Scroll Kitty images with their scroll region #126

Merged

natsukium merged 2 commits from fix/scroll-region-images into main 2026-10-12 19:46:59 +09:00

Conversation 0 Commits 2 Files changed 13 +694 -12

natsukium commented 2026-10-12 19:21:08 +09:00

Owner

Copy link

Kitty graphics placements now scroll with the text inside a scroll region (DECSTBM) and on a reverse scroll (RI at the top margin, SD), following kitty's scroll_filter_margins_func:

• A placement wholly inside the scrolled rows moves with them; one that crosses a margin or lies outside the region stays put.
• One the scroll takes wholly out of the region is removed (PlacementRemoved).
• One taken partly past a margin loses the rows past it: its extent shrinks (a requested r= with it) and its source rectangle loses the pixels those rows showed, at the scale the image draws at (kitty's src_pixels_per_screen_pixel, including the one-axis c=/r= cases). What is left keeps its size and position on screen.
• The cursor advance after a placement at the bottom margin of a region goes through the same path.

The grid queues a new PtyEffect::RegionScrolled (adjacent scrolls of one band coalesce); the daemon re-sends each survivor as Placement, since PlacementsShifted moves every anchor. No wire change.

Two cases depart from kitty where its region path disagrees with its full-screen path, recorded in the explanation doc:

• A region anchored at the primary screen's top pushes its rows into scrollback; its placements go into history whole along with those already there, as on a full-screen scroll. Kitty clips them and leaves history placements behind, off their text.
• A full-screen reverse scroll uses the region rule. Kitty moves history placements down although no history text moves, and keeps those pushed off the bottom.

Tests: unit tests for move, removal, top clip, bottom clip with a partial last row, a lone-c= clip, placements crossing or outside the band, and the into-history case; a property test that survivors stay in the band, inside their image, at their old scale, and that placements outside the band are untouched; the image-mirror property test now exercises region scrolls and compares extents and source rectangles; a daemon test for the issue's repro.

Closes #40

Kitty graphics placements now scroll with the text inside a scroll region (`DECSTBM`) and on a reverse scroll (`RI` at the top margin, `SD`), following kitty's `scroll_filter_margins_func`: - A placement wholly inside the scrolled rows moves with them; one that crosses a margin or lies outside the region stays put. - One the scroll takes wholly out of the region is removed (`PlacementRemoved`). - One taken partly past a margin loses the rows past it: its extent shrinks (a requested `r=` with it) and its source rectangle loses the pixels those rows showed, at the scale the image draws at (kitty's `src_pixels_per_screen_pixel`, including the one-axis `c=`/`r=` cases). What is left keeps its size and position on screen. - The cursor advance after a placement at the bottom margin of a region goes through the same path. The grid queues a new `PtyEffect::RegionScrolled` (adjacent scrolls of one band coalesce); the daemon re-sends each survivor as `Placement`, since `PlacementsShifted` moves every anchor. No wire change. Two cases depart from kitty where its region path disagrees with its full-screen path, recorded in the explanation doc: - A region anchored at the primary screen's top pushes its rows into scrollback; its placements go into history whole along with those already there, as on a full-screen scroll. Kitty clips them and leaves history placements behind, off their text. - A full-screen reverse scroll uses the region rule. Kitty moves history placements down although no history text moves, and keeps those pushed off the bottom. Tests: unit tests for move, removal, top clip, bottom clip with a partial last row, a lone-`c=` clip, placements crossing or outside the band, and the into-history case; a property test that survivors stay in the band, inside their image, at their old scale, and that placements outside the band are untouched; the image-mirror property test now exercises region scrolls and compares extents and source rectangles; a daemon test for the issue's repro. Closes #40

natsukium added 1 commit 2026-10-12 19:21:08 +09:00

grid, daemon: scroll Kitty placements with their scroll region ... fa460c4332

Placements moved only on full-screen scrolls, so a program with a fixed
header and footer (DECSTBM) scrolled its text under or over a stationary
image, and a reverse scroll left images behind as well. Kitty moves each
placement that lies wholly inside the scrolled band, clips one that the
move takes partly past a margin, and drops one taken wholly outside.

The grid now queues a RegionScrolled effect for every band scroll that
is not one of the two full-screen cases, coalescing adjacent scrolls of
one band. The daemon runs kitty's scroll_filter_margins_func over the
placement table: the clipped rows come off the extent (and a requested
r=) and off the source rectangle at the scale the image draws at, as
src_pixels_per_screen_pixel computes it, so what is left keeps its size
and place. Survivors are re-sent as Placement because a uniform
PlacementsShifted cannot express a partial move or a clip; no wire
change is needed.

Two cases depart from kitty where its region path disagrees with its
full-screen one. A region anchored at the primary top carries its
placements into history whole, together with those already there, as a
full-screen scroll does; kitty clips them and leaves history placements
behind, off their text. A full-screen reverse scroll uses the region
rule; kitty moves history placements down although no history text
moves, and keeps those pushed off the bottom.

Assisted-by: Claude Code

natsukium added 1 commit 2026-10-12 19:33:29 +09:00

grid: state the rounding bound of coalesced region scrolls ... f0ac3750cc

The coalescing comment claimed exact equivalence; source cuts round to
whole pixels per clip, so split and merged scrolls can differ by a
pixel of source. The bound is what callers may rely on.

Assisted-by: Claude Code

natsukium merged commit 06a77f453c into main 2026-10-12 19:46:59 +09:00

natsukium referenced this pull request from a commit 2026-10-12 19:47:00 +09:00

Merge pull request 'Scroll Kitty images with their scroll region' (#126) from fix/scroll-region-images into main

Sign in to join this conversation.

Reviewers

No reviewers

Labels

Clear labels   

Area/ci

CI workflows (.forgejo/workflows/)

  

Area/cli

felis-cli crate, the felis front-door binary

  

Area/client

felis-client crate, the winit GUI frontend

  

Area/client-core

felis-client-core crate, shadow screen, config, connect_or_spawn_daemon

  

Area/daemon

felis-daemon crate, session pool, IPC server

  

Area/deps

Dependency bumps

  

Area/docs

Documentation tree when no narrower area fits

  

Area/grid

felis-grid crate, cells, scrollback, image store

  

Area/manual

User-facing docs: tutorials, how-to, user-facing reference pages

  

Area/nix

Nix flake, home-manager module, package definitions

  

Area/protocol

felis-protocol crate, felis.proto, frame layout, version constants

  

Area/pty

felis-pty crate, PTY backends

  

Area/render

felis-render-wgpu crate, GPU pipeline, atlas

  

Area/shaping

felis-shaping crate, swash shaper, font fallback

  

Area/skills

Agent skills (skills/felis, .agents/skills/)

  

Area/transport

felis-transport crate, framing, socket/pipe, SSH carrier

  

Area/vt

felis-vt crate, VT parser, dispatch

  

Compat/Breaking

Breaking change that won't be backward compatible

  

Kind/Bug

Something is not working

  

Kind/Documentation

Documentation changes

  

Kind/Enhancement

Improve existing functionality

  

Kind/Feature

New functionality

  

Kind/Performance

Throughput, latency, or memory cost

  

Kind/Security

This is security issue

  

Kind/Testing

Issue or pull request related to testing

  

perf-regression-accepted

Passes the instruction-count bench gate for an intended cost; state why in a PR comment

  

Reviewed

Confirmed

Issue has been confirmed

  

Reviewed

Duplicate

This issue or pull request already exists

  

Reviewed

Invalid

Invalid issue

  

Reviewed

Won't Fix

This issue won't be fixed

  

Status

Blocked

Something is blocking this issue or pull request

  

Status

Need More Info

Feedback is required to reproduce issue or to continue work

  

Status

Upstream

Waiting on a change in an upstream project

No labels

Area/ci

Area/cli

Area/client

Area/client-core

Area/daemon

Area/deps

Area/docs

Area/grid

Area/manual

Area/nix

Area/protocol

Area/pty

Area/render

Area/shaping

Area/skills

Area/transport

Area/vt

Compat/Breaking

Kind/Bug

Kind/Documentation

Kind/Enhancement

Kind/Feature

Kind/Performance

Kind/Security

Kind/Testing

perf-regression-accepted

Reviewed

Confirmed

Reviewed

Duplicate

Reviewed

Invalid

Reviewed

Won't Fix

Status

Blocked

Status

Need More Info

Status

Upstream

Milestone

Clear milestone

No items

No milestone

Assignees

Clear assignees

No assignees

1 participant

Notifications

Subscribe

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference

felis-terminal/felis!126

No description provided.