Re: Extending news to EAI

Hi Charles,

> So the experimental protocol would start off with the extensions allowe=
d
> by RFC5535, and then add UTF-8 in the Newsgroups header.

I think it will also require to change the possible values for "argument"
here:

   control         =3D  "Control:" SP *WSP control-command *WSP CRLF
   control-command =3D  verb *( 1*WSP argument )
   verb            =3D  token
   argument        =3D  1*( %x21-7E )

We need to be able to use UTF-8 in "argument".


Could it also be possible to extend dist-list to allow UTF-8 distribution=
s?

   distribution    =3D  "Distribution:" SP dist-list CRLF
   dist-list       =3D  *WSP dist-name
                      *( [FWS] "," [FWS] dist-name ) *WSP
   dist-name       =3D  ALPHA / DIGIT
                      *( ALPHA / DIGIT / "+" / "-" / "_" )


Incidentally, may it exist hostnames in UTF-8 (for the Path: header field=
)
or are they always canonized in ASCII?



> It has already been established that the existing transport mechanisms
> will move such articles around without problem.

Is there any record of that?  Which news servers (and versions) were test=
ed?



> Some early USEFOR drafts show how the Newsgroups header was to be
> extended. In particular, it required some very strict normalization, so
> that a simple byte-by-byte comparison of newsgroup-names would always
> work.

The last draft seems to be -09, Section 5.5:
    http://www.eyrie.org/~eagle/usefor/drafts/draft-ietf-usefor-article-0=
9.txt

--=20
Julien =C9LIE

=AB Avez-vous remarqu=E9 qu'=E0 table les mets que l'on vous sert vous me=
ttent
  les mots =E0 la bouche ? =BB (Raymond Devos)=20