Why make a mod of this? just submit the bug again in the bug tracker
1. The bug is not a subtle one the developers missed. It is there by design. The developers who wrote the code for tracking unreads purposefully wrote code to wipe out global unread flags whenever the user reads his last unread post of a forum or posts a new post, and the code and comments they wrote make this totally clear. I suspect they made a choice to do that to avoid the additional resource overhead it would have taken to treat globals just like any other forum, since regular boards rarely will have new global posts anyway. So I would not be surprised if they choose not to fix the bug.
2. It is possible (not likely but possible
) that I thought of an approach for fixing the bug that is efficient enough so that they might change their minds and adopt something like my fix. So submitting the full code as a mod and having people test it out could convince the developers to fix a bug they were otherwise going to leave alone (and give them a starting place when they write their own fix). Of course it's also completely possible that my approach sacrifices something the developers don't want to sacrifice, or introduces a new bug that is worse than the original bug. The comments I get during validation should be interesting.
3. I wanted to fix this bug in a validated way so that I could use the fix for another mod of mine that deals with unread flags generally (my view or mark unread posts mod). Some users had reported that my mod didn't work correctly for globals and they were right, but the reason it didn't work correctly on globals was because of this bug.
Has anyone tried this yet on their boards?