- From: Alex Hall <alexhall@revelytix.com>
- Date: Wed, 2 Mar 2011 14:25:56 -0500
- To: RDF-WG <public-rdf-wg@w3.org>
- Message-ID: <AANLkTikr9A0=2VzcqsPtmJJCFEb_L_Q66J-80bnUsVZg@mail.gmail.com>
On Wed, Mar 2, 2011 at 1:55 PM, Nathan <nathan@webr3.org> wrote: > Hi All, > > Okay, I'm lost, as to what's what and what is in scope, out of scope, > possible and not. > > Is B.C. for turtle (data and consumers) to be maintained? > I was about to ask for clarification on this same issue. The only mention of backwards compatibility in the charter is with respect to graphs and entailments: any existing RDF graph/entailment must be a valid graph/entailment under the new specs. Applied to Turtle, backwards compatibility would mean that any existing Turtle document is still a valid Turtle document under the new specs. I think that is important to maintain. Under this definition, the proposed syntactic extensions to Turtle are backward compatible but TRiG is not (because triples are not allowed to appear outside a graph block). The reverse of that is forward compatibility -- if a deployment supports Turtle v1 then will it also support Turtle v2? In general I don't think it's important to maintain forward compatibility between versions, but I would assert that what we should be doing is standardizing version 1 of Turtle before we start working on version 2, or even 1.x. If we throw a bunch of new features into the existing Turtle submission and call the result "Turtle 1.0" then I suspect we'll be breaking a lot of existing deployments. -Alex
Received on Wednesday, 2 March 2011 19:26:29 UTC