10:25 <wlb> wayland Issue #251 opened by () Wayland Cross Compilation failed for arm toolchain https://gitlab.freedesktop.org/wayland/wayland/-/issues/251

20:02 <flibitijibibo> just ran into something odd with pointer constraints... for pointer_constraints_unstable_v1 is it possible (or more importantly, intended) to set a region in lock_pointer to be screen-relative rather than surface-relative?

20:03 <flibitijibibo> main motivation is a ClipCursor implementation which is thankfully not relevant to the application for wl, but i did notice the region didn't specify a coordinate system (unless i missed a line) https://github.com/libsdl-org/SDL/pull/4892

20:05 <emersion> no, the region is in surface coords

20:06 <emersion> clients don't know their position on screen anyways

20:07 <flibitijibibo> that's what i figured - i could hypothetically see how a screen-relative constraint could work without broadcasting the window position or anything, but it seemed like a stretch

20:07 <flibitijibibo> so another function in the pile of "crap i don't need to write"