- From: Jeremy Carroll <jjc@hpl.hp.com>
- Date: Tue, 24 Apr 2007 10:56:40 +0100
- To: Dan Connolly <connolly@w3.org>
- CC: GRDDL Working Group <public-grddl-wg@w3.org>
I don't find the process argument compelling. It seems to misunderstand the specific nature of a normative dependency between a spec and its test cases, and how the two together fit in a call-for-implementations. A CFI is intended to find bugs in the test cases, rather more than bugs in the spec. How about making the dependency informative, at least for now. We could always review that decision later. Jeremy Dan Connolly wrote: > On Mon, 2007-04-23 at 14:29 +0100, Jeremy Carroll wrote: >> Here are some suggestions concerning the @@'s in editor's draft. >> $Date: 2007/04/18 14:54:08 $ >> $Revision: 1.250 $ >> http://www.w3.org/2004/01/rdxh/spec > > Thanks; I these helped a lot with my TODO list. > > I think this suggestion, which I followed, is > particularly worth noting: > >> @@cite GRDDL test cases? >> ============== >> >> Yes, as a normative ref. With the following text linking to it. >> >> Under >> Preface and Companion Documents >> >> add >> [[ >> <h4 id="intro_testcases">GRDDL Test Cases</h4> >> >> <p>The GRDDL Test Cases<a class="norm" href="#testcases">[testcases]</a> >> provides a collection of tests illustrating this specification. >> Some of the tests may help clarify the intended >> reading of the normative text.</p> >> ]] > > A normative citation on the test cases blows > away the possibility of doing a CR on the spec > while the test cases are in last call. > > (I discovered this while trying to update > the detailed schedule > http://www.w3.org/2001/sw/grddl-wg/sched7 ). > > But it seems like The Right Thing. > -- Hewlett-Packard Limited registered Office: Cain Road, Bracknell, Berks RG12 1HN Registered No: 690597 England
Received on Tuesday, 24 April 2007 09:57:04 UTC