Forum OpenACS Q&A: Re: Group composition and permissions hineritance

Collapse
3: Re: Group composition and permissions hineritance (response to 1)
Posted by Dave Bauer on 12/04/03 05:38 PM
So you are saying ( i believe ) that if you grant permission to group three, that groups one and two also should get this permission?

This sounds like a similar issue I worked with on a site with 3 seperate group hierarchies.

The composition_rel works so that members of group three and also members of groups two and one in your example. Also memebers of group two are members of group one.

So if I am not mistaken you want the permissions to inherit in the reverse of how it currently works? That is a permission granted to group one is inherited by groups two and three.

Collapse
4: Re: Group composition and permissions hineritance (response to 3)
Posted by Claudio Pasolini on 12/04/03 06:14 PM
Dave,

it is exactly the contrary: if I grant a privilege, let say write, TO a party ON, let say, group one, I want that the party also get the write privilege on the objects two and three.

Unfortunately the groups don't hinerit their permissions by the parent group, but by the object corresponding to the closest subsite.

Collapse
5: Re: Group composition and permissions hineritance (response to 4)
Posted by C. R. Oldham on 12/04/03 06:22 PM
Claudio,

I think your statement above indicates a possible misunderstanding (maybe on my part 😊 ). Remember a permission is a *triplet* with one element of the triplet being an object for which the privilege would make sense.  Granting "write" to a party on a *group* has undefined semantics.  What would you be trying to achieve with that?