On Thu, 28 Aug 2003 18:27:36 -0700
Russ Allbery <rra(_at_)stanford(_dot_)edu> said...
I'm sorry, you're right. I was insufficiently precise. The current,
completely out-dated and inadequate-for-implementation NNTP standard, RFC
977, limits the message ID to around 500 octets indirectly by limits on
the length of command arguments. (It doesn't address the point directly.)
The upcoming standards track replacement, which is very close to last call
and which is a much better basis for implementation, states:
o A message-id MUST begin with "<" and end with ">", and MUST NOT
contain the latter except at the end.
o A message-id MUST be between 3 and 250 octets in length.
o A message-id MUST NOT contain octets other than printable US-ASCII
characters.
The not-quite-so-upcoming standards track replacement for RFC 1036
(draft-ietf-usefor-article-11/txt) also imposes the same length
restriction (and some slightly tighter restrictions on the rest of the
syntax).
There was recent discussion on both lists, and it was agreed that some
limit was needed for technical reasons, that 250 was about the right
size and was also built into several existing servers, and therefore was
as good a number as any.
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_)clerew(_dot_)man(_dot_)ac(_dot_)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