[NNTP] Header terminology

[Russ Allbery]

Forwarding from USEFOR.  What do folks think about making the editorial
changes noted below for compatibility with mail terminology?  See the end
of Richard's message for the exact details; other than in the base
document (which might not be worth the hassle), the changes are trivial
and I don't think they reduce the readability any.

It sticks in my mind that we may have raised this before, but I don't
remember for sure so I'm asking again just in case.


Date: Sun, 12 Jun 2005 12:52:34 +0100
To: ietf-usefor at imc.org
From: Richard Clayton <richard at highwayman.com>
Subject: Re: #1031 backward compatibility and handling incompatibilities
X-Mailer: Turnpike Integrated Version 5.02 M <Hm3$+Pcr77fJAOKLReX+dOUPoc>

In message <IHvsE0.Mpx at clerew.man.ac.uk>, Charles Lindsey
<chl at clerew.man.ac.uk> writes

>I think he is upset because we use the term "header" rather than "header
>field" and "header name" rather than "field name".
>
>I would point out that our drafts have always used this terminology (and
>have carefully documented that usage).

Although I have sympathy for the view that this is continuing the
tradition of how "news" documentation is written, I think that's a
mistake :(

Considerable leverage is gained by defining "news" as a special case of
"mail" (albeit, I sometimes wonder if -- when the minutiae is debated at
such length -- the gain is entirely worth the pain).

Whilst we continue down that path then it behoves us to use the jargon
from "mail" in a consistent and user-friendly manner.

Although there may be mileage in saying "In this standard we use frooble
to mean 'some very long phrase that is hard to keep repeating'" I don't
think that it does anything other than create confusion to use a
substitute for quite short phrases AND what's more, to use a word
"header" that so closely resembles what is being substituted.

As a data point, I'm currently in the process of finally editing my PhD
thesis (yes, they let you do PhDs even at my advanced age!) and realised
that I'd been very sloppy with my own use of the word "header". It took
less than an hour to work through the document and fix every mention to
the correct 2822 jargon.  I really do not think that it reads any worse
now that I've done it, nor is it harder to understand for a neophyte.

>RFC 1036 used the term "header" (though not consistently, since it
>sometimes used "header line" - a quick look revealed no occurrence of
>"field", though I have not grepped to confirm it).
>
>Son-of-1036 used the term "header".
>
>The NNTP draft, which has just completed its IETF Last Call period, uses the
>term "header".

of these three points, this is the _only_ one that would concern me (we
are, I trust, hoping to reach a situation where only historians consult
the other two documents, so who cares what they say?) ... so I checked
the documents recently exiting last call, which appear to be:

http://www.ietf.org/internet-drafts/draft-ietf-nntpext-authinfo-09.txt

        which contains the single phrase "This may be accomplished, for
        example, by inserting headers in the posted articles ..."

        ie: trivial change to "header fields"

http://www.ietf.org/internet-drafts/draft-ietf-nntpext-tls-nntp-06.txt

        which doesn't contain "header" at all

http://www.ietf.org/internet-drafts/draft-ietf-nntpext-streaming-05.txt

        this uses "headers" twice, both of which can easily be changed
        to "header fields"

http://www.ietf.org/internet-drafts/draft-ietf-nntpext-base-27.txt

        this uses "header" extensively in 3.6 (relating to folding) and
        would require a moderate amount of work :-(  though in my view
        that work is mechanical and should not give rise to debate.

>I move that we make no change to this long established practice.

You can argue that both alternatives are "long established".

The question is what is best going forward. And there the choice appears
to be between creating significant consistency in the set of documents
that future implementors consult; and in doing some mechanical editing
to documents that are not yet set in stone (albeit some are already a
long way towards that state).

If we are still allowed to disrupt the smooth progress of the NNTP
documents then, in my view, that is what we should do.

- -- 
richard @ highwayman . com                       "Nothing seems the same
                          Still you never see the change from day to day
                                And no-one notices the customs slip away"