trimming the reply list
On 10/14/25 11:53, [email protected] wrote:
Hi Scott,
On 14.10.25 17:16, Hollenbeck, Scott wrote:
[snip]
... and a question
9.4. of RFC8126 makes a distinction between revocation and release.
Shall it be then also defined as "Records in this registry may be
revoked or released" or it is intended to have only revocation, so the
assignee cannot initiate the removal?
[SAH] That section is about reclaiming previously assigned values for reuse. As
I read that text, the concern is about protocol values that have meaning in
some software implementation, and how a registry change can affect that meaning
in future implementations. I don't think that's a risk here since this registry
has a different purpose.
PK> Yes, that's fine. I just wanted to highlight the risk people with interpret
"revoke" as it is defined in rfc8126 when it is referred to in the same sentence,
even if the context is not the same. One fix could be as I proposed above. Other could be to
narrow down the reference to 4.10 instead of whole document.
There's nothing in the text that says who can (or can't) initiate removal, though. Maybe that should be fixed
by adding "REVOKE" to the list of action requests found earlier in 2.2.3. It also begs the question
of what REVOKE means. Does it mean remove completely, or change the registration state from
"Active" to "Inactive"? I'd prefer to keep old entries listed so that they don't get
lost. The text already describes how to change state from Inactive back to Active, so that's already covered.
PK> I support and this was my feedback before to have "REVOKE" or "REMOVE" request
mentioned in the text. I also support setting to "Inactive" instead of removal.
There should be an option to remove extensions that are inappropriate from the
registry, such as those that may have references to docs that are unacceptable
and those squatting on IETF namespaces. If we want to mark them inactive that
is fine so long as the reference to them is removed.
-andy
_______________________________________________
regext mailing list -- [email protected]
To unsubscribe send an email to [email protected]
