Skip to content

A parent is never due before its children - #1626

Merged
Rafael-SOWNet merged 3 commits into
masterfrom
a-parent-is-never-due-before-its-children
Sep 30, 2026
Merged

Rafael-SOWNet merged 3 commits into
masterfrom
a-parent-is-never-due-before-its-children

Conversation

@Rafael-SOWNet

@Rafael-SOWNet Rafael-SOWNet commented Sep 30, 2026 •

Copy link
Copy Markdown
Member

The two rules from #718 (comments 5910315961 and 5910345918), written into AGENTS.md's milestone section:

  • A parent issue's milestone is never earlier than any child's. The order is the versions, then Future, then Guiding principles. A child that moves later takes its parent with it. The other way round, as Happypig375 added an hour later (5910745181), children being worked on now go to the next minor and pull their parent forward. Goal: be able to evaluate math competition/textbook/past papers #718 itself is on Guiding principles.
  • "Under" is the sub-issue relationship, not prose. An issue is linked in the turn it is filed, and a comment says "under" only once the link exists. The command is given beside the one that sets an issue's type, because it takes the child's id rather than its number.

Audited 2026-09-30: every open parent (#233, #718, #1469, #1501, #1540) now holds.

🤖 Generated with Claude Code

https://claude.ai/code/session_012sonx8iAspMiwRwokT1Ura

The two rules Happypig375 set on #718: a parent issue's milestone is never
earlier than any child's, in the order the versions run and then Future and
Guiding principles, so a child moved later takes its parent with it; and
"under" is GitHub's sub-issue relationship, set in the turn an issue is
filed, before any comment says it. With the command that sets the link,
which takes the child's id rather than its number.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012sonx8iAspMiwRwokT1Ura
Happypig375's answer on #718 an hour later: the in-progress children move
to the next minor and take their parent with them, and #718 is on Guiding
principles. The example of #233 going to 2.8 described a state that no
longer holds, so the paragraph says the rule both ways instead.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012sonx8iAspMiwRwokT1Ura
@Happypig375

Copy link
Copy Markdown
Member

It's really just scheduling. We need calculus to be completed earlier because it's the door to university level mathematics and beyond.

Happypig375 on #1626: "It's really just scheduling. We need calculus to be
completed earlier because it's the door to university level mathematics
and beyond." The previous revision read his move of the integration work
to 2.6 as a second direction of the rule. It is not: a child delayed
beyond its parent moves the parent, and bringing work earlier is a
decision with its own reason, which the paragraph now gives as his.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_012sonx8iAspMiwRwokT1Ura
@Rafael-SOWNet

Rafael-SOWNet commented Sep 30, 2026 •

Copy link
Copy Markdown
Member Author

Thanks, I had read the move as a second direction of the rule. It now runs one way only: a child delayed beyond its parent moves the parent. Bringing the integration work to 2.6.0 is recorded as your scheduling decision, calculus first, with your reason (d6c10b6).

@Rafael-SOWNet
Rafael-SOWNet merged commit 3215618 into master Sep 30, 2026
27 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants