Re: MV counterexample
Date: Fri, 07 May 2004 00:57:55 +0200
Message-ID: <409ac2f4$0$15375$e4fe514c_at_news.xs4all.nl>
Dawn M. Wolthuis wrote:
> mAsterdam wrote:
>>Dawn M. Wolthuis wrote: >>>Combining these very conceptually simple concepts of: >>>string (of data), graph (for navigation), and function >>>(mapping) with data is so intuitive for >>>the human brain that I would guess that >>>if you put these concepts together >>>into some sort of web of information, >>>it just might catch on in a world-wide fashion. >> >>Nah. Incomplete. No enforced referential >>constraints.
>
> I don't know about that -- I see entire propositions, including child
> clauses, altogether on a single page. So, there are SOME such constraints.
> When the page is removed, the child clauses seem to be gone too.
Nice. A page as a piece of a database.
A good set of DTD's would be needed as
a cornerstone to another distributed
database (than the last one
we talked about, but similar in working).
Maybe http://web.resource.org/rss/
is a step in that direction.
>>You would get stale links in no time, and your database >>would be corrupted. End of exercise.
>
> It would surely be useless if there were any
> "page not found" responses --
> human beings would not put up with that ;-)
Totally unacceptable from a
conceptual integrity point of view.
> It isn't rocket science to prepare a web site that has no problem links if
> you have control of all linking and linked pages. But once that is not the
> case -- once you have a link from a zip code in your data to a map that
> shows where that zip code is located and you don't control the map, then you
> might have a missing link. We have seen cases where users can put up with
> that for the advantages they get when the link is there.
>
>>It would make a nice TLA, though.
>
> and it seems to be a model "IN PRODUCTION" at practially every company --
> quite impressive that WWW "data model" eh? --dawn
Yup. Received on Fri May 07 2004 - 00:57:55 CEST