At 06:28 PM 2/2/02 +0100, Jacob Palme wrote:
3.4 Change control
Change control of a header registration is subject to the same
condition as the initial registration; i.e. publication of an IESG-
approved RFC for a Normative Message Header, or on request of the
indicated author/change controller for a Provisional Message Header.
I can see problems with this. If a certain header is at the
same time non-standard, and widely used, you cannot allow
the person who happpens to register it first to have change
control. After it has been put into the registry, IETF (or
some other standards organisation) must in some way have
change control, to stop people changing the definition of
widely used, non-standard headers in ways which are not
good and not in accordance with existing use.
Interesting point. I do see a possible related problem, but not exactly
the one you describe.
There is no requirement for uniqueness in the provisional registry. In
theory, the original registrant can change their provisional registration
and if that no longer corresponds to a wideley followed practice of the
original registration then someone else may reinstate the original
registration as a new entry.
The problem I do see is that the change of registration details may occur
silently. But if the cited document is not an RFC, then there is no
(IETF-ish) control over changes to the specification document cited by the
registration.
I have some thoughts about this but, for now, I'm just observing.
Thanks,
#g
------------------------------------------------------------
Graham Klyne Baltimore Technologies
Strategic Research Content Security Group
<Graham(_dot_)Klyne(_at_)Baltimore(_dot_)com> <http://www.mimesweeper.com>
<http://www.baltimore.com>
------------------------------------------------------------