Repository navigation
libghostty: expose Kitty graphics virtual placements (Unicode placeholders) to embedders #14698
Unanswered
nckre
asked this question in
Feature Requests, Ideas
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Ghostty renders Kitty graphics Unicode placeholders (
a=p,U=1+U+10EEEEcells) fine, but libghostty does not.Tried this in Rex where direct placements (
a=T, anda=tfollowed bya=p) render correctly but placements using Unicode placeholders don't. The cells show as raw glyphs and no image is drawn.This matters e.g. for TUI apps like terminal-browser which shows a blank pane in Rex, while its page loads and takes input normally.
Gaps seems to be according to Claude review:
include/ghostty/vt/kitty_graphics.hcan identify a virtual placement (GHOSTTY_KITTY_GRAPHICS_PLACEMENT_DATA_IS_VIRTUAL), but the geometry calls return nothing for one:ghostty_kitty_graphics_placement_rect: "Virtual placements (unicode placeholders) return GHOSTTY_NO_VALUE."ghostty_kitty_graphics_placement_viewport_pos: returnsGHOSTTY_NO_VALUE"when the placement is a virtual (unicode placeholder) placement."So an embedder can see that a placeholder image exists but can't work out which cells it covers or which part of the image belongs in each cell. #12111 exposed image data through the render state; this request is the remaining piece for virtual placements.
All reactions