ietf-nntp Re: OVER extension

[Charles Lindsey]

In <20020108193853.GQ10849 at finch-staff-1.demon.net> "Clive D.W. Feather" <clive at demon.net> writes:


>Charles Lindsey said:
>>> My reading of the current specification, *and* the manual page that was
>>> presented to us, is that if an article doesn't contain the relevant header
>>> then the field should be blank.
>> 
>> OK, on a careful re-read of the present text I can see that was the
>> intention. But the wording should be rewritten to remove all doubt (and in
>> particular it should say that "A field may be emtpy ..." is speaking of
>> an absent header rather than an empty one).

>Okay, I've reworded that bit in my proposal:

>  The byte count and line count MUST be decimal integers. They MUST
>  count the entire article, both header and body. The line count is
>  the number of US-ASCII CRLF pairs, each of which counts as two bytes.
>  These values MUST be returned; the corresponding fields MUST NOT be
>  empty.

I thought somebody said that byte count included the headers, but line
count didn't.

>[Possible addition]
>? Note: even if an article contains a Lines: or Bytes: header, the server
>? MUST NOT trust it but must generate its own value.

I think I would say:

	Although articles sometimes contain a Lines header, the server
	MUST not assume this is correct, and MUST compute the line count
	itself; likewise with any header that purports to convey the byte
	count.

>  For the five mandatory headers, the content of each field MUST be
>  based on the original header with the header name and following colon
>  and space removed. If the article does not contain that header, or if
>  there is nothing following the colon and space, the field MUST be
>  empty.

>  For all subsequent fields the content MUST be based on the entire
>  header, including the name. In every case, the header is processed
>  by first removing all US-ASCII CRLF pairs and then replacing each
>  remaining US-ASCII NUL, TAB, CR, or LF character with a single
>  US-ASCII space (for example, CR LF LF TAB will become two spaces). If
>  there is no such header in the article, the field MUST be empty. If
>  the header exists but has nothing following the colon and space, the
>  field MUST NOT be empty.

Not quite. The business of removing all US-ASCII CRLF pairs and so on
applies to the mandatory headers as well as the subsequent ones.

>> Son-of-1036 took a dim view of empty headers. USEFOR does not forbid them,
>> but it carefully ensures that it does not define any headers that could be
>> empty.

>Fine, but we should still allow for it.

Indeed.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: +44 161 436 6131 Fax: +44 161 436 6133   Web: http://www.cs.man.ac.uk/~chl
Email: chl at clw.cs.man.ac.uk      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