I-D ACTION:draft-ietf-nntpext-base-06.txt

[Charles Lindsey]

In <IUCUNoDOx9Q2EwQW at on-the-train.demon.co.uk> "Clive D.W. Feather" <clive at on-the-train.demon.co.uk> writes:

>>As far as I understood, the primary intent is that an article I have
>>read already should not reappear as unread.  Wouldn't it be easier to
>>require the reading agent to check whether the superseded article has
>>already been read? This could be easily implemented by using the Xref
>>header of the superseded article! 

>Indeed. Or, since Replaces: is new, we could require the Xref header in
>the replacing article to contain both sets of numbers:

>Xref: demon.co.uk alt.fuzzy:1256/1234,alt.soggy:9987/9944

Yes, an interesting idea. It would mean a change to the Xref header, so we
need to take it back to the other group. Two questions.

1. Presumably 1256 is the new number and 1234 the old. So it gets stored
as 1256. So when you ask for news articles, NNTP downloads the whole
article to you (of perhaps just the header if that is what you asked for),
and the decision to throw it all away is made by the reading agent. Is
that acceptable? The other implementation (reusing the old number) means
that the article is never downloaded at all in normal operation. Or would
you have a variant on the ARTICLE command that allowed you to specify
which old article numbers you would not be interested in?

2. Does the new Xref header break existing newsreaders? Or, who actually
looks at the Xref header anyway? To avoid seeing a crossposted article in
other groups? Would that part of existing implementations get broken?

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Email:     chl at clw.cs.man.ac.uk  Web:   http://www.cs.man.ac.uk/~chl
Voice/Fax: +44 161 437 4506      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9     Fingerprint: 73 6D C2 51 93 A0 01 E7  65 E8 64 7E 14 A4 AB A5