eternity-lite & cancel msgs

One issue with the system is that censors will issue cancel messages
for the articles.  Not too much which can be done about this.  Some
trends tend to help this problem:

Cancel message abuse by people running cancel bots reflecting their
censorous views have resulted in some sites ignoring cancel messages.
PGPmoose provides a better method of allowing only the author to
cancel.  Or NoCEM provides a ratings system framework, which is
superior to using cancels.

Another aspect of the system is that it relies on news archival
services such as dejanews and altavista; these presumably don't listen
to cancel messages for already archived data (? guessing here).  The
archive maintainers presumably don't want to get involved in disputes
over which old articles should be purged from their archive anyway.

Adam
-- 
Have *you* exported RSA today? --> http://www.dcs.ex.ac.uk/~aba/rsa/

print pack"C*",split/\D+/,`echo "16iII*o\U@{$/=$z;[(pop,pop,unpack"H*",<>
)]}\EsMsKsN0[lN*1lK[d2%Sa2/d0<X+d*lMLa^*lN%0]dsXx++lMlN/dsM0<J]dsJxp"|dc`

Adam Back <aba@dcs.ex.ac.uk> writes:

> One issue with the system is that censors will issue cancel messages
> for the articles.  Not too much which can be done about this.  Some
> trends tend to help this problem:

The vast majority of cancels in the "control" newsgroup are the so-called
"third party cancels" (a.k.a. forgeries).

In the U.S. many major ISPs like America Online and Earthlink are ignoring
cancels outright.  It's easy - start the latest version of INN with the
-C flag, or apply Dave Hayes's patches at http://www.jetcafe.org/~dave/usenet.

> Cancel message abuse by people running cancel bots reflecting their
> censorous views have resulted in some sites ignoring cancel messages.

Quite a few, I might add.  The only reason why Demon processes cancels is
that Ade Lovett is a homosexual who wants to be able to censor "homophobic"
Usenet articles. :-)

Forging cancels is easy -- see the cancelbot I posted last year.

> Another aspect of the system is that it relies on news archival
> services such as dejanews and altavista; these presumably don't listen
> to cancel messages for already archived data (? guessing here).  The
> archive maintainers presumably don't want to get involved in disputes
> over which old articles should be purged from their archive anyway.

Correct: none of the 3 major archival services (dejanews, altavista,
reference.com) delte articles based on cancel or supersedes once they
receive them.

However an author can explicitly ask dejanews to delete his articles.

All 3 now don't archive articles that say 'X-No-Archive: yes'.

---

Dr.Dimitri Vulis KOTM
Brighton Beach Boardwalk BBS, Forest Hills, N.Y.: +1-718-261-2013, 14.4Kbps

Dimitri Vulis <dlv@bwalk.dm.com> writes:
> Adam Back <aba@dcs.ex.ac.uk> writes:
> 
> > One issue with the system is that censors will issue cancel messages
> > for the articles.  Not too much which can be done about this.  Some
> > trends tend to help this problem:
> 
> The vast majority of cancels in the "control" newsgroup are the so-called
> "third party cancels" (a.k.a. forgeries).
> 
> In the U.S. many major ISPs like America Online and Earthlink are ignoring
> cancels outright.  It's easy - start the latest version of INN with the
> -C flag, or apply Dave Hayes's patches [...]
> 
> Forging cancels is easy -- see the cancelbot I posted last year.

Seems your cancel bot has been a net boon to freedom of speech,
encouraging cancel message abuse which in turn leads to the natural
solution of removing the ability to cancel other peoples articles.
Perhaps your intention :-)

Perhaps a content aware version would improve on your content blind
one, making it easy for censorous persons to make more of a nuisance
of themselves without having to have enough of a clue to figure it out
for themselves.  Perhaps a slick windows interface is in order :-)
Select your prejudices from this pick list.  Just to help accelerate
the move away from honoring cancel requests.

> > Another aspect of the system is that it relies on news archival
> > services such as dejanews and altavista; these presumably don't listen
> > to cancel messages for already archived data (? guessing here).
>
> Correct: none of the 3 major archival services (dejanews, altavista,
> reference.com) delte articles based on cancel or supersedes once they
> receive them.
> 
> However an author can explicitly ask dejanews to delete his articles.
> 
> All 3 now don't archive articles that say 'X-No-Archive: yes'.

Does this imply that you are aware of precedents where the search
engines maintainers have removed articles on email requests after the
article was initially posting without X-No-Archive: yes?  Or were you
just refering to the X-No-Archive mechanism?

A bad precedent to act on instructions of someone purporting to be the
author requesting deletion.  Opens the way for someone to socially
engineer deletion of articles they wish to censor if this is so.  Also
opens the way for the author to be coerced into removing the article.

As an aside I think X-No-Archive is a stupid idea.  If you don't want
a post released for anyone to save for later reading, don't post it.
If you are concerned about the correlation of your net posts with your
physical identity, post using a remailer, or nymserver.

In fact I'd encourage anyone with the hardware to spare to offer an
X-No-Archive ignoring search engine.  Move it off-shore if you get
hassled.  Perhaps an X-No-Archive: yes only archive would be the best
way.  (ie only archive stuff which is specifically marked as not for
archive, that should give pretty low bandwidth requirements and make
the point).

Adam
-- 
Have *you* exported RSA today? --> http://www.dcs.ex.ac.uk/~aba/rsa/

print pack"C*",split/\D+/,`echo "16iII*o\U@{$/=$z;[(pop,pop,unpack"H*",<>
)]}\EsMsKsN0[lN*1lK[d2%Sa2/d0<X+d*lMLa^*lN%0]dsXx++lMlN/dsM0<J]dsJxp"|dc`