feat(window): add option to allow size persistence between app launches #9422
Conversation
|
should this be per window or windows in general |
|
windowrule, preferably |
okay... |
|
not sure this is possible ive tried some shit but apps start flipping out with their scaling and shiz edit - maybe we can just reapply each apps previous floating size when floating is triggered or something |
|
can you give me opinions on this @vaxerski |
|
can't you just gate |
|
if we stop telling apps they are maximized then some apps like firefox dont scale properly |
|
and they dont even remember their previous floating size |
👍
should easily be |
what about this |
|
might be our UV calculations not liking CSD. Some apps, notably gtk ones, will draw csd no matter what unless they are maximized (because then the protocol doesn't allow them) I would love if the xdg_shell spec said the same about tiled modes but it doesn't and I can't suggest the change due to being banned. |
damn bro is banned 💀 thats crazy... |
freedesktop shenanigans |
ghost
marked this pull request as ready for review
|
this is ready ✔️ |
ghost
mentioned this pull request
|
wiki : hyprwm/hyprland-wiki#983 |
|
we could do this for x windows as well, though i dont know if that will blow some app up lol or if its even necessary |
vaxerski
left a comment
vaxerski
left a comment
There was a problem hiding this comment.
I don't see where this disables the maximized hints. It doesn't, in fact.
yeah actually we dont need to disable maximize hints, if we do, some apps like firefox wont scale properly. if we do it this way, apps like firefox will scale properly, and all apps will remember their size after app restart. do you want me to rename the window rule to somethin else? |
|
we could just keep it how it is, and just masquerade that it disables maximize hints |
|
I'll just rename the window rule |
|
so ive found a bug with this, or is it a feature instead? open 1 window resize it down to a size, and keep it open open a new window, make it float, see that it has become the same size as window 1 now resize window 2 to a different size, and then reopen it. window 2 will not keep the new size, instead it will keep window 1's size until window 1 has been closed window 1 and 2 are the same app btw |
|
this doesnt happen for different apps. so not an issue in that department |
|
the current implementation treats windows differently based on class+title, so if a user matches with an asterisk; this would become a problem in the users perspective? |
|
any opinions on this "bug"? this is ready to be merged otherwise |
|
hm, good question. Dunno what we should do in that case. |
|
conflicts |
|
any reason for the holdup on this @vaxerski is it cuz #9422 (comment) ? |
|
i dont think thats really fixable with our current approach |
|
yeah, the reason is I forgor |
|
already done |
|
As already mentioned in #9171 , this doesn't solve the issue, and in general I think it's the wrong approach. There should just be an option to not tell windows that they are maximized. No workaround with storing sizes or whatever. |
|
|
@jbirnick our uv calculations are already a complete spaghetti due to there being so many things that can alter a surface's presentation wrt. the texture, and it seems that adding xdg geometries breaks the windows. The problem is that Only maximized does that, which means we have to rely on it to avoid weird textures for now. |
…size` is set. fix hyprwm#9422.
…s set (#10212) * fix(window): use stored size for new floating window when persistentsize is set. fix #9422. * fix: replace `std::any_of` with `std::ranges:any_of` * fix: use initialClass and initialTitle when storing sizes on close * fix: add `xdgTag` as a new indicator * fix: no {} * fix: format with clang-format
3 participants
fixes #9171