fix(yes-core): accept any RFC 9562 UUID version, not only v4 (2.1.1) - #37
Merged
Conversation
Same bug as yousty-eventsourcing 17.0.2, in the other gem family. pg_eventstore 3.0.0 moved event-id generation off the database's gen_random_uuid() (always v4) to SecureRandom.uuid_v7, and those ids reach us as causation_id / correlation_id. The v4-pinned pattern makes TransactionDetails.new raise Dry::Struct::Error, which fails the handler and kills the subscription once its restarts run out. apprenticeship_presentation died that way in production on 2026-08-22 through yousty-eventsourcing. company_manager and application_management_system are exposed through this gem and have not been hit yet only because the specific handlers have not fired -- worth fixing before they do, particularly since the failed_subscription_notifier is still unmerged here (#36), so a death in those two services would be SILENT. Version is now [1-8] rather than a hard-coded 4. Still a real constraint: variant nibble stays 8/9/a/b per RFC 9562. Specs cover v4/v7, non-UUID, bad variant and bad version. README gains the pg_eventstore compatibility note. Based on origin/main so it is independent of the open #36 branch.
CI failed at "Set up Ruby" with bundle exit 16 across all five jobs -- a lockfile/gemspec mismatch, not a test failure: bumping only yes-core to 2.1.1 left every committed Gemfile.lock pinning yes-core (2.1.0), which bundler refuses in frozen mode. Following the convention the 2.1.0 release used (b9b418d), which bumped all five version files together and updated the lockfiles: lib/yes, yes-auth, yes-command-api and yes-read-api go to 2.1.1 alongside yes-core, and the yes-* entries in all five Gemfile.lock files are updated to match. Only yes-core has a functional change; the others move for version lockstep.
PR #36 added the failed-subscription notifier without a CHANGELOG entry, and it is now part of this release. Worth pairing explicitly: the UUID fix addresses a bug that KILLS subscriptions, and the notifier is what tells you when one has died -- yes-core services had neither until now.
This repository is public. The 2.1.1 notes and the code comments named internal services, an internal ticket id and a private sibling gem -- none of which mean anything to an outside reader, and none of which we should be publishing. Removed from CHANGELOG, types.rb, railtie.rb and failed_subscription_notifier.rb. The technical substance is unchanged and arguably clearer: what the bug is, what it breaks, and what the constraint still enforces, described in terms of pg_eventstore and this gem's own behaviour. Note CHANGELOG.md:98 (a 2.0.0-era entry) still names the private gem. Left alone rather than rewriting released notes -- worth a separate decision.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Same bug as yousty-eventsourcing 17.0.2, in the other gem family
pg_eventstore 3.0.0 generates event ids with
SecureRandom.uuid_v7(moved off the database'sgen_random_uuid(), which always produced v4). Those ids reach the gem ascausation_id/correlation_id, andTypes::UUIDpinned the version nibble to4, soTransactionDetails.newraisesDry::Struct::Error— the handler fails and the subscription dies once its restarts are exhausted.apprenticeship_presentationdied exactly that way in production on 2026-08-22, via the same bug in yousty-eventsourcing.Why fix it here too, now
company_managerandapplication_management_systemare exposed through this gem. They have not been hit yet only because the specific handlers have not fired — and 🔴 thefailed_subscription_notifieris still unmerged here (#36), so a death in those two services would be silent. No Sentry alert, just a subscription quietly not processing.The change
Version nibble becomes
[1-8]instead of a hard-coded4. Still a real constraint — the variant nibble must remain8/9/a/bper RFC 9562.not-a-uuidSpecs cover all of the above. Version → 2.1.1, root CHANGELOG entry, and a pg_eventstore compatibility note in the yes-core README.
Based on
origin/main, so it is independent of the open #36 branch.