https://bugs.kde.org/show_bug.cgi?id=506097
--- Comment #15 from [email protected] --- the same rules apply for the offsets of other monitors, so they can be side by side no matter in which direction you round the resolution itself is meaningless here, just that the (scaled) offsets of the monitors do not match the sum of the resolutions On 13/10/2026 20:08, Zamundaaa wrote: > https://bugs.kde.org/show_bug.cgi?id=506097 > > --- Comment #14 from Zamundaaa <[email protected]> --- > (In reply to armin.novak from comment #8) >>> A gap-less Wayland output configuration can have gaps in Xwayland. I'm not >>> sure that is avoidable. >> how so? This sounds like integer rounding basics? > It's literally mathematically impossible to have that work in all cases, > because we have (at least) two differently scaled integer coordinate systems. > > Let's say you had a screen at position 3, with size 4. The second screen would > be at position 7 then... Now translate that with an Xwayland scale of 120% and > you get a gap, even though the actual positions are perfectly touching: > 3+4=7 > round(3*1.2)+round(4*1.2) = 4 + 5 = 9 != round(7 * 1.2) = 8 > > The only way to reliably make that work without gaps is if we modify the > resolution of each screen on the X11 side, which I don't think is an > acceptable > workaround. > -- You are receiving this mail because: You are watching all bug changes.
