Read on in a thread from where reading is, not from its root #255

Merged

littlefox merged 4 commits from issue-251-thread-walk into main 2026-10-22 01:50:50 +00:00

Conversation 0 Commits 4 Files changed 4 +202 -18

littlefox commented 2026-10-21 23:31:36 +00:00

Owner

Copy link

Closes #251
Closes #259

In threads mode, every step listed the whole thread from its root (one get-text-stat per round trip, up to 2 000) and then looked for the first unread text. Now ReadState::thread_at remembers the text last read; the next unread is looked for after it (below, next siblings, ancestors' next siblings), and only if nothing is unread there, from the root. Both stop at the first unread text, and a post's replies are fetched together. thread_root no longer stops 64 steps up.

Measured with tools/perf.sh (#225, merged in here for that), 300 steps along the 1 500-long chain, 30 ms each way:

	baseline	after
slowest step	13 822 ms	127 ms
server calls	3 448	901 (−74 %)
step p50	115 ms	105 ms

Also #259: commands.spec's skip-tree test turns read-by-seen off (a slow runner read the comment by seeing it before the skip).

Verified: new devserver test long_chain_reading (121-text chain, at most 12 calls a step; the old code took 126 in one step, checked); all 15 devserver tests incl. thread_reading, reading_order, related_first and the mute tests pass; full e2e suite on three profiles 326 passed, 1 flaky (mute.spec on iPhone, a lost click, noted on #229); lint.

🤖 Generated with Claude Code

https://claude.ai/code/session_01XJW2ocwLunyaLiUwCBHEP8

Closes #251 Closes #259 In threads mode, every step listed the whole thread from its root (one `get-text-stat` per round trip, up to 2 000) and then looked for the first unread text. Now `ReadState::thread_at` remembers the text last read; the next unread is looked for after it (below, next siblings, ancestors' next siblings), and only if nothing is unread there, from the root. Both stop at the first unread text, and a post's replies are fetched together. `thread_root` no longer stops 64 steps up. Measured with `tools/perf.sh` (#225, merged in here for that), 300 steps along the 1 500-long chain, 30 ms each way: | | baseline | after | |---|---|---| | slowest step | 13 822 ms | 127 ms | | server calls | 3 448 | 901 (−74 %) | | step p50 | 115 ms | 105 ms | Also #259: commands.spec's skip-tree test turns read-by-seen off (a slow runner read the comment by seeing it before the skip). Verified: new devserver test `long_chain_reading` (121-text chain, at most 12 calls a step; the old code took 126 in one step, checked); all 15 devserver tests incl. `thread_reading`, `reading_order`, `related_first` and the mute tests pass; full e2e suite on three profiles 326 passed, 1 flaky (mute.spec on iPhone, a lost click, noted on #229); lint. 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_01XJW2ocwLunyaLiUwCBHEP8

littlefox added 4 commits 2026-10-21 23:31:36 +00:00

Measure first: a big meeting, a perf spec and a desktop memory run ... c5abe1fe63

0.6 is about performance (D-057), and every fix should come with a
before/after number. tools/perf.sh seeds a big meeting (NEOKOM_SEED_PERF:
~2 800 texts, a 1 500-long chain, 300 more channels, 2 500 letters) on a
lyskomd kept in memory, puts 30 ms each way between relay and server,
and runs e2e/perf/perf.spec.ts in Chromium: memory (JS heap and the
app's wasm), DOM nodes, main-thread time, long tasks, frames while
scrolling, step times and server calls. tools/desktop-perf.sh measures
the desktop app's RSS and idle CPU. tools/perf-table.py compares runs.

The app exposes its wasm memory and the session's call count to the page
for this; Chromium shows no WebSocket frames sent from wasm.

Refs #225

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XJW2ocwLunyaLiUwCBHEP8

Read on in a thread from where reading is, not from its root ... 085f29d93b

In threads mode, every step listed the whole thread from its root, one
text stat per round trip, and then looked for the first unread text. In
a 1 500-long chain the first step made ~1 500 sequential calls (12 s and
more over a real network). Now the engine keeps the text last read in the
thread and looks after it: below it, then its next siblings, then its
ancestors' next siblings; only if nothing is unread there, from the root.
Both stop at the first unread text and fetch a post's replies together.
thread_root no longer stops 64 steps up.

The devserver test long_chain_reading reads a 121-text chain and allows
at most 12 calls a step; the old code took 126 in one step.

Closes #251

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XJW2ocwLunyaLiUwCBHEP8

Perf: a scenario for reading cards by scrolling past them ... 91bc19544c

Reading with space marks few cards by being seen, the case #228 is
about. This scrolls slowly through a post's 120 unread replies.

Refs #225

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XJW2ocwLunyaLiUwCBHEP8

Merge remote-tracking branch 'origin/issue-225-measure' into issue-251-thread-walk f61aa2fffd

littlefox referenced this pull request 2026-10-22 00:23:39 +00:00

Flaky: commands.spec "skip a comment tree", the comment read by being seen #259

littlefox added 2 commits 2026-10-22 00:23:44 +00:00

commands.spec: no read-by-seen while skipping a comment tree ... 8a58aaba83

On a slow runner the comment was read by being seen before the skip.

Refs #259

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XJW2ocwLunyaLiUwCBHEP8

Merge remote-tracking branch 'origin/main' into issue-251-thread-walk f4d9cdcb49

littlefox merged commit dcebe3137e into main 2026-10-22 01:50:50 +00:00

littlefox referenced this pull request from a commit 2026-10-22 01:50:51 +00:00

Merge pull request 'Read on in a thread from where reading is, not from its root' (#255) from issue-251-thread-walk into main

littlefox referenced this pull request 2026-10-22 02:26:34 +00:00

Reading a thread at a time walks the whole thread from its root, every step #251

littlefox referenced this pull request 2026-10-22 02:26:34 +00:00

Flaky: commands.spec "skip a comment tree", the comment read by being seen #259

littlefox referenced this pull request 2026-10-22 14:11:05 +00:00

Release 0.6.0 #281

littlefox referenced this pull request from a commit 2026-10-22 14:11:05 +00:00

Release 0.6.0

Sign in to join this conversation.

Reviewers

No reviewers

Labels

Clear labels   

area/ci

Forgejo Actions, runner, tests

  

area/desktop

Desktop build

  

area/packaging

Arch, Debian, Fedora, NixOS packages

  

area/protocol

kom-proto, wire format, charset

  

area/reading

Unread tracking, reading order, review

  

area/ui

Dioxus app, layout, interaction

  

area/web

Web build and relay

  

area/writing

Composing and sending texts

  

kind/bug

Something is broken

  

kind/feature

New functionality

  

kind/infra

Tooling, tests, CI, dev environment

  

kind/request

requests for features or changes from users

  

needs/user

Waiting for a decision or action from littlefox

  

parked

Kept for reference; no work planned until the user says so

  

perf

Run the perf workflow and its budgets on this pull request (#287)

  

security

Security-relevant

  

upstream

A proposal for Protocol A or lyskomd, for the LysKOM development group

No labels

area/ci

area/desktop

area/packaging

area/protocol

area/reading

area/ui

area/web

area/writing

kind/bug

kind/feature

kind/infra

kind/request

needs/user

parked

perf

security

upstream

Milestone

Clear milestone

No items

No milestone

Projects

Clear projects

No items

No project

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

littlefox/neokom!255

No description provided.