- From: Dave Beckett <dave@dajobe.org>
- Date: Wed, 03 Oct 2007 07:05:22 -0700
- To: Ivan Herman <ivan@w3.org>
- Cc: Story Henry <henry.story@bblfish.net>, Michael Schneider <schneid@fzi.de>, semantic-web@w3.org
Ivan Herman wrote: > My 2 cents... > > Regardless of the advantages, issues and objections on this very > proposal... I think we should be careful about priorities. In my view > the goal should be to have an absolutely stable turtle language that is > as closely compatible to the language used in SPARQL as possible (it > cannot be 100%, because, for example, SPARQL uses PREFIX and turtle uses > @prefix). Personally, I would prefer _not_ to start discussions about > new language features for now, much as it can be an enjoyable technical > discussion:-). > > Maybe we can do something like that in a next round, but the priorities > should be elsewhere in my view. > > And thanks to Dave for starting this... > > Ivan > > Story Henry wrote: >> What about = >> >> That would be a nice addition. :-) >> >> Especially as it allows one to group statements together like this: >> >> :me foaf:knows [ = <http://eg.com/joe/>; >> foaf:name "Joe Smith" ] . >> Many of the major reasons I'd not want to do this have been mentioned 1) It's a specific vocabulary burnt in. 'a' should be the only one. 2) It is not in SPARQL's triple syntax. 3) I am trying to stabilise Turtle by not adding new syntax now. 4) It means two things; an abbreviation for owl:sameAs and naming a blank node. If I understand it correctly. This seems bad. plus my own: 5) It has been rarely requested, rare enough I wouldn't consider adding it. 6) Naming a blank node [ = <uri> ] I find muddies what [] means. 7) I find it hard to read. Dave
Received on Wednesday, 3 October 2007 14:05:48 UTC