Re: optimizing container pages serialization to enable streaming

On 11/11/2013 01:25 PM, Wilde, Erik wrote:
> On 2013-11-11, 10:18 , "Alexandre Bertails" <bertails@w3.org> wrote:
>> On 11/11/2013 12:48 PM, Eric Prud'hommeaux wrote:
>>> ). While strictly true, the issue here is not monotonicity but
>>> pragmatic ordered serialization to enable a client to interpret
>>> the data as it arrives. Let's call that "streaming".
>> Why am I thinking that specifying an order on the streamed RDF triples
>> does not feel like RDF anymore?
>
> is it really the case that LDP is attempting to put constraints on RDF
> beyond the RDF model and serializations? if yes, i'd like to point out
> that this might not be such a great idea. this would disallow the use to
> standard RDF serializers that (rightfully so) do not allow to control
> these aspects that RDF does not constrain in any way.

That was pointed out during this morning teleconf.

> it reminds me a little bit of XML's early years when out of convenience,
> some specs would constrain XML in ways beyond the XML spec (there are many
> way of doing that, for example requiring attributes to be sorted by name,
> or disallowing character entities). this led to *a lot* of breakage that
> was impossible to fix without rebuilding whole software stacks. i may be
> misreading this thread, but if LDP attempts to constrain any RDF syntax in
> any way, i think that would be a Really Bad Idea.

+1. A server would be lying to a client if it returns [[ Content-Type:
text/turtle ]] while adding some extra constraints like paging. Paging
is not a conservative extension for already defined RDF resources and
this cannot simply be fixed with a Link header.

Alexandre.

>
> cheers,
>
> dret.
>
>

Received on Monday, 11 November 2013 18:34:59 UTC