Jim Whitehead ejw@cse.ucsc.edu
Wed, 19 Dec 2001 12:34:53 -0800

> For goodness' sake, Jim, Crit predated WebDAV by about four years.
> If you're really going to ask this, then i suggest that the opposite
> question is more appropriate: why did no one from your group consider
> how WebDAV could be designed to interoperate with Crit?

I would claim that DAV, as designed, does provide a useful service layer
that Crit could use.

If I understand it correctly, Crit's architecture is proxy-based: pages are
fetched from an origin server, and annotations are added by the proxy. The
proxy controls the storage of the annotations.

This could be modified so that Crit shares storage of annotations. If the
origin server supports DAV, and allows property modification by the Crit
server, then the DAV server could store the annotations in properties on the
source HTML document. If these conditions are not met, Crit would default to
normal operation, which is to provide annotation storage.

> Why does no one in your group or the W3C yet acknowledge or cite Crit
> in your documents as related work?  When has anyone from your group
> consulted me or expressed interest in collaborating on it?

Two factors. I didn't become personally aware of Crit until 1999/2000, and
then it's not clear to me that Crit is *directly* related work for WebDAV.
It doesn't seem quite as related as, say, the literature in the CSCW
community on remote collaborative authoring.

But, if you'll look at the Remote Collaborative Authoring Resources page I
maintain at:

There is a link to Crit.org at the bottom of the page (and this link has
been there for a long time). So I do recognize that Crit is related to
remote collaborative authoring, construed broadly.

> (I'm not angry at you, Jim.  But if you must raise the question,
> then i have a few questions of my own.)

As for the W3C, since I am not a W3C staff member or supported researcher, I
cannot comment. Although, I'll note that the Annotea group does reference
the Crit Hypertext'98 short paper in their WWW10 paper.

> The XLink/XPointer and annotations groups have completely ignored
> this prior work, instead building a vastly more complicated and
> difficult-to-understand model that is to this day not usable or
> accessible from any common browser, while Crit supported them all
> from day one.  Where is my motivation to continue working on Crit
> when the W3C has consistently ignored real working implementations
> like Crit and MINSE in favour of complex unimplemented specifications
> like XPointer and MathML?

Steve DeRose seems like a better person to ask than me.

If I were to speculate as to why XLink/XPointer didn't use your work as a
starting point, it might be due to having a broader set of requirements than
are handled by

- Jim