Re: [Editorial Errata Reported] RFC5537 (1980)

Julien ÉLIE <[email protected]> Mon, 28 Dec 2009 19:47:57 +0100
Newsgroups gmane.ietf.usenet.format
Organization TrigoFACILE -- http://www.trigofacile.com/
Message-ID <4E0AAD5953E54641B28EF7EAAD8D7202@Iulius>
Hi Alfred,

I have two major remarks:

1/ If your report is verified, there is also Section 3.4, fourth paragrap=
h,
to amend:

    predating this specification do not add an Injection-Date header.

"field" is missing.


2/ When you say:
>  Similarly, denoting as "header field" a "header field value" is
>  confusing -- items (e), (f), (g), (i), and (j) above.

Where do you draw the "header field value" terminology?
Isn't it "header field body"?  RFCs 5322 and 5536 define
"header field body" but not "header field value".




And also a few other points:

Note that when for instance we have in Section 6.1 of RFC 5537:

   To best ensure that it will be relayed to the same news servers as
   the original message, a cancel control message SHOULD have the same
   Newsgroups header field as the message it is cancelling.

or in 6.1:

   Cancel control messages are not required to have the same Newsgroups
   header field as the messages they are cancelling.  Since they are
   sometimes processed before the original message is received, it may
   not be possible to check that the Newsgroup header fields match.


of course it is not exactly the same header field because
   Newsgroups: group1,group2
and
   Newsgroups:    groups1,group2
are equivalent...  WSP and FWS should not be taken into account...

Should such terminology also be fixed?



And in Section 5.5:

   Contrary to [RFC1036], the relayer-name MUST be given as the last
   argument in the Control header field.

Wouldn't it be better to say it is the header body?



> Rationale:
>  Contrary to its companion document, RFC 5536, this RFC mixes precise
>  IETF terminology for protocol elements and colloquial abuse of it in
>  various places.  For clarity and consistency, it should also
>  inequivocally make use of the standard terminology; the fields
>  of the "header" that a protocol layer or sub-layer adds to its
>  payload are "header fields", not "headers" in itself.


Shouldn't RFC 3977 also be checked for that?

   The distrib.pats list is maintained by some NNTP servers to assist
   clients to choose a value for the content of the Distribution header
   of a news article being posted.

   Lines or Bytes header

and other similar terminology.  RFC 3977 defines headers this way:

   The headers of an article consist of one or more header lines.  Each
   header line consists of a header name, a colon, a space, the header
   content, and a CRLF, in that order.

--=20
Julien =C9LIE

=AB Ne parlez jamais de vous, ni en bien, car on ne vous
  croirait pas, ni en mal car on ne vous croirait que trop. =BB (Confuciu=
s)=20