Re: 'services' and LDP (was: Interaction model vs data model)

hello roger.

On 2013-01-24 12:22 , "Roger Menday" <roger.menday@uk.fujitsu.com> wrote:
>Thanks for mentioning these two categories Arnaud.
>It does seem that most people here are looking at category 1 type
>applications.

and i think that's what the charter is telling us to do. are you reading
http://www.w3.org/2012/ldp/charter that we're chartered to do specific
protocols?

>As covered in the conversation on Monday call, some of us are interested
>in category 2. we would like to have an issue so that we can monitor the
>parts needed for category 2, track what we won't cover in LDP, etc.

i think it's a good idea to keep this in mind, and maybe we could even, if
there is enough interest and somebody driving it, add non-nromative parts
to the spec that would explain how more specialized services and clients
can be based on LDP.

like i said earlier, i think that by virtue of RDF's openness, we're
pretty well covered when it comes to the openness and extensibility of the
data model. that's almost out of the box with RDF. the greater challenge
may be the interaction model, because then we probably need to come up
with patterns how servers can expose additional affordances in a coherent
way (basically, an equivalent of atom's <limk rel=""/> mechanism), so that
LDP-based protocols look and feel the same way, and generic clients could
still make extensions at least visible ("i see there's an additional link
made available by the server that has the relation type foo, but i have no
idea what it means and how i am supposed to interact with it.").

cheers,

dret.

Received on Thursday, 24 January 2013 12:40:28 UTC