[Mailman-Users] GMane?

Stephen J. Turnbull stephen at xemacs.org
Fri Feb 17 13:45:58 CET 2006

>>>>> "jam" == John A Martin <jam at jamux.com> writes:

    sjt> I missed at least one of your posts, receiving Brad's reply
    sjt> to it almost 24 hours in advance of your post.  Even today
    sjt> this is common for netnews.

    jam> Sorry, Gmane is not netnews.  Gmane is not Usenet.

OK, so that's not due to Gmane, it's due to the fact that you have
been using Gmane.  Different mechanism, same result.

    jam> I am at a loss however to understand any of this being
    jam> attributable to or in any way related to netnews or to Gmane.
    jam> Perhaps you would be kind enough to show the evidence upon
    jam> which you base this connection?

The circumstantial evidence.  Be thankful nobody's in a hanging mood. :-)
(Yes, I know about the "post hoc" fallacy.)

    sjt> Gmane started its service for no apparent reason without
    sjt> notifying anyone,

    jam> Lars should have consulted you in 2002 before offering a
    jam> mailing list archive to The Net, right?

Of any lists I administer, yes.  I was a bit peeved when I found out
months after the fact that they had been subscribed, despite having no
objection in principle.

    sjt> they stopped it for no apparent reason

    jam> ?!?

    jam> You mean Gmane stopped carrying the mailman-* lists?


    jam> Lars said, "The mailman people requested that they be
    jam> removed", see <http://article.gmane.org/gmane.discuss/9291>.

Brad says he didn't, at least not in the past week.  There are others
who can speak for Mailman, but Lars apparently doesn't see the need to
identify who he's listening to.  So maybe he acted on the rumor that
Brad was _going_ to ask him.  Is that any way to run a railroad?

    sjt> without notifying anyone.

    jam> See <http://article.gmane.org/gmane.discuss/9293>.

No, I'm not going to look, and you're way out of bounds for suggesting
that anybody *should* look.  It's the other way around; those who want
the Gmane service should take it up here (more precisely, with
mailman-users-owner at python.org).

Also, Gmane should get confirmation from the list owner whenever it
subscribes or unsubscribes, for exactly the same reason that Mailman
gets confirmation from a user when it subscribes or unsubscribes.

    sjt> They have a history of being an attractive nuisance
    sjt> (publishing email addresses and other spam-facilitating
    sjt> activities).  It's fundamentally irresponsible, but that is
    sjt> the way they operate.

    jam> Did you have a similar view of Altavista ten years or so ago?

Yes and no.  Yes; of course I support the robots.txt protocol, and
that is because there were a number of incidents of undesired

On the other hand, no.  First, at the time there was no protocol for
saying "don't index".  This isn't true for finding list owners, the
informal use of LIST-owner and/or LIST-admin goes back to early
majordomo at least, maybe back to UCSD listserv.

Second, Altavista et al simply link to what is publically available
anyway, which is part of the advertised functioning of the web, and is
a practice that goes back to the earliest days of writing.  Gmane on
the other hand copies and retransmits (specifically forbidden by
copyright law, unless you have explicit permission from the copyright
holder) and can also _change_ what is publically available on the web.
The potential problems with those behaviors have been known since
ancient times, too (which is part of why copyright exists, and and
most of why many jurisdictions have inalienable "author's rights" as

Third, if third parties subscribe a list to Gmane, the list-owner's
membership-management prerogatives are usurped.  This is not true of

    jam> List owners (or more accurately, whoever subscribes a list to
    jam> Gmane) have a number of choices (Posting allowed, Read only,
    jam> List member only posting, No posting through Gmane, Encrypt
    jam> addresses, Spam tagging, and more).  ISTM choosing the
    jam> appropriate posting option and encrypted address would go a
    jam> long way toward reducing the nuisances.

I agree 100%.  The problem is your parenthetical remark.  The Mailman
admins are on the record as not wanting these lists on Gmane.  I think
it reasonable to suppose they did not subscribe Gmane to the lists.
Therefore somebody else did, and since they are not the admins, they
are far too likely to be careless about such nuances.

    jam> Gmane is a Free Public Service.  Like the man said, "If it
    jam> cannot be abused, it is not free".  Such is the world.

    jam> Blame Gmane because bad people use it. :)

No, I blame Gmane because they don't care enough whether it is used
for "bad" purposes.  If I were running Gmane, the subscription process
would only be open to list owners.  Evidently it is not so restricted.
In other words, Gmane is asking those who do _not_ want to be
subscribed to correct any mistakes and bear the burden of any harm

    jam> I have suggested to list owners that sooner or later someone
    jam> will likely subscribe their list to Gmane.  This is because
    jam> it is there and there are folks that like to use it.

This is only possible because Gmane does not follow its stated policy
of checking with list owners.  (I'm taking Brad's word on the stated
policy; I know that they don't check with list owners because they
didn't check for any of those I administer.)

    sjt> the Mailman list admins should be free not to use Gmane, or
    sjt> to require Gmane improvements as a condition of using Gmane,

    jam> It is not as if Lars or his minions are trying to persuade
    jam> anyone in particular to subscribe a particular list.  List
    jam> subscribers are likely sooner or later to try to subscribe a
    jam> list, or persuade a list owner to do so, unless steps are
    jam> taken to prevent it.

Exactly.  Gmane should take steps to prevent non-owners from trying.
I welcome your attempt to persuade the Mailman list admins to allow
Gmane to subscribe and gateway the lists.  I think that should be the
*only* sanctioned way to do it.

    sjt> The fact that Gmane *re*subscribed to Mailman lists in
    sjt> violation of both their own policy and a previous request to
    sjt> cease and desist speaks volumes for the risks and their lack
    sjt> of respect for others' privacy, IMHO.

    jam> Do you really think anyone at Gmane subscribed the mailman-*
    jam> lists?  Don't you imagine that someone, probably a
    jam> subscriber, subscribed the list?

Of course.  Under the legal doctrine of "attractive nuisance", you are
legally responsible for harm caused because you don't warn people
about a hazard caused by property you own.

Gmane is an attractive nuisance, and IMHO IANAL morally responsible
for this kind of "mistake", since their "policy" makes no effective
attempt to discourage it.

School of Systems and Information Engineering http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
               Ask not how you can "do" free software business;
              ask what your business can "do for" free software.

More information about the Mailman-Users mailing list