W3C home > Mailing lists > Public > public-rdf-wg@w3.org > September 2013

Re: Remaining Work To Do

From: Gregg Kellogg <gregg@greggkellogg.net>
Date: Wed, 11 Sep 2013 11:41:37 +0800
Message-Id: <159F7F0F-0619-4C0C-A1FA-8EFBF420CC80@greggkellogg.net>
Cc: RDF WG <public-rdf-wg@w3.org>
To: Sandro Hawke <sandro@w3.org>
On Sep 11, 2013, at 3:44 AM, Sandro Hawke <sandro@w3.org> wrote:

> It looks like we have about 14 or 15 more weeks of this WG.    I'm looking over what's left, trying to see what we need to do right away and/or get people particularly motivated to work on.
> 
> In general, I'm seeing the remaining work falling into a few categories:
> 
> 1.   CR stuff: defining what is tested, defining tests, approving tests, creating implementations, receiving test results, generating an implementation report from the test results.
> 
> A few thoughts:
> - RDF Concepts says it's not testable; RDF Semantics looks like it should be testable the same way it was in RDF 1.0, with positive and negative entailment and consistency tests, but ... we neglected to say this in RDF Semantics, so where can we say this?   I guess we can say it non-normatively in the test suite.    That's a bit weak.   Anyway, we at least need someone to take the 2004 test suite and get it into a form people can use now, and add/modify all the tests to show all the changes in semantics since 2004.   Pat and Peter, are you up for writing those new/modified tests?
> 
> - Gregg, you've been doing the implementation reports for the syntaxes -- are you up for doing it for Semantics, and anything else that comes along?    How hard would it be for someone else to adapt/run your code, if they don't know ruby?

I'm certainly up for more reports. It's completely re targetable, and is used for about 4 different. It's re targetable, and anything that uses a manifest structure similar to Turtle's will work out of the box.

Of course, I can help in getting the test infrastructure together too.

Gregg

P.S. out of the office for the next couple of weeks.

> 2.  Handling User Comments.   Obviously this is work, and not (usually) a lot of fun.   It's one of the things that's holding Turtle back at the moment.    I'm not confident everything sent to public-rdf-concepts has been tracked on the page for the right deliverable.    Is there anyone who is currently taking responsibility for that?
> 
> 3.  Hard decisions about issues users raise.  This is the big unknown.   We should make sure any non-editorial comments get RAISED as issues immediately, so the chairs can try to fit any necessary discussions into the remaining meetings.  For example, I see Jan's W's Turtle comment about how to name the two versions of Turtle, but it's not an ISSUE.   We've got 5-10 hours of telecon time to close all these issues, including ones that haven't been reported yet.
> 
> 4.  The unpublished documents.   RDF Schema, RDF Primer, RDF/XML Syntax, and the Dataset WG Notes promised by me and Antoine.   I think we also need something, maybe just a web page, saying what RDF 1.1 is, as compared to RDF 2004.  All the WG's I've been involved with *since* RDF 2004 have made an Overview document.   We may want to set deadlines for each of these, by which if the group doesn't have a draft to review for publication, we say we wont publish them.   We may want to come up with contingency plans -- is it okay to not touch the RDF/XML spec?  (I can add an alert box pointing folks at some other spec without republishing it.   But we'd need something to point to, like that "what's new in rdf 1.1", which links to turtle, trig, json-ld, and maybe lets folks know rdfa exists.)    I believe Dan is on the hook for Schema, Guus for Primer, and Antoine and I for our notes.    Anyone for RDF/XML and Overview?    Which of those are Rec Track?   If Schema, Primer, RDF/XML Syntax, and Overview want to be/remain on the Rec Track, we need drafts *very* soon.
> 
> 5.  Publication mechanics.    We're looking at 12-14 documents, most of which need to be published 2-3 times in the next three months. They need to all point to the correct version of each other, with every publication, and keep passing pubrules.   That's a lot of publications.   We either need editors all standing by, making sure they can do the right things at the right time, or we need to make sure everyone's using identically configured respec, so someone [me] can do the publications without knowing about idiosyncrasies. Probably we should do both.
> 
> Am I missing anything big here?
> 
>       - Sandro
> 
> 
Received on Wednesday, 11 September 2013 03:42:13 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 22:02:16 UTC