Skip to content

Date two overtaken readings in the decisions that carry them #137

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
iderex merged 2 commits into from
Aug 23, 2026
Merged

Date two overtaken readings in the decisions that carry them #137

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
4f1162c
Correct the hand step the publication stopped being (#130)
iderex Aug 23, 2026
3cbb25d
Date the address grep and put the reading that holds beside it (#136)
iderex Aug 23, 2026
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
29 changes: 23 additions & 6 deletions decisions/manifest-address.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -93,16 +93,32 @@ published nothing and the declaration switched off for the first publication are
`decisions/first-release.md`, and they are absent by decision rather than by
failure.

Two tracked files printed the address, and #35 removed both:
Two tracked files printed the address, and #35 removed both. What follows is the
reading taken that day rather than the state of the tree:

grep -rno 'https://[a-zA-Z0-9./_-]*manifest.json' -- README.md docs/ ; echo "exit=$?"
exit=1

Run 2026-08-08. Nothing was printed and the grep exited 1. What the
operator-facing files said instead was that installation is not available yet,
which was what an address answering 404 left them able to say. Now that the entry
exists they may print the address, and whether they do is the instruction in #55
rather than this rule.
which was what an address answering 404 left them able to say.

The same command answers with five sites now, and those five are the operator
instruction:

grep -rno 'https://[a-zA-Z0-9./_-]*manifest.json' -- README.md docs/ ; echo "exit=$?"
README.md:17:https://flowfin.dev/manifest.json
README.md:54:https://flowfin.dev/manifest.json
README.md:63:https://flowfin.dev/manifest.json
README.md:73:https://flowfin.dev/manifest.json
docs/index.html:31:https://flowfin.dev/manifest.json
exit=0

Run 2026-08-23 against the tree at 1f88020. The first is the address in the
three-step instruction, the three below it are inside the commands an operator
runs when the repository looks empty, and the fifth is the same address on the
served page. Printing them is what the entry in `internal/address.Answered` made
permissible, and it is written rather than still to be decided.

The 404 pair stays here because this is where the address is argued, and a
superseded reading is worth more in place than deleted: it is the pair that
Expand Down Expand Up @@ -142,8 +158,9 @@ That check returned early for as long as the list was empty, and it says so on
its own output rather than passing quietly. With an entry it makes the request,
so the day the entry landed is the day the check began reading anything at all.

The operator instruction that prints the address is the other half of #34. The
list is no longer what stands in front of it.
The operator instruction that prints the address was the other half of #34. It is
written, the five sites above are it, and the list stopped standing in front of
it on the day the entry landed.

## What this costs

Expand Down
19 changes: 17 additions & 2 deletions decisions/release-procedure.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -52,8 +52,23 @@ it.
Naming these is most of the value, because the hand steps are the ones done
differently the second time.

Running the publication. Nothing schedules it today; #32 is where a schedule and
a run that cannot race itself are decided.
Deciding that a run's verdict calls for a merge, and asking for a run when a
release lands between the scheduled ones. Starting the publication is not the
hand step here. `.github/workflows/publish.yml` declares a daily schedule, a
request trigger, and a concurrency group naming the destination so that two runs
never write one file. What such a run leaves is a verdict and nothing else: its
permissions are read-only, none of its three steps pushes, and its checkout goes
with the runner. So `wrote new bytes` on the default branch is the sentence
saying the published catalogue no longer matches what the declarations come to
and a merge is owed, and reading that, then producing the bytes the request
below carries, is the part that stayed a person's.

The sentence that stood at this entry until this correction is superseded rather
than deleted. It said the publication had no schedule, and it sent a reader to
#32 as the place where a schedule and a run that could not race itself were
still being decided. #32 was answered on 2026-08-20 and the workflow file above
carries both, so what that sentence pointed forward to has arrived and what is
left by hand here is the paragraph above rather than starting the run.

Opening the pull request that carries the placed file, and merging it. This is
the same route as any other change and it has the same gate in front of it.
Expand Down
Loading