POP3 Support for UTF-8
draft-ietf-eai-pop-09

My original Discuss read...

> I think it would be valuable (and help with the discussions about an
> Experimental RFC updating a Standards Track one) to document the scope
> of the experiment. That is:
> - why is this an experiment?
> - how is the experiment confined?
> - what are the risks if the experiment escapes?
> - how will you judge the success (or otherwise) of the experiment?

I am going to clear my Discuss because I don't think I should block publication on this issue, but I am not completely happy with the answers.

This document is aimed at being published as an Experimental RFC. It seems to me that this means it is either an experiment in its own right or part of a larger experiment. The former case would have been easy to address by including text to answer my quesitons. The latter case (which I infer from your emails is the actual situation) is even more easily addressed by the inclusion of text such as:
   This document forms part of the larger EAI experiment described
   in RFCxyz (or in the charter of the xyz working group) and will be
   evaluated as part of that experiment.

I detect (from the tone of both Randy and Chris' emails) a slight disatisfaction with the fact that your charter forces you to publish as Experimental. That is a separate discussion that you need to have with your AD. You need to understand and support the limitations of your charter or get them changed.

Personally, I think that Experimental publication is very important and is to be encouraged. Treating the Internet with respect and making changes with caution should be part of the philosophy of the IETF. Introducing new work as Experimental should not delay its development, but should induce appropriate care in how the features are rolled out. Progression from Experimental to Standards Track is not hard, when the time comes.

Nevertheless, I appreciate your efforts to meet me half way, and will clear.

=======

The start of section 2 is a little cryptic! Could you arrange to begin
with some English text that introduces the formal definitions?

Ditto section 3.

---

Agree with Russ that the French needs to be checked, although I disagree
with his interpretation of correct French :-)

support Lars discuss on update

  Lars Eggert's DISCUSS; as Experimental, this cannot update RFC 1939.

  The Gen-ART Review by Brian Carpenter on 2009-10-17 asks some good
  questions:

  I would have expected a reference to RFC 5198 (PS for UTF-8 in
  protocols) as well as RFC 3629.

  I wonder whether any French person has checked the examples such as
  >
  > La Language commande a ete execute avec success
  >
  "Language" is not a French word. The French word for a specific
  language such as French is "langue".  Also, "success" is not a French
  word. It should be "succès" (that's a grave accent on the e, if UTF-8
  didn't quite get through). There are four other accents missing in the
  sentence.  I know we can't yet use UTF-8 in drafts but in that case, I
  suggest either using the usual U+HHHH notation for the accented
  characters, or choosing an example language that doesn't need accents.
  In any case the examples should be checked by a native speaker.
  (Actually the sentence makes very little sense anyway: "The language
  ordered has been executed with success.")

  The examples as presented would make the IETF look a bit silly in France.