CUNY Academic Commons

Boone, apologies. I had all of that information written down and either I wrote it in the wrong place or accidentally deleted it because none of it is showing up now. Not sure what happened, probably user error. I will have to ask Alyson if she can repeat it again and write it down.

Here's what I remember. It was Alyson Vogel. We noticed it today.
She had a group invitation to a group that started with Lehman.
That group also showed up in the Memberships area.
So it was showing up in both the invites and the membership tabs of the Groups area in Profile.
If she clicked on it to accept the invite in the invite area it registered an error that read something like:
group invitation cannot be accepted

She was able to accept other membership requests (I had one to a group I sent her) and as I said she was actually a member of the group.

I've experienced something like this in the past, two -- cases where someone is both in an invited state and a member state for the same group.

Thanks for the details, Chris. It may sorta be related to https://buddypress.trac.wordpress.org/ticket/3969

But in any case, it sounds like it may be a BP bug. The bug is not so much that you can't accept the invite - it makes sense that accepting an invitation would fail when you're already a member of the group in question. The real problem is that these duplicates exist in the first place, and the secondary problem is that, the duplicates being in the system, the invitation is shown to the user.

I'll have a closer look in the upcoming days. In the meantime, if the user is already a member of the group, I assume this is not a pressing issue.

I decided that the best strategy was to allow group memberships to be "accepted" in cases where the user is already a member of the group. I went this way because it was behavior I could actually reproduce, as opposed to the multiple group membership bug, which could be from a much earlier version of BuddyPress.

BP ticket: https://buddypress.trac.wordpress.org/ticket/4182
Manually ported to CAC: https://github.com/castiron/cac/commit/68ff8f6bed06f383495bc03923786e64dcf5b749

Makes sense. thanks Boone.