From: John Stanley (stanley@peak.org)
Date: Wed Feb 04 2004 - 12:45:26 CST
J. B. Moreno (planb@newsreaders.com):
>Extra stuff added to a header that isn't authorized by the user and isn't
>required by the protocol -- that sounds like a protocol issue to me,
Stuff that isn't required by the protocol is a protocol issue? WTF?
>partiuclarly when the extra information interferes with the proper
>operation of the header.
What is interfering with the "proper operation" of the header? It's
unstructured; the only thing that could interfere with that is if some
broken software tries to enforce structure on it. And what is the "proper
operation" of the header? It has no function in the news system other
than to report to the reader what the subject of the article is. "Re: "
has nothing to do with that.
>If you allow "Re^2: " and "SV:" and "Anwr:" where does it end and become a
>protocol issue?
It is an unstructured header. You cannot NOT allow "Re^2: " or anything
else. And it is an unstructured header; treating it that way will not
ever be a protocol problem.
>If the client slips "Your mama wears combat boots:" in at
>the start of every Subject and then hides that from it's user, ...
It is the user's choice to use that client. If the user chooses such
a client, who are we to tell him he cannot?
> ... is that just a display issue?
It certainly isn't an issue for this standard.
>What the hell is wrong with saying that the structure of the Subject is:
That's not what the "structure" of the subject is defined to be. You
cannot say "the structure of an unstructured header is X". We're
talking about the section of this standard that deals with "Re: ", not
with the base definition of the header.
And, just to be pedantic, your new definition of the subject header is
functionally identical to the current one. If the subject happens to be
"Re: foo is good", then it fits the "subject-content" definition either
with or without the optional "back-reference" element. A news reader has
no way of knowing whether the "Re: " is "back-reference" or just part of
the following unstructured "pure-subject". It would know that "Sv: foo is
good" does not have a legal back-reference, which is not different from
the existing definition.
>And article grouping/threading by the client should be considered just as
>important as grouping by the server -- break it and news becomes unusable.
Yes, and that is why References is mandatory -- it is how articles are
threaded, by definition.