Re: Discovery/Affordances

hello arnaud.

On 2013-06-14 7:52 , "Arnaud Le Hors" <lehors@us.ibm.com> wrote:
>Andy Seaborne <andy@apache.org> wrote on 06/13/2013 11:23:38 AM:
>> concretely then - is httpd (or any other web server supporting
>> GET/PUT/DELETE/HEAD) an LDP server?  If not, where does it fail?
>I think the answer is that it can be.
>I just checked the spec and there is really only one MUST that you'll
>have to fulfill in addition:
>4.2.2 LDPR servers MUST provide a text/turtle representation of the
>requested LDPR [TURTLE
><https://dvcs.w3.org/hg/ldpwg/raw-file/default/ldp.html#bib-TURTLE>].
>With that taken care of the answer to your question would be yes.
>Besides, I'm told that in fact the conformance section allows for LDP
>servers not to support LDPCs. All it takes is for the server not to ever
>serve any LDPC. :-)

it just seems this whole conformance section starts from a strange
foundation. it's like saying "to be a web site server you MUST serve
text/html". it would be looking at everything backwards, instead of just
saying "serving HTML representations is what constitutes a web site
server; if something out there in the universe doesn't provide HTML
representations, the question of whether it is a web site server makes no
sense".

so i think i agree with some other that i find the whole line of thinking
in that section a bit out of line with how things work when you look at
them from the web angle. you can "validate" HTML without a web server
being around, just be looking at HTML from the file system. that may be
different for LDP because it may have more to say about HTTP headers, but
then it's still all about what kind of behavior you expect to see in the
uniform interface, when you as a client follow a link and expect to start
interacting according to LDP.

cheers,

dret.

Received on Friday, 14 June 2013 15:05:41 UTC