Entry eight of #46 is answered, and this issue writes the answer down where a
decision lives. The answer, decided 2026-08-24: this board takes experiments
from anybody.
That is what a public board usually means, and it is the only way this one
becomes useful to anybody who is not already here. Taking none, and saying so,
would have cost the board its reach, and it would have left the contributing
guide, the templates and the sign-off gate written for an audience that does not
exist - a repository that reads as open while accepting nothing wastes the time
of whoever believes it.
The residual is the part that has to survive into the record. The cost is not
review effort, which is ordinary. It is that three rules in this plan depend on a
person having read them before they start: what may never be committed, what
happens when an experiment finds a flaw in shipped software, and what an
experiment may do with real data. Somebody arriving for the first time is exactly
the person least likely to have read any of them, and the first two are the ones
where a mistake is public and permanent. Nothing here can refuse those
violations, which the issues holding them state plainly, so the residual sits on
whoever reads the change. The record says that, in those terms, rather than
recording an open door and leaving the residual in the issue that is about to be
closed.
The contributing guide carries the answer in its first paragraph, because the
alternative is somebody finding out after doing the work, and the three rules
above are linked from that same paragraph so the person least likely to have read
them meets them where they arrive. CONTRIBUTING.md is on the default branch
already, written under #9, so this is an edit to a landed guide rather than a
blank page:
git ls-tree --name-only origin/main CONTRIBUTING.md
The templates under .github/ISSUE_TEMPLATE/ landed under #48 and follow the
same answer.
Entry three of #46 only becomes a question once there are contributors, and this
answer is what makes it one. That entry is answered separately and its record is
its own issue.
What section three has to list: taking no outside experiments and saying so, and
taking proposals as issues rather than as changes. Section four says what each
would have cost, and entry eight of #46 sets both out - including that the middle
option turns the cheapest thing this board offers, which is trying something,
into a request somebody has to answer.
Done when docs/decisions/0024-who-may-run-an-experiment-here.md exists on the
default branch, carries the four sections record 0000 fixes, says the board
takes experiments from anybody, names the three rules a newcomer must have read
and states that nothing refuses a violation of them, says the answer belongs in
the first paragraph of CONTRIBUTING.md with those three linked from it, and
lists in sections three and four both rejected options with what each would have
cost.
Entry eight of #46 is answered, and this issue writes the answer down where a
decision lives. The answer, decided 2026-08-24: this board takes experiments
from anybody.
That is what a public board usually means, and it is the only way this one
becomes useful to anybody who is not already here. Taking none, and saying so,
would have cost the board its reach, and it would have left the contributing
guide, the templates and the sign-off gate written for an audience that does not
exist - a repository that reads as open while accepting nothing wastes the time
of whoever believes it.
The residual is the part that has to survive into the record. The cost is not
review effort, which is ordinary. It is that three rules in this plan depend on a
person having read them before they start: what may never be committed, what
happens when an experiment finds a flaw in shipped software, and what an
experiment may do with real data. Somebody arriving for the first time is exactly
the person least likely to have read any of them, and the first two are the ones
where a mistake is public and permanent. Nothing here can refuse those
violations, which the issues holding them state plainly, so the residual sits on
whoever reads the change. The record says that, in those terms, rather than
recording an open door and leaving the residual in the issue that is about to be
closed.
The contributing guide carries the answer in its first paragraph, because the
alternative is somebody finding out after doing the work, and the three rules
above are linked from that same paragraph so the person least likely to have read
them meets them where they arrive.
CONTRIBUTING.mdis on the default branchalready, written under #9, so this is an edit to a landed guide rather than a
blank page:
The templates under
.github/ISSUE_TEMPLATE/landed under #48 and follow thesame answer.
Entry three of #46 only becomes a question once there are contributors, and this
answer is what makes it one. That entry is answered separately and its record is
its own issue.
What section three has to list: taking no outside experiments and saying so, and
taking proposals as issues rather than as changes. Section four says what each
would have cost, and entry eight of #46 sets both out - including that the middle
option turns the cheapest thing this board offers, which is trying something,
into a request somebody has to answer.
Done when
docs/decisions/0024-who-may-run-an-experiment-here.mdexists on thedefault branch, carries the four sections record
0000fixes, says the boardtakes experiments from anybody, names the three rules a newcomer must have read
and states that nothing refuses a violation of them, says the answer belongs in
the first paragraph of
CONTRIBUTING.mdwith those three linked from it, andlists in sections three and four both rejected options with what each would have
cost.