Repository navigation
Conversation
This was referenced Sep 21, 2026
This was referenced Oct 19, 2026
…hase 3)
Owner ruling (2026-10-08): "{now} should be per file, {start} the
capture's start — that's the behavior I had intuited from both
existing, so if they're the same today by design that's even more
confusing than I expected." `resolve_run_path` picked one `Local::now()`
per run and `LogWriter` reused that `base` for every split; the panel's
preview only re-asked the host when a template, the project, the
logger's name, or the capture's start changed. So `{now}` read as a
timestamp frozen at the run's start, identical to `{start}` whenever a
capture had no wall-clock anchor of its own.
`LogWriter` now opens through a `FileNamer` closure — called once to
open the run and again at every roll — instead of holding a fixed
`base`. The production namer (`make_run_namer`) resolves the folder
once, rooted and read before the run starts (nothing in a folder
template has reason to differ file to file), closes over the capture's
wall-clock start once, and resolves the *file* template fresh against
`Local::now()` on every call. `{start}` therefore stays the capture's
start across every split; `{now}` names each file's own creation. The
collision rule is unchanged and is now the only reason a suffix
appears: a resolved name already on disk — because the template has no
`{now}` at all, or a `{now}` pattern repeated within its own precision
— takes the next `-002`, `-003`… the same way at the run's first file
and at every later split (`first_free_part`/`split_path`, now always
scanned from part 1 since each open supplies its own fresh base rather
than a running part counter).
The logger panel's preview ticks once a second while mounted and a
template holds `{now}` — the interval `DateTimePatternEditor` already
runs for the `date_time_pattern` setting — skipped entirely when
neither template carries the token, so an unaffected logger issues no
extra IPC.
Docs: `export_template.rs`'s `{now}` doc, `logger.rs`'s module doc and
the `LogWriter` / `FileNamer` / `make_run_namer` rustdoc, README's
logger section (per-file `{now}`, the unchanged `{start}`, the
collision rule, the preview tick). No ADR states the logger
run-naming rule — ADR 0062 covers only the date-pattern notation, not
when a logger evaluates it — so none is amended.
Verified: `cargo test -p cannet-gui` (1449 passed, two new — a roll
resolves the file template from its own fresh instant; `{start}` stays
fixed across a roll while `{now}` advances), `cargo clippy -p
cannet-gui --all-targets -- -D warnings`, `cargo fmt --all -- --check`,
rustdoc `-D warnings`; `pnpm --dir apps/gui test` (3875 passed, three
new on the preview tick), `pnpm --dir apps/gui build`; `pnpm --dir
apps/gui tauri build --no-bundle` (release host, runs). Python lanes
skipped — the diff cannot reach them.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
The owner's comments on section 1 resolve four of five items: the ADR 0050 undo-withdrawal sentence stands (168), the calendar fraction stays as specced (167), a failed manual send gets a system-log row plus an indicator on the transmit element (160, now section 2), and ruff 0.16's default rule set is adopted with the code fixed to satisfy it (136, now section 2, launching). The rounding question is re-put in plain words. Section 2 gains the exit-cue item in groomed form and the 166-168 entries read by PR number. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
…t fix landed, rounding and no-noqa fixes ordered Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
…26-10-20) Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
…e queue Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
…pe (0169, queue) Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
…-10-20) Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
… observation Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
…updated Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
…owner ruling 2026-10-20) Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
…updated Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
…from it Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
…updated Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
…ask 171; task 133 item 2 Task 171 captures the 2026-10-26 four-day Kvaser session's signal-panel hang: the first error frame armed the trace's default error-frame collapse, whose non-narrowable predicate built a filter index over the whole capture while holding the DBC mutex, so every decoding view blocked and Disconnect did not cancel it. Timeline, mechanism with code citations, scope, open questions and exit criteria; placed first in the 0.10.x section, ungroomed. Task 133 gains item 2: the window's own scrollbars flicker on and off. Dockview lays its grid out at Math.round of a fractional, unclipped dock area, and each scrollbar changes the other axis by half a CSS pixel at 200 %, so the two chase each other one resize-observer tick at a time. Fix shape: the app shell never scrolls at the document level. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Five tasks landed today on top of doc-closeout (147, 148, 142, 146 on
new branches; 145 inserted below the python client branches). This
folds what their phases and the overseer's reviews turned up into the
owner review queue: four plot rulings to confirm or reverse (tile
layering under the series, the
Points: autolane exemption, the~1.5-column held-code boundary, the wide ΔH clip), the fzf port's
diacritic boundary, the per-service token gate and grpcio-tools pin as
FYI, the five tasks awaiting acceptance, and the close-out housekeeping
owed: the ev-zonal layout autosave to re-apply, the stale sidecar and
client lockfiles to re-lock on
task136-core-bus, the roadmapretirements, and the tip perf report's disposition.
Co-Authored-By: Claude Fable 5.1 noreply@anthropic.com