Re: The Interlinked Cypherpunks Lists? (fwd)

Hi,

> > * Is it possible for the interlinked lists to send out announcements of
> > list problems to subscriber of all such lists? (Each list owner could do a
> > periodic "who cypherpunks" of the other lists, then use this list to send a
> > message if a problem comes up. More sophisticated cross-processing could
> > eliminate duplicates, etc.)

Sounds like time to start another list subscribed to the existing nodes
and offer such services to the subscribers at the operators whim.

I oppose any action which would force the list to become more 'cooperative'
in the sense that each node operator would be forced (sorta anti-cpunks I
believe) to comply with some set of scripts and such they didn't develop in
the first place. It would be just another means to limit the ability of
people in general, and the cpunks in particular, to communicate with as
little structure and 'authority' as possible. Unless I am seriously
mistaken on this, the point was to make the list more robust and less
centralized. Pretty soon we'll have to join some damn union or something.

Now some of you are going to say that this is taking it to extremes. That
what is being proposed is a good thing and not something that could be used
to 'manipulate' the lists. The road to hell is paved with good intentions,
and the guy driving the paving truck only wants to help you (as he paves
your petunia's).

> > * Alternatively, if one of the sites goes down, such as
> > "cypherpunks@cyberpass.net" seems to have done, could the other sites
> > automagically pick up the task of distributing articles until the site
> > comes back up?

This again is anti-cpunks, it forces a level of cooperation and information
sharing that is in direct opposition to beliefs in anonymity and privacy.
I don't want it known who is subscribed to lists through SSZ, in short:
It isn't any of your damn business and quit asking.

Solution? Get a life or start an archive site.

> The only true reliability can be achieved by subscribing to two or more
> cypherpunks sites, and eliminating duplicates. Reliability has its costs.

As do freedom, privacy, anonymity, etc.


                                                  Jim Choate
                                                  CyberTects
                                                  ravage@ssz.com

At 06:50 PM 5/13/97 -0800, Tim May wrote:
>Instead, imagine this "service":
>- a distribution point (= list) which subscribes to all of the various
>lists (cyberpass, algebra, ssz, etc.)
>- it sends out to subscribers the first instance of any message it receives
>- duplicates (see discussion below) would not be sent
>- it would, ideally, be on a robust machine

As a filtering service, it would be great, though most of the duplicates
I see are articles posted on one or more of (cypherpunks, coderpunks,
cryptography), and I'd see a few more dups if I read alt.cypherpunks
regularly.

On the other hand, one of the reasons we have multiple list servers
is to reduce the risk of a single cypherpunks list site dying again
(as well as just the effect of several people volunteering at once.)
A filter site is a single point of failure, though it's a less serious
failure if there are still multiple lists feeding it.
If the filter can set the headers appropriately so that the
author and originating-list information still works, 
then it matters even less if it goes down.


#			Thanks;  Bill
# Bill Stewart, +1-415-442-2215 stewarts@ix.netcom.com
# You can get PGP outside the US at ftp.ox.ac.uk/pub/crypto/pgp
#     (If this is a mailing list, please Cc: me on replies.  Thanks.)

At 4:34 PM -0800 5/13/97, Jim Choate wrote:

>> > * Is it possible for the interlinked lists to send out announcements of
>> > list problems to subscriber of all such lists? (Each list owner could do a
>> > periodic "who cypherpunks" of the other lists, then use this list to
>>send a
>> > message if a problem comes up. More sophisticated cross-processing could
>> > eliminate duplicates, etc.)
>
>Sounds like time to start another list subscribed to the existing nodes
>and offer such services to the subscribers at the operators whim.
>
>I oppose any action which would force the list to become more 'cooperative'
>in the sense that each node operator would be forced (sorta anti-cpunks I
>believe) to comply with some set of scripts and such they didn't develop in

I wasn't advocating either:

a) any "forcing" of anything

b) any compliance with scripts, cooperation, etc.

Instead, imagine this "service":

- a distribution point (= list) which subscribes to all of the various
lists (cyberpass, algebra, ssz, etc.)

- it sends out to subscribers the first instance of any message it receives

- duplicates (see discussion below) would not be sent

- it would, ideally, be on a robust machine

The "duplicates" issue has been discussed by others. Even if message IDs
are not enough to find duplicates--someone reported that the same message
from algebra and cyberpass have different IDs--I would think that using the
sender, message title, and date of origin ought to be more than enough to
spot duplicates.

Thus, the message "Re: The Interlinked Cypherpunks Lists? (fwd), 4:34 PM
-0800 5/13/97, Jim Choate" should be unique.


>Now some of you are going to say that this is taking it to extremes. That
>what is being proposed is a good thing and not something that could be used
>to 'manipulate' the lists. The road to hell is paved with good intentions,
>and the guy driving the paving truck only wants to help you (as he paves
>your petunia's).

No, this is just another filtering service, one designed to collect
messages from as many inputs as possible and send them out, without
duplicates.


>> > * Alternatively, if one of the sites goes down, such as
>> > "cypherpunks@cyberpass.net" seems to have done, could the other sites
>> > automagically pick up the task of distributing articles until the site
>> > comes back up?
>
>This again is anti-cpunks, it forces a level of cooperation and information
>sharing that is in direct opposition to beliefs in anonymity and privacy.
>I don't want it known who is subscribed to lists through SSZ, in short:
>It isn't any of your damn business and quit asking.

No one has ever said you, or SSZ, or anyone, has to participate.


>Solution? Get a life or start an archive site.
>

"Get a life." What an original insult.

Reminds me why you usually reside in my filter file, Jim. Back in it you go.

--TCM

There's something wrong when I'm a felon under an increasing number of laws.
Only one response to the key grabbers is warranted: "Death to Tyrants!"
---------:---------:---------:---------:---------:---------:---------:----
Timothy C. May              | Crypto Anarchy: encryption, digital money,
tcmay@got.net  408-728-0152 | anonymous networks, digital pseudonyms, zero
W.A.S.T.E.: Corralitos, CA  | knowledge, reputations, information markets,
Higher Power: 2^1398269     | black markets, collapse of governments.
"National borders aren't even speed bumps on the information superhighway."

-----BEGIN PGP SIGNED MESSAGE-----


In <v03007805af9ed5fd9760@[207.167.93.63]>, on 05/13/97 at 08:50 PM,
   Tim May <tcmay@got.net> said:


>The "duplicates" issue has been discussed by others. Even if message IDs
>are not enough to find duplicates--someone reported that the same message
>from algebra and cyberpass have different IDs--I would think that using
>the sender, message title, and date of origin ought to be more than enough
>to spot duplicates.

>Thus, the message "Re: The Interlinked Cypherpunks Lists? (fwd), 4:34 PM
>-0800 5/13/97, Jim Choate" should be unique.

I have been doing some work in this area and IMHO the above is not enough
to properly weed out the duplicates. Not that it woun't catch them but it
will also throw out some non duplicate messages. I would recomend that the
criteria be extended to include the seconds in the time stamp (4:34:25). If
the seconds are not available perhaps generating a hash of the messages
body and compairing would work.

I'm not trying to quibble but I have seen in the past where an author will
post a message and there will be multiple replies to the message. The
author will then reply to the replies and several of these messages will
have the only difference will be the seconds on the time stamp.

Awhile back on one of my e-mail client list their was a disscussion on
wether the time stamp of a message should be generated at the time the
message is composed or at the time the message is sent. We had come to the
conclution that the time the message was composed was the best route to
take. Other e-mail clients may have taken the other route which would
increase the chances on non duplicate messages haveing the same hour:min in
the time stamp.


- -- 
- -----------------------------------------------------------
William H. Geiger III  http://www.amaranth.com/~whgiii
Geiger Consulting    Cooking With Warp 4.0

Author of E-Secure - PGP Front End for MR/2 Ice
PGP & MR/2 the only way for secure e-mail.
                          
Finger whgiii@amaranth.com for PGP Key and other info
- -----------------------------------------------------------
 
Tag-O-Matic: OS/2: Your brain.  Windows: Your brain on drugs.

-----BEGIN PGP SIGNATURE-----
Version: 2.6.2
Comment: Registered User E-Secure v1.1 ES000000

iQCVAwUBM3kwS49Co1n+aLhhAQGOtQP+JTDtd65amyJ7dO8vzsKMB6Z3s/2XNp3D
cALj/LF2UsNGpky/3FUWoAK4QNnUZkH75yMs8h9Cy8Z5Qe5DZrmvlHW4fEztzrsG
3fQs9H8qZDMIwMDKYM6YF+VmHsQOLy9tS6YLr7z1H4T6HG/cy0Qa6saYYyM9UDl3
JcvmnevnIyU=
=5Rtx
-----END PGP SIGNATURE-----

Tim May wrote:
> At 4:34 PM -0800 5/13/97, Jim Choate wrote:
> 
> >> > * Is it possible for the interlinked lists to send out announcements of
> >> > list problems to subscriber of all such lists? (Each list owner could do a
> >> > periodic "who cypherpunks" of the other lists, then use this list to
> >>send a
> >> > message if a problem comes up. More sophisticated cross-processing could
> >> > eliminate duplicates, etc.)
> >
> >Sounds like time to start another list subscribed to the existing nodes
> >and offer such services to the subscribers at the operators whim.
> >
> >I oppose any action which would force the list to become more 'cooperative'
> >in the sense that each node operator would be forced (sorta anti-cpunks I
> >believe) to comply with some set of scripts and such they didn't develop in
> 
> I wasn't advocating either:
> 
> a) any "forcing" of anything
> 
> b) any compliance with scripts, cooperation, etc.
> 
> Instead, imagine this "service":
> 
> - a distribution point (= list) which subscribes to all of the various
> lists (cyberpass, algebra, ssz, etc.)
> 
> - it sends out to subscribers the first instance of any message it receives
> 
> - duplicates (see discussion below) would not be sent
> 
> - it would, ideally, be on a robust machine
> 
> The "duplicates" issue has been discussed by others. Even if message IDs
> are not enough to find duplicates--someone reported that the same message
> from algebra and cyberpass have different IDs--I would think that using the
> sender, message title, and date of origin ought to be more than enough to
> spot duplicates.

Tim, what you are proposing is no different from any other cypherpunks
node. So if you suggest to add more nodes, I am all for it.

What you describe above is precisely what algebra, cyberpass, and
ssz.com do now.

It is not a "just another service" as you suggest below, it is
exactly the same service, but on a new node.

Actually, I think that it is important to add one more node to the
network: ssz.com has mail delivery problems and the two more or less
reliable nodes are cyberpass and algebra. I do not see it as *sufficiently*
robust, since the probability of both of them being down is on the
order of 0.5%.

> No, this is just another filtering service, one designed to collect
> messages from as many inputs as possible and send them out, without
> duplicates.

> "Get a life." What an original insult.

> Reminds me why you usually reside in my filter file, Jim. Back in it you go.

Tim, if you read what goes into your filter file, putting anyone there
does not mean that you ignore them.

	- Igor.

Jim Choate <ravage@ssz.com> writes:
> [...]
> I don't want it known who is subscribed to lists through SSZ, in short:
> It isn't any of your damn business and quit asking.

Oh.  I guess that explains the blank message I got back from
majordomo@ssz.com!  I counted that as no one subscribed (or inactive).
So I guess the count of cypherpunks subscribers is an underestimate
having discounted anyone subscribed through ssz.com.

Would you be interested to reveal the number of subscribers for
counting purposes?  (Not a big deal if you're not).

> > The only true reliability can be achieved by subscribing to two or more
> > cypherpunks sites, and eliminating duplicates. Reliability has its costs.
> 
> As do freedom, privacy, anonymity, etc.

If you have a shell account, you could run an email concentrator
(remove duplicates) with procmail, and then down load only the non
duplicates.  With this setup you could subscribe to all 3 lists and
not notice any outages.

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`