Record the answers to entries 8 and 9 where the work reads them [#7]

[iderex]

What was wrong

The Done-when on #7 asks that every one of the nine entries carry an answer
written down where the work that depends on it can read it, and that an entry
whose answer belongs in a decision record point at the record carrying it. All
nine have answers. Two of them, entries 8 and 9, were answered on 2026-08-24 and
were written only on the tracker. I measured that before writing:

for pat in imprint funding 'provider identification' 'contact alias'; do
  printf '%-24s ' "$pat"; git grep -l -iE "$pat" -- decisions/ docs/ content/ data/ | tr '\n' ' '; echo
done
imprint
funding
provider identification  content/legal.txt
contact alias

Run 2026-09-05 at 2add524. The one hit is the legal notice's prose naming the
duty, and no file carried the answer to it.

The legal notice already reads a file whose every value points at the entry:

jq -r 'to_entries[] | "\(.key)  \(.value.state)  \(.value.waiting)"' data/publisher.json
publisher  undecided  entry 8 of issue 7
contact  undecided  entry 8 of issue 7
postal  undecided  entry 8 of issue 7

Run 2026-09-05 at 2add524. So the work depending on entry 8 is built, reads a
pointer at the tracker, and the tree held nothing it could point at instead.

What this does

Two records, in the shape decisions/README.md and 0000-template.md fix, with
the next two free numbers.

0017-what-the-site-says-about-who-publishes-it.md carries entry 8: full
provider identification through a paid imprint-address service rather than a
home address, and a rotatable contact alias on the legal page, with the
alternatives and what each would have cost as the entry laid them out. It
carries the readings the entry was taken against, and the state of the values
file today, under ## What was measured.

0018-whether-the-site-asks-for-money-and-in-what-form.md carries entry 9: a
plain text link to the existing funding providers, with the embedded button left
refused by the byte budget. Its measurement re-reads the organisation funding
file, which has moved and gained a provider since the entry read it, and pastes
the provider keys with the account values cut.

Each names the other under ## What this record does not decide, because the
two were answered on one day and the donation link is the point where they
touch.

Neither record decides anything. Both state answers already taken, and where a
record and the entry differ, the entry on #7 is the authority and the record is
wrong.

Why now, against the reading of 2026-08-26

That reading on #7 left these two unwritten on purpose: both belong to pages
rather than to the build, and both name work outside this repository before a
page can state anything. I have taken the other side of that, for two reasons
that are the directory's own. Its README says a record exists before the code
that assumes it, and the legal notice is code that will assume entry 8 the day
the values arrive. And a record states the decision, not the value: 0017 names
no address and no alias, so the booking finishing changes the values file and
supersedes nothing here. What the page should say in the meantime about an
answer that is decided and a value that is pending is a page question, and it
is now #236 rather than a clause of either record.

Closes

Closes nothing. #7 closes on its own Done-when, and after this change every one
of the nine entries has its answer in the tree or in a record another repository
holds. What is left on #7 is written there already: the reading of whether the
second half of its condition reaches closed issues, and the label. I have not
restated either.

What was run

go run . ci
gate: 7 legs, in order: format, vet, test, build, links, sitemap, invariants
  needs-network was not asked for. Asking costs a request to the public name from whatever machine runs it, and a verdict that moves when somebody else's service does rather than when this tree changes. Ask with: go run ./harness/needs-network
  format: ok, 74 file(s)
  vet: ok
  test: ok, 37 test file(s)
  build: ok, 22 file(s)
  links: every reference that stays inside this site resolves to a file the build wrote
  sitemap: every page the build wrote is listed once, and every entry has a page behind it
  invariants: ok, 39 rule(s) decided, 2 owed and not decided
7 of 7 legs ran. None was skipped.

go run . hygiene origin/main HEAD
hygiene: 1 non-merge commit(s) in origin/main..HEAD, origin internal
  f0fa7bf37c58: subject carries its reference
1 commit(s) judged, none refused.

git diff --name-only origin/main...HEAD
decisions/0017-what-the-site-says-about-who-publishes-it.md
decisions/0018-whether-the-site-asks-for-money-and-in-what-form.md

All run 2026-09-05 at f0fa7bf.

The formatter that judges markdown on a pull request here reaches further than
the gate does, so I ran it at the pinned version rather than assuming it:

npx --yes "prettier@$(jq -er '.[] | select(.id == "prettier") | .version' pins.json)" --check --end-of-line auto "decisions/*.md"
Checking formatting...
All matched files use Prettier code style!

The needs-network set was not asked for and no result from it is claimed here.
Nothing was skipped for a display or for elevation.

No guard is added by this change, so there is none to show biting. What it adds
is prose, and the README beside it says plainly that nothing reads these
headings and no check refuses a record that omits one. That is the residual and
this sentence is the disclosure, not a softening of it.

The means

Markdown in decisions/, which is where this repository already puts a decision
of record, in the shape the template carries and the README states. A record
written anywhere else would be a second home for the same class of thing. It
adds no language, no runtime and no dependency.

Who read it

Nobody but me. This board requires no approving review and no status check on
the default branch, so this pull request is merged by whoever opened it, and
there is no second reader on it tonight. The evidence above stands in place of
one.