AUR package, packaging river master with the mix of zig and wlroots versions floating around with breaking changes doesn't go well

20:15 <novakane> right without zig and wlroots it would be good but like this it's too much

20:25 <novakane> well now all wlroots bind for IME support in river are here for dnkl to add it :P