You removed the ?, but threadInfos could be missing an entry for threadID, right?

What if timestamps are equal? Should we have some conflict resolution mechanism?

In theory, it can, but in practice, it shouldn't because canBeProcessed would return missing_thread in such a case.

It is unlikely to have a conflict. Also, we have some more likely issues, e.g. https://linear.app/comm/issue/ENG-9125/membership-list-can-become-inconsistent-in-thick-threads. Created a task to track https://linear.app/comm/issue/ENG-9159/thick-threads-timestamps-conflict-resolution but we can't prioritize it before launch.

Where is canBeProcessed used that would stop this?

In https://github.com/CommE2E/comm/blob/master/lib/shared/dm-ops/process-dm-ops.js#L140-L160 we return early if we can't process the operation.

I don't think this is the best practice in general to depend on code run outside of the function, since someone could call this function somewhere else. But in this case iI think t's fine, if you really don't want to change it

Makes sense!

If no changes end up being applied, is there a way we can skip this update, instead of propagating an empty update?

Created https://linear.app/comm/issue/ENG-9232/consider-skipping-creating-updates-when-there-are-no-changes-to-a track.