ietf-nntp Comments on the lastest rfc977bis draft

[Stan Barber]

I have substatially clarified the LIST EXTENSIONS response. I believe folks
should be able to actually implement this now. I also simplified and generalized
the extensions concept a bit. This is something we will need to discuss on
Monday.

I have taken all the comments I got over the course of the last four months
an intergrated them into the document. There are a couple of troublesome 
areas:
	1. AUTHINFO continues to be trouble. I am tempted to remove it from
	the current document and do it as a seperate extension so that all
	the various issues in that area could be delt with independent of the
	main document. Such a move might be too radical since there would then
	be no identification mechanism in base document. I think the biggest
	problem here is that there are implementations that do certain things
	in certain ways and either we continue down that path (and add other
	paths as well) or we don't.

	2. The document does not deal well with responses to BODY or ARTICLE
	that are binary. It also does not deal well with POST when the body is
	binary. The problem here is that the termination sequence is CRLF.CRLF.
	However, we are not clear on wether the first CRLF is part of the
	body of the article or not. We need some consensus here.

Finally, we need to be sure we are confortable with all the UTF-8 versus US-ASCII
issues. I think we've done ok, but I am no expert in the UTF-8 stuff. Comments
from experts in that area continue to be most welcome.

Thanks in advance for your thoughts on any of this. Please be sure to quote
relevant parts of the current draft when you comment and offer specific
changes that will clarify any of these issues.