Re: Background information onesocialweb

On Fri, Feb 5, 2010 at 10:19 AM, Weisscher, Alard, VF-NL <
Alard.Weisscher@vodafone.com> wrote:

> > On 5 Feb 2010, at 02:27, Weisscher, Alard, VF-NL wrote:
>
> > > 2. Do you have any reference to the ACL language that you mentioned
> during the call?
> > >
> > > You can see the general picture of it in the XMPP protocol proposal:
> http://onesocialweb.org/spec/1.0/xep-osw-activities.html#sect-id2774754 We
> expect more discussions on this at the XMPP summit and plan to develop this
> further before our March release.
>
> > General question....did you look at existing ACL/Permission languages to
> use instead?
>
> > (Does the web need yet another access control language?)
>
> > Cheers...  Renato Iannella NICTA
>
> Hi Renato,
>
> If possible we will use existing alternatives whenever we can. We did look
> at a few alternatives, but found them often too complicated. We however
> expect to have more discussions on this topic and are open to any good acl
> candidates. Any suggestions are welcome.
>
> Cheers, Alard
>

You might want to go in another direction altogether and use formal code
instead of ACLs. There's a talk by Zed Shaw that makes this point fairly
clearly [1] but it's part of a 70-min rambling session of his (all very
interesting but not very focused). The beef of the ACL part starts at 5:56
and ends around 17:00

Anyway, I think ACLs on the Social Web are just a recipe for trouble.
They'll be way too complicated and multi-layered to be really a) airtight or
b) understandable by users & developers.

[1] http://vimeo.com/2723800

Cheers,
Tim