As a community, we are still struggling with where to draw the
boundary between what is generally accepted and thus should be
standardized, and what is still unknown enough and thus needs
competition to figure out what the best approach is and for what use
cases.
It appears that we agree on for sure:
- discovery (aka Yadis)
- shared-secret-based authentication (aka OpenID authentication)
It appears that we might be able to reach consensus on:
- a minimal vocabulary (at least I have not heard different
factions argue in favor of different, competing, vocabularies, for
reasons other than the next item ... see below ...)
So far, there has been disagreement on:
- the underlying data model for vocabularies (based on tree, a
directed graph, or simply name-value pairs)
- the timing and triggers that cause data to move
- the mechanisms by which data moves.
Depending on the day of the week, we seem to all be more or less
confident on whether we can (or should, at this time!) get agreement
on the things that we don't always agree on.
I would agree with you that we should avoid premature standardization
for things that don't have sufficient consensus yet.
Others argue that it is better to pick one of the alternatives as a
standard, and fix it later if needed, than to not have a standardized
facility for an important function (an argument that I understand).
The current approach seems to be to go down the standardization route
and see how far we can go. It is certainly possible that as we go
down that route and improve the proposed specs, some of the
objections will disappear ...
On Sep 26, 2006, at 0:36, Martin Atkins wrote:
Johannes Ernst wrote:
I thought we had consensus that Drummond and I owned the
vocabularies/ontologies/schema/whatever-we-call-it action item?
The more the merrier, in my opinion.
Let the market decide!
_______________________________________________
general mailing list
[EMAIL PROTECTED]
http://openid.net/mailman/listinfo/general
Johannes Ernst
NetMesh Inc.
http://netmesh.info/jernst
_______________________________________________
specs mailing list
[email protected]
http://openid.net/mailman/listinfo/specs
