-----Original Message-----
From: Keith Moore [mailto:moore(_at_)cs(_dot_)utk(_dot_)edu]
Sent: Monday, February 11, 2002 10:12 AM
... however we may disagree about how "light-weight" the
registration should be.
I still want to see a brief review period, with community
input, before registration is accepted, and I want to have
the possibility for the registration to be deined - especially
if it's abusive or misleading. I think that will work better
than automatically accepting whatever extensions are proposed.
I think the default action should be to register a field, with
provisions for de-registering abusive ones (assuming you mean personally
abusive, like "So-and-so-is-a-doofus: Yes"). The reality is that
refusing to register a field does not mean that it won't be deployed, it
just means that there is no place for me, as an implementer, to find out
anything about that header when I encounter it. I don't think a
registration should ever be revoked just because it's a bad idea - if
it's out there, it's out there and I want to know that I'd better not
use the name "FooBar" for a field because there's already some dumb UA
using that name in a totally stupid manner and who knows what will
happen to my UA if it ever receives one of its messages.
The registry is not, IMO, about approval - it is about notifying the
Internet community of a field and making information about it readily
available. That information should include the community's assessment
of the value/danger of the field, but we should not refuse registration
just because a field is deemed dangerous. We should, in fact,
*encourage* it because the registration provides a forum for the
community to warn about the dangers of the field (assuming we can't
influence it to make it non-dangerous).
-- jeff