fix(package): Linux Tauri zip: drop move-to-aw-modules.sh and the loose awatcher, rewrite README

[ErikBjare]

The Linux Tauri AppImage should be self-contained and run from anywhere. #1499 bundles awatcher + aw-sync inside the AppImage, .deb and .rpm, so nothing needs to sit next to it. This tidies the zip to match:

• Drop move-to-aw-modules.sh. It copied everything into ~/aw-modules as a workaround for the AppImage having no modules, and only included awatcher if the shell it ran from had XDG_SESSION_TYPE=wayland.
• Rewrite README.txt. The AppImage is self-contained, and the .deb/.rpm are the system-wide alternative. The other folders (aw-watcher-input, the classic window/AFK watchers; aw-notify is bundled by build(linux): bundle aw-notify into the Linux Tauri deb/rpm/AppImage #1510) are optional extras to copy into ~/.local/share/activitywatch/aw-tauri/modules/, which is a default discovery path.
• Leave the loose awatcher/ out of the zip. It's bundled inside the AppImage/deb/rpm by build(linux): stage awatcher + aw-sync into aw-tauri before Tauri bundle step #1499, and awatcher/ isn't a name module discovery descends into anyway.

Goes with #1499; the README describes the bundle as it is once #1499 lands. Earlier revisions of this PR (and of ActivityWatch/aw-tauri#283) also made aw-tauri search the folder next to the AppImage. Dropped: the AppImage shouldn't depend on its surroundings.

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-07T13:27:15.346129Z	20b3b7e	New commits

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[ErikBjare]

@greptileai review

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 8f372abc70

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[greptile-apps]

Confidence Score: 5/5

[Low risk] Updates packaging scripts and documentation for Linux builds.

This PR appears safe to merge; no actionable defect was established.

What we checked:

• Classic packages keep their watcher: The script removes awatcher/ only when the platform is Linux and TAURI_BUILD is true.

Summary

Updates the Linux Tauri zip to use self-contained bundles instead of a module-copy helper.

• Removes move-to-aw-modules.sh and the loose awatcher/ copy.
• Rewrites README.txt with bundle and optional-module instructions.
• ErikBjare explicitly describes bundling from build(linux): stage awatcher + aw-sync into aw-tauri before Tauri bundle step #1499 and build(linux): bundle aw-notify into the Linux Tauri deb/rpm/AppImage #1510 as intended dependencies.

_{Reviews (3) · Last reviewed commit: "fix(package): Linux Tauri zip: drop move..." · Reviewed by Greptile}

[TimeToBuildBob]

🤖 AI code review

This PR removes the move-to-aw-modules.sh helper script from the Linux Tauri packaging, rewrites the README.txt to describe the AppImage/deb/rpm as self-contained with optional extra modules, and changes package-all.sh to copy only the README into the zip while deleting the loose awatcher directory before zipping.

Not safe to merge — 1 P1 open

Confidence 3/5

2 findings · ❌ 1 P1 · ⚠️ 1 P2

❌ P1 high — scripts/package/README.txt:15

The README now instructs users to copy optional module folders into ~/.local/share/activitywatch/aw-tauri/modules/ (or $XDG_DATA_HOME/activitywatch/aw-tauri/modules/), but the PR description says this is a 'default discovery path' for aw-tauri. The diff does not include any change to aw-tauri's module discovery code, and the removed move-to-aw-modules.sh used ~/aw-modules/ as the discovery directory. If aw-tauri's actual discovery path is still ~/aw-modules/ (as the old script and README stated), then the new README points users to a directory that the application does not scan, so optional modules copied there will never be loaded. The observable consequence is that users following the README to add aw-watcher-input or the classic watchers will see no effect after restarting ActivityWatch. This is a documentation/contract mismatch: the README asserts a discovery path that the code it documents does not implement in this repository. The fix is to either update aw-tauri's discovery path to match the README or revert the README to the actual path.

_{How this was verified: Checked the README lines 13-18 and the deleted move-to-aw-modules.sh which used ~/aw-modules. The PR description says the AppImage is self-contained and should not depend on its surroundings, but the README instructs copying to a data directory. Without the actual aw-tauri discovery code, the path may be wrong.}

⚠️ P2 medium — scripts/package/package-all.sh:80

The package-all.sh change removes dist/activitywatch/awatcher with rm -rf before zipping, but the PR description says awatcher is 'bundled inside the AppImage/deb/rpm (staged into aw-tauri's modules/ by the Makefile)'. If the Makefile staging happens after package-all.sh runs, or if the AppImage/deb/rpm are built in a separate step that does not include the loose awatcher, then deleting the loose copy could remove the only copy of awatcher from the distribution. The observable consequence is that the zip (and possibly the AppImage if it depends on the loose directory) would lack the AFK watcher, breaking the self-contained claim. The diff does not show the Makefile staging order, so this is a potential ordering defect: the rm -rf assumes the bundled copy already exists in the AppImage/deb/rpm, but if that staging is not guaranteed to have happened, the deletion is premature. The fix is to verify the staging order or move the rm -rf to after the AppImage/deb/rpm are built.

_{How this was verified: The diff shows only the removal of the loose awatcher; no Makefile changes are present. The PR description claims the Makefile stages it, but that staging is not in this diff, so the ordering is unverified.}

1 advisory finding (summary-only, not scored)

These P2 guard, heuristic, trade-off, or documentation claims are retained for judgment without opening review threads.

⚠️ P2 medium — scripts/package/README.txt

This is a fix(...) PR but no test files are included in the diff. Erik's feedback: 'where is the repro & fixes they are supposed to catch' (gptme#3441), 'that measurement should come with a regression test' (gptme#3446). Add a test that would have caught this bug. (Advisory: Erik merged all such PRs but consistently requested tests.)

Add a test file that reproduces the bug before the fix and passes after it.

_{How this was verified: static preflight: fix-commit + touched-files scan (rule 7)}

Files changed (3) — the diff as I read it

• scripts/package/README.txt — Rewrites the README to describe the Tauri AppImage/deb/rpm as self-contained, with optional extra modules copied into ~/.local/share/activitywatch/aw-tauri/modules/.
• scripts/package/move-to-aw-modules.sh — Deletes the shell script that copied modules to ~/aw-modules/ for aw-tauri discovery.
• scripts/package/package-all.sh — Changes the Tauri Linux packaging block to copy only README.txt and remove dist/activitywatch/awatcher before zipping.

Previous review passes

commit	score	findings	engine	when
8f372abc70f9	5/5	0	llm	2026-10-07 10:17 UTC
7030d487ba99	4/5	1	llm	2026-10-07 12:52 UTC

_{Reviewed 20b3b7e7b47b · openrouter/deepseek/deepseek-v4-flash-0731 · llm engine · 34s · about this reviewer}

_{Maintainer commands}

_{@TimeToBuildBob review (own line) — fresh review · @TimeToBuildBob fix — a worker acts on the findings. Once per comment; 👀 = received.}

[ErikBjare]

@greptileai review

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: dd66db79d8

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[ErikBjare]

@TimeToBuildBob review

[TimeToBuildBob]

The review from 10:19 UTC is current — PR head is still 7030d487ba99, same SHA reviewed earlier. Score 5/5, no findings. Nothing new to review.

[ErikBjare]

@greptileai review
@TimeToBuildBob review

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 20b3b7e7b4

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".