05:59 <LarstiQ> scherso: ah, PATH was bare?

17:10 <leon-p> I think git master accidentally re-introduced that one bug we had ages ago, where mapped keys would still send the un-press event to clients

17:11 <leon-p> ifreund: ^

17:58 <ifreund> leon-p: hmm, I did touch that logic in the last month or so, I could have broken it

18:03 <ifreund> anyhow I don't see anything obviously wrong from looking at the code and don't have the time to test myself/pull out gdb right now

18:03 <ifreund> please do open an issue though if you're certain there's been a regression

19:56 <tiosgz> leon-p & ifreund: well wev confirms it. the main problem is in Keyboard.handleKey() that the last else-if ignores that on release we've already sent the event if needed

19:57 <tiosgz> but i have a suspicion there's some sneaking-about depending on what eats the key and it could still end up in the input method or somewhere

19:58 <tiosgz> thus i can't just fix it rn. will try tomorrow when it isn't so late

19:58 <tiosgz> *depending on what eats the key and whether there's a -release mapping

20:45 <adamcstephens> Is there any fix I can apply to resolve windows that don't get a border? For example Firefox doesn't get one, and neither does ghostty (GTK)

20:45 <adamcstephens> For ghostty if there's something I can fix on the app side I can look into doing so

21:37 <novakane> adamcstephens: look for rule-add ssd in riverctl man page

22:21 <adamcstephens> novakane: Cool I’ll check into fixing it in my config, but I’d still like to know how to fix at least ghostty so it doesn’t require everyone to work around it

22:42 <novakane> adamcstephens: something like `riverctl rule-add -app-id 'ghostty' ssd`

22:43 <novakane> of course you need to use the right app-id for ghostty because I have no idea what it is