Re: "DOM" for RDF?

Yes Richard,  i agree and mybe my reply ws not so clear regarding what i
mean. I like the ld-api (such as Elda) which uses ldpath much like xpath is
used in the XSLT or DOM context... the next step in that direction could be
marmotta, i agree on that.

I also love very much Sail or the gremlin language (via tinkerpop [1]),
which can be very helpful, as the api can be seen as part of the query
language, again. But those are server side approach, and they are probably
under the same "optimized by the language" umbrella suggested before by Tim.
Since RDF models domain much more than a DOM can do i feel we have not to
search to go too much on the DOM-like direction, and a good compromise is
by the usage of expressive domain language.

Just my 2 cents


[1] http://www.tinkerpop.com/


2013/12/2 Richard Light <richard@light.demon.co.uk>

>
> On 02/12/2013 11:10, Alfredo Serafini wrote:
>
> Hi Richard
>
>  from my point of view the DOM-like approach does exists yet, and it's by
> SPARQL and LDpath. What are them lacking? Do you feel there should be an
> object-oriented approach? As for the Jena model or Sesame internal Graph
> representation? If this is the case it could be interesting from my point
> of view an approach similar to the current implmentation of the Sail
> interface.
>
>   Alfredo,
>
> I think a query language on its own isn't enough.  I would like to see an
> environment in which I can create an in-memory graph, load it with one or
> more graphs and/or query results, create, delete and edit graph nodes and
> triples within it, query it, and serialize all or part of the result to any
> of the popular serialization formats, plus (X)HTML, ideally using a tool as
> powerful as XSLT. :-)
>
> Looking up LDPath I came across Marmotta [1], which seems rather closer to
> what I have in mind.
>
> Richard
>
> [1] http://marmotta.apache.org/
>
>
>  My 2 cents
>
>  Alfredo Serafini
>
>
> 2013/12/2 Richard Light <richard@light.demon.co.uk>
>
>>  Hi,
>>
>> I'm sure this has been discussed many times and/or ages ago, but I am
>> struck by the absence of a DOM-like W3C framework for RDF. By this, I mean
>> "an application programming interface (API) for [RDF graphs]", which will
>> be "a standard programming interface that can be used in a wide variety of
>> environments and applications. The [RDF] DOM is designed to be used with
>> any programming language". (Quotes taken from [1])
>>
>> A quick search turns up a number of PHP-based libraries, and the odd one
>> for javascript, Delphi, Python and Ruby, but as far as I can see there is
>> little, or no, commonality of approach or functionality amongst these
>> offerings.  This means that a programmer (a) has to decide which of these
>> widely varying approaches to adopt, (b) only gets whatever documentation
>> each chooses to provide and (c) is faced with a complete rewrite, should
>> they decide to switch RDF platform.
>>
>> Might this situation be a significant factor in the slow take-up of RDF
>> by mainstream developers?
>>
>> Richard
>>
>> [1] http://www.w3.org/TR/REC-DOM-Level-1/introduction.html
>>
>> --
>> *Richard Light*
>>
>
>
> --
> *Richard Light*
>