Re: Google Docs as capabilities as data

Re-reading this thread again now, I apologize ... it appears that I read
something into the messages that isn't actually there (or perhaps it is and
I'm not finding it now, I'm not sure if my mind is failing me one, the
other, or both times ...).

At any rate, I thought I had seen someone wondering why the traditional
identity-based ACL model of access to resources is often implemented, and
the "one webkey for everyone" model is implemented, but no one ever (apart
from my friends at HP Labs, of course) implements the "ACLs via multiple
webkey models".

I was attempting to show that the UX for the single webkey model has clear
benefits, but the UX for the identity-based ACL model and the webkey-based
ACL model is roughly comparable. However, *implementing* the webkey-based
model (from scratch) is (probably) more work than either of the other two
and hence why we don't see it all that often.

I was *not* claiming that you couldn't build a UX that included individual
revocation that was understandable for web keys, or that such a UX was
actually worse than the identity-based UX.

Hopefully that helps,

-- Dirk

On Tue, Jan 28, 2014 at 8:41 PM, Stiegler, Marc <[email protected]>wrote:

>  So, I’m not sure if you’re trying to describe what’s involved from a
> user perspective or a developer perspective. From a developer perspective,
> there is a little more work involved to enable individual revokers than a
> single revoker. From the user perspective many of these steps do not exist.
>

_______________________________________________
cap-talk mailing list
[email protected]
http://www.eros-os.org/mailman/listinfo/cap-talk