Follow-up to #361: Auditing models with UUID pk Currently, only the result of calling `to_i` on a UUID string is stored, which gives the illusion of working correctly: ```ruby > "78035962-1234-1234-1234-9c394a3e5a0b".to_i => 78035962 ``` * When the UUID contains hyphens, `to_i` stops at the first alphabet. * If there are no hyphens, the integer may exceed the 4‑byte limit and raise an error. To address this, I propose one of the following: 1. Update the README to explicitly state that UUIDs are not supported. 2. Extend PR #333’s column‑type configuration to include `auditable_id` and `associated_id`, allowing full UUID support. 3. Convert all ID columns to string types so they can store complete UUIDs. PR #333 currently only lets you change `user_id`’s type. Extending support to these other ID columns would enable full UUID compatibility. Related: #651
Follow-up to #361: Auditing models with UUID pk
Currently, only the result of calling
to_ion a UUID string is stored, which gives the illusion of working correctly:to_istops at the first alphabet.To address this, I propose one of the following:
auditable_idandassociated_id, allowing full UUID support.PR #333 currently only lets you change
user_id’s type. Extending support to these other ID columns would enable full UUID compatibility.Related: #651