Answer A-487d: a read of the newest revision names the newest again while it moved on, bounded by progress
Claude-Session: https://claude.ai/code/session_01JG73KHUnBze48rdua8u2sy
This commit is contained in:
@@ -2004,3 +2004,10 @@ the principle that settles it, cited as "design 47 G A-n".
|
||||
all of its events (`meshcatalog.SubjectCatalogRevisionAll`), the one the grants part grants, and its manifest
|
||||
consumes `mesh-catalog.catalog-revision`; where G-12 names `catalog-revision.>`, the contract's subject holds. (From
|
||||
the review of A-487 and A-487a, round 1.)
|
||||
- **A-487d** (A-487, A-487b; principle 7) A read of the newest revision of a branch whose list or one of whose entries
|
||||
is removed under it (the pin judge removes a revision as soon as a newer one is kept) names the newest again and
|
||||
reads that, as long as the newest is newer than the one that failed; the same revision failing twice is the read's
|
||||
failure, said as it is. Progress bounds the loop, so it has no number and no setting. This holds for every read that
|
||||
names "the newest" (the kept trunk's start read and its read on `catalog-revision`, `TrunkView`, `BranchView`); a
|
||||
read of a revision by its number keeps failing loudly when that revision is gone. (From the review of the fix of the
|
||||
live race on 2026-11-19 11:55: trunk.488 removed while the catalog half read its entries.)
|
||||
|
||||
Reference in New Issue
Block a user