Say 0.16.21 where the docs still said 0.16.20 #299

New Issue

Closed

opened 2026-09-07 00:10:22 +00:00 by jcoffey-dev · 0 comments

jcoffey-dev commented

2026-09-07 00:10:22 +00:00

Owner

Caught by review: the README badge still read 0.16.20.

The badge said 0.16.20 in both its label and the shield image it links to. It is the first version number a reader sees, and it was the one place the earlier prose update missed, because it is HTML rather than Markdown.

Two entries had gone past stale and were wrong. FEATURES stated that occurrence ids are not stable across a write, and KNOWN-ISSUES carried the same thing as a live hazard, with the five-week series that demonstrated it. 0.16.21 fixed precisely that: an occurrence is identified by its recurrence id, and an id held across a write stays on its own date.

Both entries now say so. The old behaviour and the evidence for it are kept rather than deleted, because the client supports 0.16 as a whole and 0.16.20 is still a server someone may be running, and each records what replaced it and how that was confirmed — the same shape of test, re-run on a real 0.16.21.

The defence in the client stays either way, and the reasoning is written down instead of left implicit: re-resolving by recurrenceId costs one lookup, a date can still leave a series, and the newest release is not the only supported one.

The KNOWN-ISSUES header now says the live instance runs 0.16.21 and, unlike every upgrade before it, that this one was re-run rather than read against the diff, naming what was exercised by hand.

Left alone deliberately: the benchmark figures in the install docs, which record what they were measured against and would become false if edited, and the minimum-version notes that correctly say 0.16.20 or newer.

Merged 2026-09-06 as coffey-labs/ihasmail@0e05bee69a

_{Rebuilt from: git history, session transcript.}

Caught by review: the README badge still read 0.16.20. **The badge** said 0.16.20 in both its label and the shield image it links to. It is the first version number a reader sees, and it was the one place the earlier prose update missed, because it is HTML rather than Markdown. **Two entries had gone past stale and were wrong.** FEATURES stated that occurrence ids are not stable across a write, and KNOWN-ISSUES carried the same thing as a live hazard, with the five-week series that demonstrated it. 0.16.21 fixed precisely that: an occurrence is identified by its recurrence id, and an id held across a write stays on its own date. Both entries now say so. The old behaviour and the evidence for it are kept rather than deleted, because the client supports 0.16 as a whole and 0.16.20 is still a server someone may be running, and each records what replaced it and how that was confirmed — the same shape of test, re-run on a real 0.16.21. The defence in the client stays either way, and the reasoning is written down instead of left implicit: re-resolving by `recurrenceId` costs one lookup, a date can still leave a series, and the newest release is not the only supported one. **The KNOWN-ISSUES header** now says the live instance runs 0.16.21 and, unlike every upgrade before it, that this one was re-run rather than read against the diff, naming what was exercised by hand. Left alone deliberately: the benchmark figures in the install docs, which record what they were measured against and would become false if edited, and the minimum-version notes that correctly say 0.16.20 or newer. **Merged** 2026-09-06 as coffey-labs/ihasmail@0e05bee69a79 <sub>Rebuilt from: git history, session transcript.</sub>

This repo is archived. You cannot comment on issues.

1 Participants

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: coffey-labs/ihasmail-github-archive#299