Re: a new dateTime/timezone design, with datatypes

On Thu, 2004-04-15 at 10:23, Masahide Kanzaki wrote:
> At 9:34 AM -0500 04.4.15, Dan Connolly wrote:
> >Nobody should be thinking that this namespace is completely stable.
> >
> >Our policy is...
> >
> >"
> >      * we announce all changes to the schema www-rdf-calendar
> >      * if anyone screams, within a week or so, we'll back out the
> >        changes (for further discussion)"
> 
> I understand what you mean, and some minor changes might be fine.

Yes, it's a question of what's "minor" and so on.

>  But this
> is not a laboratory test tube project. There are existing systems that
> expect the current schema such as eventSherpa. I've tried to introduce
> RDFical, and generated certain amount of RDFical data.

Yes, and I think that's great.

>  It's not happy to
> make those invalid and make people think that RDFical is
> unreliable/unusable.

I understand the costs of change.

I also understand the costs of maintaining multiple
versions, and I understand the cost of keeping a design
fixed when it could have been improved.

I have evaulated all that, and I am reasonably confident
that the new timezone design is worthwhile. But I can understand
if you (or anybody else) come to a different conclusion.

> >If you don't think this change is a sufficient improvement to merit
> >the cost of change, please say so.
> 
> I do not disagree with changes and improvements. Just do not want to make
> existing RDFical invalid. Is there any way to allow but obsolete current
> form (e.g. use new property names for new idea) ?

Hmm... I don't know. Nothing occurs to me just now, but I can imagine
that somebody might come up with a design that does that.


-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
see you at the WWW2004 in NY 17-22 May?

Received on Thursday, 15 April 2004 12:36:50 UTC