Re: ISSUES: First draft of open issues document.

On Mon, 2002-03-04 at 17:20, Smith, Michael K wrote:
> Looking for feedback.

Great start...

>  I have erred on the side of over-inclusion in order
> to avoid dropping issues through the cracks. 
>  
> 1. Have I got the W3C boilerplate right for this doc?

There aren't any rules for what issues lists look like.

Whatever serves our purposes is the Right Thing.

For my purposes, making each of the "origin" thingies
an actual link to the origin of the issue is *very*
valuable. Critical, I think.

Also, note that the whole world can read this thing, so
we need to be clear about whether stuff in it is
	-- agreed by the whole WG?
	-- proposed by one member of the WG?
	-- something in between?

It's reasonably clear in its present state, but
you could add that (a) the list was produced
by just one person, not reviewed yet, and (b)
the text of the summaries is just yours,
not agreed by the group.



Also... there are rules that say we have
to close all our issues before we're done:

"5.2.2 Last Call Working Draft

Entrance criteria. Before advancing a technical report to Last Call
Working Draft, the Working Group must:

   [...]
   2. formally address all issues raised by Working Group participants,
other Working Groups, the Membership, and the public about the Working
Draft."

	-- http://www.w3.org/Consortium/Process-20010719/tr.html#last-call


So we should excercise some discipline in putting stuff on the issues
list, since we'll need to resolve it before we can claim victory.

Hmm... another point: is the numbering you're using stable?
i.e. do you promise that 2.1 will always refer
to "URI naming of instances"? or is there a chance that
you'll add a section betwewen 1 and 2? We need issue
names/numbers that we can use for the next year or so.

I like the combination of a short name and a number,
like we're using in the TAG:
whenToUseGet-7 http://www.w3.org/2001/tag/ilist#whenToUseGet-7

The serial number serves all the purposes that serial
numbers are good for; but serial numbers alone
create the sort of obscurity PeterPS has warned
about recently; it's intimidating to join a group,
look at the archives, and see long threads
under the sujbect Re: Issue 208.

"I think that we should not be in the business of making our
messages unnecessarily cryptic."
	-- http://lists.w3.org/Archives/Public/www-webont-wg/2002Mar/0011.html



> 2. For the sections on implicit and unmentioned requirements, we need some
> more text before these are worth discussing.  

What's there suffices, for my purposes (once you add the
hypertext links).

> 3. What have I missed?

Hmm... is now the time to collect issues about DAML+OIL, while
we're at it?

e.g. the joint-committee didn't reach consensus about
splitting the domain of discourse among datatype values
and objects; we planned to revisit it. This
seems like the time/place. I guess I'll send mail
separately about that...


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

Received on Monday, 4 March 2002 22:44:07 UTC