Re: status of http://www.w3.org/2002/03owlt/

On Tue, 2002-08-27 at 08:34, Peter F. Patel-Schneider wrote:
> 
> From: "Jeremy Carroll" <jjc@hplb.hpl.hp.com>
> Subject: Re: status of http://www.w3.org/2002/03owlt/ (Re: TEST: Functional and InverseFunctional tests for approval)
> Date: Tue, 27 Aug 2002 13:00:21 +0200
> 
> > Peter asks:
> > > What is the status of this directory?
> > 
> > > Can anyone put tests in it?
> > 
> > Practically no.
> > You need to jump through a few hoops with Dan to get CVS access to the
> > archive.
> > I see no reason why you shouldn't though. The whole directory is such a mess
> > that a few more arbitrary additions couldn't make it much worse.
> > 
> > I would be pleased if you did create some tests and added them.
> > 
> > A proposed clear up approach is to machine process the manifest files to
> > generate index files that only show the good stuff eventually.
> > Frankly I wouldn't be surprised if we need a much more dictatorial approach
> > in which a dictator with sole control of a new directory copies approved
> > tests into it, making editorial changes at whim. Currently we only have
> > about five approved tests.
> 
> I propose the following:
> 
> 1/ No tests be added to http://www.w3.org/2002/03owlt/ until the working
>    group approves a document that specifies how tests are to be created and
>    maintained and this document is pointed to by the WG main web page or
>    the WG main test page.
>
> 2/ No tests be placed in http://www.w3.org/2002/03owlt/ unless they are
>    approved by the WG.
> 3/ Part of the test document is a categorization of tests.
> 4/ No tests be approved by the WG unless they have an accompanying
>    description and belong to one of the test categories.

That's more of a constraint than I'm interested in working under.

> > > If not, why is Jeremy Carroll putting tests in it in
> > > advance of working group approval?
> > 
> > When I add a test I add an entry in the Manifest file of that directory
> > showing that the status is "PROPOSED". When and if it is approved the status
> > line changes, and a link to the approving minutes should be added.
> 
> I would much prefer that tests in this directory only be approved tests.

Why does it matter how the filesystem is arranged?

I can arrange for only the approved test cases to be linked from
the main test repository page; would that suffice?

> 
> > Jeremy
> 
> peter
-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/

Received on Tuesday, 27 August 2002 09:46:27 UTC