[89] in Commercialization & Privatization of the Internet
Re: Is SunFlash misuse of the network?
daemon@ATHENA.MIT.EDU (John Gilmore)
Sun Nov 11 10:09:53 1990
Date: Sun, 11 Nov 90 02:07:09 PST
From: gnu@toad.com (John Gilmore)
To: com-priv@psi.com, sun!sunvice!flash@uunet.UU.NET, gnu@toad.com
The SunFlash mailing list tree was formed by a salesman in their Ft.
Lauderdale, Florida office. I had been beating on Sun Marketing for
years to send their press releases by email (over a private uucp link
between my site and sun), but they wouldn't. Sun Marketing emails them
to every employee, but until SunFlash, not to outsiders. The enterprise
of John McLaughlin, the Ft. Lauderdale salesman, was a godsend to those
Sun customers tired of reading garbled reports in the 'trade press' two
weeks later.
The SunFlash mailing list itself is really a list of exploders at
various sites around the country. My understanding is that typically
these are Sun sales offices (which get the message via the internal Sun
Wide Area Network). On my site, which is fed from the uucp site "sun",
there is a "sunflash" alias that feeds the list both to me, and to a
colleague one uucp hop away; this is typical.
Undoubtedly some messages are leaking out into the Internet. Given
that nobody else basically gives a sh*t about whether their email
correspondence is in support of education or research, why would
someone in a remote sales office of a network company even conceive of
a problem with sending press-releases to "@" addresses? They have no
idea over what networks those addresses are reached.
Sunflash is a poor example, if you want to pick on Sun; they did *try*
to route it internally and then over uucp links.
Let's do a thought experiment.
Let's have them clean up the small percentage of addresses that traverse
the NSFnet. Now, how should they address mail to these people? Or
should the list maintainer simply refuse to send the mailing-list to them?
(This begs the question of how they discover that an addressee receives mail
that traverses the NSFnet, and how they keep up with changes in that status,
but let's just deal with the issue of what alternatives they have.)
They could set up a uucp link with uunet, which is authorized to
forward mail between Internet and uucp sites. This seems pretty stupid
since Sun itself is already on the Internet, but hey, rules are rules.
They could set up a link to Compuserve, and Compuserve would gateway
the messages through the Internet gateway. They could force each site
to set up a uucp link (if it's a Unix site). For a few addressees,
there might be a twisted addressing syntax that would cause the message
to go via back-roads which manage to avoid the NSFnet (and any
regionals that lack freedom of speech). Does anyone have problems with
these possible fixes, or have other suggested fixes?
John
