Re: IANA assignments

[Thomas Narten <narten@us.ibm.com>]

>   All I am saying is, IANA can have a guideline for long lived
>   drafts, such as:
>   
>      If there is a shipped implementation with an old deprecated
>      value, that value MUST be reserved to prevent future misuse.
>      It MUST be noted in the subsequent drafts and final RFC.

Well, a more proper way to express the above (process wise) would be
to have an IANA considerations Section (for the name space at issue)
of something like:

   values in the range 0-128 are assigned via Standards Action or
   Expert Review. The purpose of Expert Review assignments is for the
   assignment of code points in cases where the document is expected
   to become published as a standards track RFC (e.g., it is a Working
   Group document), but for which experimentation is desired before
   the RFC is published.

Tune further as appropriate. Then, when a WG needs an assignment, it
asks IANA to make the assignment, IANA checks with the expert, and
assuming the request is proper, the assignment gets made.

The point is that there are ways for the WG to retain control over
assignments that IANA is managing. That is what IANA Considerations
sections are for.

Thomas