Skip to content

docs(drift): a remedy tested in the direction that works, parked #619

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
JulienMartel merged 1 commit into from
Sep 21, 2026
Merged

docs(drift): a remedy tested in the direction that works, parked #619

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 22 additions & 0 deletions docs/drift.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -337,3 +337,25 @@ about why. Row 18 shares the skeleton but goes wrong later, when a fix narrows
a limit; this was wrong when written. Row 26's catch is the closest — diff at a
second dimension — except that there the instrument is narrower than the claim,
and here the measurement is sound and the explanation is confounded.

**A REMEDY that works in the direction anybody would TEST it and fails in the
direction anybody would NEED it.** `haus capture` snapshots a preference domain
and `haus revert-settings` restores it, and the README, the installer, the
command's own closing line and five pages of the site all promise it puts your
settings back. It restores through `defaults import`, which merges rather than
replaces, so it can return a key the snapshot HAS and cannot remove one added
since. Measured on a guest running haus `ef6e808d`: `com.apple.dock tilesize`
captured at 48 and changed to 67 comes back to 48, and the same key absent at
capture and written to 67 is still 67 after `✓ com.apple.dock restored` prints.
Anyone checking the verb reaches for a setting they had already chosen, which is
the working half; the half that fails is the settings a rebuild ADDS, and on a
Mac that had no opinion about a domain that is most of them. Thirteen sentences
across two repos carried the unqualified promise when this was measured, one of
them the phrase the source comment used of itself, "byte for byte". **Caught
by** asking of a restore not "does it put my value back" but "what does it do
about a key that was NOT THERE when I captured" — the empty cell in the
before-state, which a fixture written from the happy path never has. Row 27 is
the near neighbour and not the same: row 27's remedy produces nothing at all and
is falsified inside the same rev by the same repo, where this one half works,
passes the obvious test, and nothing in the repo had ever contradicted it. The
tell is a remedy whose verification and whose purpose do not take the same input.
Loading