Re: Creating JSON from RDF

On Mon, Dec 14, 2009 at 10:23 AM, Richard Light
<richard@light.demon.co.uk> wrote:
> In message <C74BADC3.20683%t.hammond@nature.com>, "Hammond, Tony"
> <t.hammond@nature.com> writes
>>
>> Normal developers will always want simple.
>
> Surely what normal developers actually want are simple commands whereby data
> can be streamed in, and become available programmatically within their
> chosen development environment, without any further effort on their part?
>
> Personally I don't see how providing a format which is easier for humans to
> read helps to achieve this.  Do normal developers like writing text parsers
> so much?
>
> Give 'em RDF and tell them to develop better toolsets ...

RDF tooling still has some rough edges, it must be said. I am as
enthusiastic about RDF as anyone (having been involved since 1997) but
I've also seen the predictable results where on occasion people (eg.
standards groups) have been 'arm twisted' into using the technology
against their judgement and preferences. We don't have a solid
well-packaged and tested RDF/XML parser for the Ruby language yet, for
example. And while we do have librdfa integration into the
Redland/Raptor C toolkit, it hasn't yet propagated into all the easy
install settings we'll eventually find it - like my Amazon EC2 Ubuntu
box, or the copy of Fink I installed recently on my MacBook Pro. And
in PHP we have a fantastic RDF toolkit in ARC2, but it relies on MySQL
for all complex querying. Plenty of scope for toolkit polish and
improvement, nothing to worry massively about, but also lots of things
that will cause pain if we take a stubborn "RDF or nothing" approach.
I wholeheartedly applaud the pragmatic approach from Jeni and others.

> Come to that, RDF-to-JSON conversion could be a downstream service that
> someone else offers.  You don't have to do it all.

That could be useful for some, and inappropriate for others. Every new
step in the chain introduces potential problems with latency, bugs,
security and so on...

cheers,

Dan

Received on Monday, 14 December 2009 09:43:47 UTC