Re: [Auth]Freport Update

[John]

Hans Zandbelt wrote:
> 
> > You've just described the design of the Freport Databank retreival
> > mechanism.
> 
> And a local Profile Manager!. So I guess we all agree that we
> can make a robust secure implementation of a local Profile
> Management system after all.

Don't jump to conclusions Hans, and don't misstate what either of us
wrote. Read in context:

"Won't this be a problem for *any* sort of third-party
convenience-enhancing option?"...
"A small change could fix this: if the client grabbed their profile from
the Profile Manager and sent it to the Profile Requester (instead of the
Profile Requester getting the profile from a Manager directly), then the
Manager wouldn't know which Requester was getting whose Profiles. Plus,
the user still gets all the fun convenience (so long as they can run
their profile client)."

The paragraphs explicitly refer to the "third-party". As you and I both
know, a third-party is only needed in the Remote Manager situation.
Local Manager doesn't enter into the paragraphs. Neither he, nor I, ever
mentioned or eluded to the local manager.

His paragraph referred to the elimination of the LINKAGE which occurs in
the ID-Sec messaging between the Service and the Remote (third party)
Profile Manager.  He suggested severing the linkage between the Service
and the Remote Manager and passing everything through the Owner. He did
NOT suggest the *removal* of the Remote Profile Manager itself (at least
I don't believe he did). Once the linkage is removed and the data
request is forced to be passed from the Service through the Owner to the
Remote, and the resulting response is passed from the Remote to the
Owner and back to the Service; the resulting network connection *IS* the
Databank retrival mechanism of FrePort.

The net result of this change to the messaging topology is that one
piece of the transactional metadata is eliminated from mining by the
Manager Provider. There's more transactional metadata that must be
eliminated to completely clean the topology, but if you use my methods
to eliminate those other bits of incidental data; the result would be
that you'd be developing FrePort and not ID-Sec!

John Le'Brecage.