ietf-nntp NNTP and 16-bit charsets

Clive D.W. Feather clive at demon.net
Wed May 2 01:32:21 PDT 2001


Charles Lindsey said:
> I have just one problem with a binary extension to NNTP. An article
> normally travels through many relayers. What happens if only some of them
> support the extension? The last thing I would like to see is for relayers
> en route to be given carte blanche to change the encoding from binary to
> base64 and back. We have seen the terrible mess that has arisen from the
> propensity of SMTP servers to do just that.

The issue of encoding, and changing any encoding, is one for Usefor, not
for this list. We should be clear that any binary transport extension is
simply to allow a wider character set through the transport.

I don't think you can say that relayers must not recode. Rather, it is an
issue for Usefor to decide whether and when it is allowed.

> So I think it would have to be clear that articles encoded with CTE binary
> would have to be rejected by (or not offered to) servers not supporting
> the extension.

I think the way you put it is:

    If a multiline continuation (whether part of the command as with
    POST or part of the response as with ARTICLE) contains byte sequences
    that do not conform to the rules concerning multiline content,
[this refers to your stuff about NUL and CRLF]
    then it may only be transmitted using the binary extension. If either
    end does not support this extension the material MUST NOT be recoded
    and MUST NOT be sent; if it is part of a response the server MUST send
    a failure response.

Argh; that's bad wording, but I'm sure you know what I mean.

> This would be fine for cooperating subnets that agreed to
> support the extension, but it would mean that articles would have poor
> propagation, or would flood around some very circuitous routes, on the
> wider Usenet. In time, of course, they would propagate better as the
> extension became more widely implemented. Subject to that understanding,
> the extension should, of course, be welcomed.

I think that's the only workable approach. If people want binary articles,
the extensions will appear.

-- 
Clive D.W. Feather  | Work:  <clive at demon.net>   | Tel:  +44 20 8371 1138
Internet Expert     | Home:  <clive at davros.org>  | Fax:  +44 20 8371 1037
Demon Internet      | WWW: http://www.davros.org | DFax: +44 20 8371 4037
Thus plc            |                            | Mobile: +44 7973 377646 



More information about the ietf-nntp mailing list