- From: Reto Bachmann-Gmuer <reto.bachmann@trialox.org>
- Date: Thu, 3 Mar 2011 09:01:03 +0100
- To: nathan@webr3.org
- Cc: Melvin Carvalho <melvincarvalho@gmail.com>, semantic-web@w3.org
Received on Thursday, 3 March 2011 08:01:36 UTC
On Wed, Mar 2, 2011 at 9:23 PM, Nathan <nathan@webr3.org> wrote: > hey mate > > > Melvin Carvalho wrote: > >> On 2 March 2011 21:02, Nathan <nathan@webr3.org> wrote: >> >>> Melvin Carvalho wrote: >>> >>>> On 2 March 2011 20:40, Nathan <nathan@webr3.org> wrote: >>>> >>>>> is the message here "blank nodes are useless, get rid of them / don't >>>>> use >>>>> them" ? >>>>> >>>> I would be very happy not to use bnodes >>>> >>> why, out of interest, what specific problems cause you to say this? >>> >> >> I know not everyone will agree with this, but I personally find URIs >> simpler to deal with. >> >> I also wonder how to do things like reiificaiton with bnodes. >> > > if there were a canonical representation of RDF so that graphs could be > signed / hashed / compared irrespective of serialization specifics, would > that solve the issues? > you might be interested in "Signing RDF Graphs" by Jeremy Carroll ( http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.58.3198&rep=rep1&type=pdf ) Apache Clerezza also offers a stable serializer which is used for graphs stored into subversion, this makes sure that small changes in the graph result in small changes in the textual diff. reto
Received on Thursday, 3 March 2011 08:01:36 UTC