Re: [foaf-protocols] semantic pingback improvement request for foaf

2010/4/17 Story Henry <henry.story@bblfish.net>

>
> On 17 Apr 2010, at 11:34, Melvin Carvalho wrote:
>
> >>
> >>  0. Search engine solution
> >>  -------------------------
> >>
> >>  Wait for a search engine to index the web, then ask the search engine
> >> which people are linking to you.
> >>
> >> Problems:
> >>
> >>  - This will tend to be a bit slow, as a search engine optimised to
> search
> >> the whole web will need to be notified first, even if this is only of
> minor
> >> interest to them
> >>  - It makes the search engine a core part of the communication between
> two
> >> individuals, taking on the role of the central database in closed social
> >> networks
> >>  - It will not work when people deploy foaf+ssl profiles, where they
> >> access control who can see their friends. Search engines will not have
> >> access to that information, and so will not be able to index it.
> >>
> >
> > A great summary, Henry
> >
> > What about using W3C recommended standard of SPARQL (Update)?  I refer to
> > the architecture sketch for web 3.0:
> >
> > http://www.w3.org/DesignIssues/diagrams/social/acl-arch.png
> >
> > It strikes me a (hyper) data file *should* be, first and foremost,
> updated
> > (write) using SPARUL or WebDAV and HTTP should be used for read
> operations.
>
> SPARUL seems to me to be asking a lot of technology for something that is
> really simple, and that is much easier done using much more widely deployed
> technology. This does not stop it from being deployed later. But we will
> have a lot
> more chance integrating foaf into web 2.0 applications if we don't require
> SPARUL,
> especially if it is clear that one can do without it.
>

Makes sense.  But if you look at the diagram it did have an extra arrow for
the web 2.0 servers.  As you did raise 6 different ways, I thought I would
add the standards compliant web 3.0 way.  If you're saying we should
concentrate on the low hanging fruit first, it's a great point, but let's
not forget that recommended standards do exist here.


>
> It is important to have this work in browsers too, so that people can
> create
> a friend request and have a web page, they can see the request at. This
> would
> allow them to then also DELETE that ping request, or edit it.
>
> This can all easily be done using POST, GET and DELETE.  And mostly even
> just POST.
>
> Furthermore as we saw doing updates on graphs is still very new, and can
> easily lead
> to all kinds of problems.
>

Right, so perhaps need some testing / proof-of-concept here ... let's
experimant a bit and see what we find.


>
> Finally a ping is arguably not a request to update a graph. It is an action
> to notify someone of something. That is all. As I mentioned in another
> thread, what the
> person notified does with this notification is up to them: it could be
>  - to add the person to their foaf as a friend
>  - to add them as an aquaintance
>  -                a spammer
>  -                a follower
>  -             ignore them
>  - to warn their friends
>  - to call the police
>

Good point.  One classic 2.0 pattern is twitter emailing you 'X has become
your follower' to represent <#X> siot:follows <#you>.  Of course blogs and
we're back to the classic pingback.

I wonder if something like a Talis Changeset could be sent to a
'notification queue' of a recipient ... that may be something like an inbox
for them, but instead containing semantically rich updates.  When the user
and the queue are both online the updates can be process via some kind of
workflow.  Some updates can be automatic.


>
> I really don't see that I want to give other people any decision in
> updating my graphs.
> I can give them a space to make a comment, but updating my graph will be
> something I
> am going to be very careful about allowing.
>
> >
> > So I add you as a friend to my FOAF.  And also send you a message with
> the
> > triples I'd like you to add to your FOAF (use discovery if necessary to
> find
> > your sparql server, but as tabulator does, you could just send it to the
> web
> > 3.0 enabled file).
>
> As stated above, if that is what you want to do, then you don't need
> SPARUL.
> You could post a request which contains the triples that you want.
>
> Perhaps we can design the ping in such a way, that a change request can be
> posted,
> for occasions when you noticed an error in my foaf file....
>

Right, you can do it that way too, but web 3.0 servers will be more and more
often sparql enabled.


>
>
> >  You can peform some sanitization/validation/reputation
> > check before adding the friend to your FOAF.  It's a simple workflow to
> get
> > you started, but you can build layers on top (push/pull, realtime,
> > notifications, approvals etc.).  Also compatible with FOAF+SSL Auth.
>
> I'd be for that, as long as we can start with a very simple ping mechanism
> that
> is easy to implement. And that would favor a POST, using an html form.
>
> Also it would be nice if this could be done in an intutive way so that we
> can have
> deployments with human readable html, that reminds people of SN friend
> requests.
>

I'll thing about some proofs of concept.


>
> Henry
>
>
>
>
>