Re: #1083 USEPRO 7.5: Rules for Generating message-ID

["Charles Lindsey" <chl@xxxxxxxxxxxxxxxx>]

In <FE15FB978B17722F1CB21BDF@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> Harald Tveit Alvestrand <harald@xxxxxxxxxxxxx> writes:

> >> - What needs to be said in order to "play nice" with the $altz
> "control."
> >> convention (for instance, not to outlaw it)

1. There is a clear requirement [USEFOR/RFC 2822] that message identifiers
be globally unique.

2. It is well known that if you deliberately break that requirement you
can bring about a situation whereby two different versions of what is
ostensibly the same article can propagate, but that any given server will
see only one of them.

3. In general, such deliberate breach of the standard would be regarded as
malpractice, meriting LARTing the perpetrator to his ISP. However, in a
few exceptional situations Usenet (more specifically, the consensus view
of the operators of servers) has consciously decided to turn a "blind eye"
to the breach.

In my opinion, we should leave it that way, by remaining resolutely silent
on the issue, certainly within USEFOR and USEPRO. I have reviewed the
earlier thread on this topic, and I find no calls for any specific action.

On looking at the current text of USEAGE, I find that I have written:

   ..... It is also in order to include
   additional information of significance to the poster within the <id-
   left>, and even to deliberately make a non-unique identifier in cases
   where the identical message is to be posted by several posters (for
   example, a cancel for an article which may also be cancelled by
   others).

I now think that maybe I ought to tone that down somewhat. Opinions?

Here is a little more background on the matter:

Usenet is a strange social phenomenon. There Is No Cabal - there is just
the body of server administrators world wide who, de facto, have ultimate
control.

Shortly after the Green Card affair, when a large number of spammers
started to cause widespread disruption of Usenet, Cancelmoose started to
issue mass cancels for spammed articles. Since such cancels were clearly
against all rules of good order on Usenet, he did so anonymously, through
anon.penet.fi, and I am not aware that, even to this day, his true
identity has ever become known. But he was quite open about it, posting
and discussing in news.admin.misc what he was doing.

Rather to everyone's surprise, Usenet as a whole hailed him as a hero.
There was an apparent consensus to turn a "blind eye", and various people
joined in to refine the process, resulting in the $alz convention, the
syberspam convention, and the Briedbart Index. That is a typical example
of how Usenet governs itself. We should leave it that way.

That is the approach I have taken in USEAGE. There are the usual
definitions of 1st, 2nd and 3rd party cancels, and what it says for 3rd
party cancels is:

   3. Other entities MAY be entitled to issue a cancel message for that
      article, in circumstances where established policy for any
      hierarchy or group in the Newsgroups header, or established custom
      within Usenet, so allows (such policies and customs are not
      defined by this document). ........


>>> - Whether, and if so where, the requirement not to go mess with
> other
> >> people's namespace needs to be documented

We are talking here of people who construct <msg-id>s which contain the
names of other people's domains in their <id-right>. Assuming that they
have taken care to choose an <id-left> such that the <msg-id> as a whole
is still globally unique, they are not in breach of USEFOR; nothing will
break as a result, hence they are not in breach of USEPRO. Nevertheless
the practice is, to say the least, extremely rude and probably constitutes
an attempt to throw doubt on the true origin of the article (since
<msg-id>s are one the places people will regularly look when trying to
trace the origin of an article). Therefore, this could be a proper concern
for USEAGE.

Here is a possible text for USEAGE, though I am not at all sure that it is
worth inclusion:

        NOTE: To use, in the <id-right>, a domain name or IP address not
        associated with the originating site could defeat the object of
        achieving global uniqueness and would, moreover, be an affront
        to the true owner of that domain (perhaps leading people to
        believe the article was somehow associated with that other
        domain). Such practice is therefore deprecated.
[possible text for #1083.]

Opinions?

-- 
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@xxxxxxxxxxxxxxxx      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