On Fri, 18 Aug 2023 18:00:48 GMT, Jose Pereda <jper...@openjdk.org> wrote:
>> Following the great insight @hjohn vocalized in one of the earlier PRs, at >> the end we are always dealing with integer pixel coordinates. >> >> so, to be correct, any computation that returns pixel coordinates must use >> snapped values. (it also means that any computation that does not result in >> pixel coordinates, might use unsnapped values, such as when we try to >> distribute a single pixel among many columns or nodes). > > Please, see my edited comment. If we are to change to snapped values > _everywhere_ (not only in the proposed fix of this PR), that would be a > broader change... no, I did not propose to change it everywhere. In this PR, in the code we change, we probably should [start to] use snapped coordinates. wouldn't you agree? ------------- PR Review Comment: https://git.openjdk.org/jfx/pull/1203#discussion_r1298721309