W3C home > Mailing lists > Public > semantic-web@w3.org > April 2010

Re: next steps

From: Jiří Procházka <ojirio@gmail.com>
Date: Sun, 11 Apr 2010 00:46:09 +0200
Message-ID: <4BC0FFB1.6010204@gmail.com>
To: Danny Ayers <danny.ayers@gmail.com>
CC: Semantic Web <semantic-web@w3.org>
Hi,

On 04/11/2010 12:10 AM, Danny Ayers wrote:
> I believe there is going to be a face to face discussion on what
> happens next to RDF etc.
> 
> I can't afford the travel, but I hope I can make certain points when
> it comes to me addresses the SWIG (that is you).
> 
> Some of the things happening right now seem to be losing the good
> bits. The ability to to blah about blah.
> 
> Things like Facebook have engaged  half the Western world. Twitter works too.
> 
> The RDF model is a nice easy one, and it can cover most things in my
> life (not every weird little bit, obviously).
> 
> Do we have to throw that away for a bit?
> 
> People have been much more productive with other bits.
> 
> Maybe we should just kill RDF for 10 years. Return when the
> correspondences have been noted.
> 
> Personally I think the stuff like Web of Data will happen anyway, it's
> just a matter of how fast.

I agree, question is how... Many people say now is similar to the time
before it was clear HTML will dominate the web page description. While
we are not talking about web page description, but something much bigger
- formal model for description of anything...

I think most of us would rather skip this phase and work with the
winner. Not many of us have in depth understanding of RDF semantics, but
have still decided to go with RDF, maybe because there was no other
alternative, or maybe we were sold it by good evangelism and W3C
backing. The point is this doesn't matter - lot of work has gone into
things based on RDF already and doing changes is going to be costly. I
am for literally setting RDF in stone, not changing it in any way (there
is no versioning mechanism anyhow). Maybe compatible subset of RDF would
be worth something, for faster implementation, or maybe some
specifications closer to hardware, but I think all what is necessary now
is done by general public - a) software infrastructure b) standards wise
I think all we need is more efforts like for example recent Linked Data
Patterns book [1].

Best,
Jiri Prochazka

[1] http://patterns.dataincubator.org/book/

> Love,
> Danny.
> 
> 
> 


Received on Saturday, 10 April 2010 22:46:45 UTC

This archive was generated by hypermail 2.4.0 : Tuesday, 5 July 2022 08:45:17 UTC