Re: [regext] [Ext] regarding adopting new documents and milestones

[Patrick Mevzek <[e-mail]>]

On Mon, Oct 29, 2018, at 21:21, Andrew Newton wrote:
> Perhaps priority should be given to those I-Ds with running code.

This is one of the point I mage in my July message:
https://mailarchive.ietf.org/arch/msg/regext/DuitffLvC4Q5RR32AnN1yDEUEYg

Running code is useful/needed but maybe too high a barrier for an I-D to become a working group document. It can happen during its life I guess, but more important for me would be to have at least 2 completely separate *registries* implementation, which then could be a good idea of something that was in my July message.

In short, not all extensions may become used by more than one registry. As sad as it may be for registrars (if the extension solves a generic issue that happens elsewhere or could benefit elsewhere), it is the state we are in.

However in that, how much of working group energy should be devoted to that, and is the Proposed Standard path really necessary?
Maybe just an addition of the extension in the EPP Extension Registry should be enough?
There are a lot of extensions out there and discussed, and very few recorded in the extension registry. Basically only 3 registries took the effort to add their extensions there.

And as said, sometimes an extension really tailored to one use case and one registry will need a lot of work to be more generic, and various feedback about needs of proper specification and taking care of interoperability might not be so much welcome in fact.

So in summary I could say that I am not in favour of hardcoding any specific number of IDs to work on, but I would advise that 1) not all extensions need to become an RFC, being a proper ID with correct structure and registration in the extension registry is already a very good step and 2) for those wanting to be an RFC and before that wishing for the working group to work on, then they should really make sure to both display that the need spans multiple registries/registrars and that the document is really open to be updated to take into account more generic cases as needed as well as various implementation enhancements that are often only discovered when implementations are done.

-- 
  Patrick Mevzek
  [e-mail]