Grouper Developers Forum

[Chris Hyzer <>]

If we allow it for memberships, we should allow it for composites.  So for you example below, it stops once a group is a member of itself (right?)

So if A is a composite group with factors unioned B and C.  And D is a member of B.  And A is a member of D.  Then the members of D are everything they normally are (without a circular membership), but stop with D since we aren’t adding members twice for circular memberships…  right?  So D would have all the membership A, B, C etc but not D again.  However composite memberships work…

 

Regards,

Chris

 

From: Shilen Patel [mailto:]
Sent: Monday, September 22, 2008 11:13 AM
To: Grouper Dev
Subject: Re: [grouper-dev] Draft Minutes: Grouper Call 17-Sep-08

 

Just a quick follow up to this....

 

Should we allow a factor of a composite to have a direct or indirect membership to the composite group.  Is there a use case for this?

 

So for instance, say that A is a composite group and has factors B and C.  And D is a member of B.  We can start to get a circular membership path if we allow A to be a member of B or if we allow A to be a member of D.  Should we allow this?  By member, I'm only referring to the default list membership.  

 

Thanks!

 

-- Shilen

 

 

 

On Sep 22, 2008, at 5:48 AM, Emily Eisbruch wrote:

Concerning circular memberships,

(https://bugs.internet2.edu/jira/browse/GRP-168)

TomZ gave an example of the need for “circular membership”: 

an instructor teaches course A and course B and this instructor 

wants all mail sent to A to go to both and B to go to both.  

Chris and Shilen agreed that for such cases — in the interest of less bloating 

to the membership table — group A does not need to be an 

effective or indirect member of group A.