December 2003 - cypherpunks-legacy

Re: Cypherpunks Distributed Remailer (fwd)
by E. Allen Smith 17 Dec '03

17 Dec '03

From: IN%"ravage(a)EINSTEIN.ssz.com" "Jim Choate" 12-FEB-1997 23:19:40.35 >> From: "E. Allen Smith" <EALLENSMITH(a)ocelot.rutgers.edu> >> Someone with more experience correct me if I'm wrong (I'm just >> starting to learn about mailing list management), but isn't saving >> at least the first bounce from a given address good in order to check >> what address is the one that's actually bouncing? Admittedly, this only >> applies to badly-formed bounces. The rest can just have the address >> noted. >Hmmm, I can see saving the address causing the bounce in some sort of array >that is indexed to the majordomo subscription list for a given remailer. It >would not cost anything but overhead to save the entire message. I guess my >motivation was not that I cared their mailbox was full or it was a invalid >address but that I wouldn't be able to deliver at this time. I just don't >see any reason that it helps me to know why they can't receive mail only that >they can't. Well, if you can't deliver mail because their mailbox is full, then you should simply wait a few days before trying to deliver mail to that address. If it's because the address isn't there any more, you should delete it from the subscription lists. Otherwise, the instant people's mail quotas overflow (e.g., they're on AOL and they haven't been able to get through the clogged lines), they get bounced... or you're left with a lot of invalid addresses. The address causing the bounce is not necessarily determinable automatically from the bounce message; if it isn't, then you need to keep a copy around for a human to look at. [re: bi-directional news-mail gateways] >> >Want to volunteer? >> >> Give me some idea of how much programming knowledge, time, etcetera >> is required, please, before I can answer. I'd be renting space on some other >> computer (e.g., an account on cyberpass.net) to do it in; the only higher-level >> programming language (i.e., beyond Applesoft Basic) that I know anything of >> is Perl (and I don't know that all that well); and I am somewhat busy with >> other stuff. In other words, it depends. >Hmmm, not shure how much programming it would take - never done it myself. >I think Perl would be an acceptable language. I certainly hope to do the >majority of the scripts on the CDR in Perl myself. Perhaps there is somebody >who knows the fine points of how INN cache's messages. Redbeard and I had a >short discussion about this issue today relating to tracking messages to >reduce mail-loops. I will discuss it with him and see what happens. Given that there are pre-existing news-mail gateways, I'd think that the programs should be available. My concern on the programming is more in the realm of keeping the thing up and running. >I know exactly what you mean. I have two full-time jobs and hadn't really >thought of taking on the cpunks list in any manner. But hey, that is the >spice in life... Understood; my busy-ness with other stuff comes from being a grad student. >Give me a couple of days to look at the situation and if somebody don't beat >me to it I'll post what I find out. Cool? Looks OK. I'll also take a look around for what information I can find. -Allen

1 0

Re: distributed mailing list architecture
by Thomas Roessler 17 Dec '03

17 Dec '03

Maybe you have some use for the appended perl script. I created it for a list setup quite similar to what is currently being discussed on cypherpunks. It has never ever been used in production; you'll notice yourself that the code is not too nice. The script currently tries to handle majordomo and SmartList exploders; at least SmartList needs to be hacked a little bit to avoid daily subscription approvals to people being moved between different sub-lists. Distribute and use this script freely; credit is appreciated. tlr ------------------------------ #!/usr/bin/perl ############################################################ # # $Id: distlist.pl,v 1.2 1997/01/24 17:20:38 roessler Exp $ # # # Handle distributed mailing lists. # require 'getopts.pl'; $c_sublists='lists'; # Data about the sublists $c_distfile='dist'; # A list of mail addresses $c_datafile='dist.data'; # _Our_ list of addresses $c_contact='roessler(a)sobolev.rhein.de'; # Whom to contact in case of problems $sendmail="| /usr/sbin/sendmail -t -odq"; $debug=0; # Debugging output. Can also be # turned on by using the -x switch. $signature="-- \nThis mail has been generated automatically by the distlist\n". "program. Please contact $c_contact in case of problems.\n"; @sublists=(); @newsubs=(); # # Command line processing # &dprint (1, 'Processing the command line...\n'); &Getopts('s:d:D:c:x:'); $c_sublists=$opt_s if $opt_s; $c_distfile=$opt_d if $opt_d; $c_datafile=$opt_D if $opt_D; $c_contact=$opt_c if $opt_c; $debug=$opt_x if $opt_x; # # Read the data and config files, perform various checks. # # # sublists # &dprint(1, "Reading $c_sublists..."); # format: addr:maxsubscr:listtype:password:adminrequest:maintainer open(SUBLISTS, $c_sublists) || die("Can't open $c_sublists"); SUBLIST: while(<SUBLISTS>) { next SUBLIST if /^#/ || /^$/; chop; ($addr, $maxsubscr, $listtype, $password, $admin, $maintainer) = split(/:/, $_); $s_maxsubscr{$addr}=$maxsubscr; $s_listtype{$addr}=$listtype; $s_password{$addr}=$password; $s_admin{$addr}=$admin; $s_maintainer{$addr}=$maintainer; $s_onlist{$addr}=0; $s_addem{$addr}=""; $s_delem{$addr}=""; &dprint(3,"Parsed list $addr: maximum $maxsubscr people; listtype $listtype; password $password.; administrative address $admin; maintainer $maintainer"); @sublists=(@sublists, $addr); } close(SUBLISTS); # In addition, we define a special list for people to be redistributed later. $s_maxsubscr{'later'}=100000; # Very big. ;) $s_listtype{'later'}=""; $s_password{'later'}=""; $s_onlist{'later'}=0; $s_addem{'later'}=""; $s_delem{'later'}=""; # # Read our dist file # # format: user:list if(open(DATAFILE, $c_datafile)) { &dprint(1, "Reading $c_datafile..."); USERLINE: while(<DATAFILE>) { next USERLINE if /^#/ || /^$/; chop; ($user, $list) = split(/:/, $_); # Check if the list exists; handle user. if(! $s_maxsubscr{$list}) { &dprint(2, "While checking user $user: $list does no longer exist."); $list="later"; } elsif( $s_onlist{$list} >= $s_maxsubscr{$list} ) { &dprint(2, "Warning: $list is full. Redistributing people later."); $s_delem{$list}=join(':',$user,$s_delem{$list}); $list="later"; } $s_onlist{$list}++; $u_list{$user}=$list; $deletem{$user}=$user; } close(DATAFILE); &dprint(1,"$c_datafile finished."); } else { &dprint(1, "Warning: Can't read $c_datafile."); } # # Now, read the real distribution file. # &dprint(1, "Reading $c_distfile..."); open(DISTFILE, $c_distfile) || die("Can't open $c_distfile"); DISTLINE: while(<DISTFILE>) { next DISTLINE if /^#/ || /^$/ || /^\(/; chop; if(!$u_list{$_}) { # A new member. &dprint(2, "Found a new member: $_."); @newsubs=($_, @newsubs); } else { # We know him. &dprint(3, "Well-known: $_."); delete $deletem{$_}; } } close(DISTFILE); # # Handle deletions. # foreach $user (keys %deletem) { $list=$u_list{$user}; delete $u_list{$user}; $s_onlist{$list}--; $s_delem{$list}=join(':', $s_delem{$list}, $user); &dprint(3, "Removing $user from sublist $list."); } # # Handle postponed subscriptions: The later list. # foreach (keys %u_list) { next unless $u_list{$_} eq "later"; @newsubs=($_, @newsubs); delete $u_list{$_}; &dprint(3, "Adding $_ to the list of new subscriptions. Was postponed."); } &dprint(1, "Distributing new subscriptions..."); NEWSUBS: foreach $user (@newsubs) { $avg=0.0; foreach $l (@sublists) { $avg += $s_onlist{$l}/$s_maxsubscr{$l}; } $avg = $avg / ($#sublists + 1.0); if ($avg >= 1) { &dprint(2, "Warning: All sublists are full while trying to insert $user."); } undef $possible; foreach $l (@sublists) { if($s_onlist{$l} <= $avg*$s_maxsubscr{$l}) { $possible=$l; } if($s_onlist{$l} < $avg*$s_maxsubscr{$l}) { last; } } if($possible) { $l = $possible; } $s_addem{$l}=join(':', $s_addem{$l}, $user); $s_onlist{$l}++; $u_list{$user}=$l; } # # Write our own data file. # if(open(DATAFILE, ">$c_datafile")) { foreach (keys %u_list) { printf DATAFILE "%s:%s\n", $_, $u_list{$_}; } close(DATAFILE); } else { &dprint(1, "Warning: Can't write $c_datafile."); } # # The lists have been put together. Commit the changes. # &dprint(1, "Committing the changes..."); foreach $l (@sublists) { if($s_listtype{$l} eq "majordomo") { &commit_majordomo($l); } elsif ($s_listtype{$l} eq "smartlist") { &commit_smartlist($l); } else { &dprint(1, "While trying to commit changes for $l:"); &dprint(1, "Unknown list type $s_listtype{$l}.\n"); } } # # To be done: Print out some statistics. # print "Distlist results:\n"; print "-----------------\n"; print "\n"; printf "There are currently %d subscribers on %d sublists.\n\n", scalar(keys %u_list), $#sublists+1; $full_lists=0; printf "%-40s on max\n", "Name"; printf "----------------------------------------------------\n"; foreach $l (@sublists) { printf "%-40s %4d %4d", $l, $s_onlist{$l}, $s_maxsubscr{$l}; if($s_onlist{$l} >= $s_maxsubscr{$l}) { print " *** This list is full ***"; $full_lists++; } print "\n"; } if($full_lists) { print "\n$full_lists of the sublists are *full*. Please get in touch\n"; print "with the maintainers.\n"; } print "\n\n"; print $signature; ############################################################ # # Some helper functions. # # Print out diagnostics. sub dprint { if($_[0] <= $debug) { printf STDERR "%s\n", $_[1]; } } sub commit_majordomo { my $list=$_[0]; my @bla; my $ll; my $user; &dprint(2, "Committing changes to majordomo list $list"); &dprint(3, "to be added: $s_addem{$list}"); &dprint(3, "to be deleted: $s_delem{$list}"); ($ll, @bla)=split(/@/, $list); if(!open(SENDMAIL, $sendmail)) { &dprint(1, "Warning: Can't start sendmail when committing changes to $list"); } print SENDMAIL "To: $s_admin{$list}\n"; print SENDMAIL "From: $c_contact\n"; print SENDMAIL "\n\n"; foreach $user (@bla=split(/:/, $s_addem{$list})) { print SENDMAIL "approve $s_password{$list} subscribe $ll $user\n" if $u_list{$user}; } foreach $user (@bla=split(/:/, $s_delem{$list})) { print SENDMAIL "approve $s_password{$list} unsubscribe $ll $user\n" unless $u_list{$user} eq $list; } print SENDMAIL $signature; close(SENDMAIL); } sub smartlist_xcommand { my ($list, $to, $passwd, $command, $maintainer) = @_; if(!open(SENDMAIL, $sendmail)) { &dprint(1, "Can't start sendmail when committing changes for $list"); } print SENDMAIL "To: $to\n"; print SENDMAIL "From: $c_contact\n"; print SENDMAIL "X-Command: $maintainer $passwd $command\n"; print SENDMAIL "\n\n"; print SENDMAIL $signature; close SENDMAIL; } sub commit_smartlist { my $list=$_[0]; my @bla; my $user; &dprint(2, "Committing changes to smartlist list $list"); foreach $user (@bla=split(/:/, $s_addem{$list})) { &smartlist_xcommand($list, $s_admin{$list}, $s_password{$list}, "subscribe $user", $s_maintainer{$list}) if $u_list{$user}; } foreach $user (@bla=split(/:/, $s_delem{$list})) { &smartlist_xcommand($list, $s_admin{$list}, $s_password{$list}, "unsubscribe $user", $s_maintainer{$list}) unless $u_list{$user} eq $list; } }

1 0

Re: Cypherpunks-announce? Do we need to recreate this?
by paul＠fatmans.demon.co.uk 17 Dec '03

17 Dec '03

> Now that cypherpunks is moving, to both a newsgroup and one or more > mailing lists, there's still a need for cypherpunks-announce or > something like it. So far it's carried 1-2 announcements per month. > > Should we create it as a mailing list, moderated or unmoderated, > or as a newsgroup? (It's been a moderated mailing list, run by John and > Hugh.) I newsgrouped several sub groups along with alt.cypherpunks. alt.cypherpunks.announce alt.cypherpunks.social alt.cypherpunks.technical but they do not appear to have taken effect (my local news server refuses to see them). Can anyone else see them? - If not I`ll have a word with my newsadmin. Datacomms Technologies web authoring and data security Paul Bradley, Paul(a)fatmans.demon.co.uk Paul(a)crypto.uk.eu.org, Paul(a)cryptography.uk.eu.org Http://www.cryptography.home.ml.org/ Email for PGP public key, ID: 5BBFAEB1 "Don`t forget to mount a scratch monkey"

1 0

Re: distributed mailing list architecture
by Peter Hendrickson 17 Dec '03

17 Dec '03

At 12:11 PM 2/7/1997, Adam Back wrote: >Igor Chudov <ichudov(a)algebra.com> writes: >> I'd suggest a simplier solution: to connect each server with a couple, >> or maybe three, other servers. This scheme is rather robust, does not >> consume too much CPU time and bandwidth, and is easy to implement. > I'm not sure what the architecture you are suggesting is, but this is > what I suggest as the simplest to set up. > Have one main majordomo. > Have many mail-exploders. > You subscribe to the main majordomo request address, and it forwards > your subscription request to a random mail-exploder. > You unsubscribe to the main majordomo request address, and it forwards > your subscription to all the mail-exploders request addresses > (unsubscribe traffic is low anyway, keeping track of who is subscribed > where at the main major domo doesn't seem worth it). > Each person who wishes to run an exploder is subscribed (manually) to > the main majordomo. > You submit articles to the main majordomo, and it sends copies of the > articles to it's subscribers (the mail-exploders). > The mail-exploders send mail to the address on their subscriber lists. > (John Gilmore suggested this architecture, as a simpler alternative). This mostly solves the performance problem, but performance was not the only choke point. It is important to have many machines which accept posts and which send them to other machines. (Also, many interconnections means that few messages have to go through more than two machines which should make message distribution extraordinarily fast and highly redundant. Standing in the way of the cypherpunks list will be like standing in the way of the wind.) I suggest we consider the problem of subscriptions to be independent of the problem of posting and distributing messages. Some machine owners may wish to pool subscription activities. The fast and easy approach is to have an official cypherpunks subscription page which lists all of the mail servers which will accept new subscribers. (The person who maintains this page would ideally not be somebody running a mail server.) The interested person looks at the page and then picks a majordomo to subscribe to. It seems beneficial to me from a technical and social standpoint to have a web of relationships between list managers and subscribers. If as a subscriber you have some sort of problem (technical or social), it will work better if there is just one person to talk to rather than a loose network of volunteers. Peter

1 0

Re: anonymity and e-cash
by Sean Roach 17 Dec '03

17 Dec '03

At 10:28 PM 2/12/97 -0800, Hal Finney wrote: >From: Sean Roach <roach_s(a)alph.swosu.edu> ... >> Another idea that has been festering. If we could get a CPA involved in >> this forum, I would be willing to have h[im/er] sign my key, (which is >> seldom used, mostly because this is the only place I use e-mail), for that >> reasonable fee that CPA's can charge. I know that it is not a standard, or >> even legally recognized, post for CPA's, but I think that enough people >> would trust them. > >The big question with identity certificates is what procedures were followed >in verifying the identity when the cert was issued. If the CPA publishes >some standard method, and his reputation is strong enough that people will >trust him to follow it, then it might well be worth money to you to have >him sign it. This is the traditional role of the Certification Authority. > >> This would take care of some of the "newly minted" key problems. Since >> getting someone who is trusted to sign your key is a recognized method of >> getting people to believe you are who you say you are. > >It depends on the circumstances where you expect to use your key. Within >a small to medium circle of associates there may be some group members >who sign keys and are trusted by other members of the group. There is >no particular reason for it to be difficult. If you want a signature which >will be accepted by everyone in the U.S. you have a harder problem. As far as I am aware, a CPA is supposed to require a photo ID. Since this is the method by which absentee voting is accomplished. A supposedly anonymous system, (using two envelopes, the postmaster only knows that you voted, the vote counter only knows that it came by mail, the public only knows how many people are absentee voters, and who, the voter assumes that the system works), requiring the CPA as a witness to the identity of the voter. The problem would be getting the public to recognize the CPA as a key signing authority. This would presumably require a public announcement, the backing of existing trusted users, etc. Personally, I think that the netherlands? toying around with the post office would be ideal for keys used solely for signing. I just wouldn't trust the government to have access to the one I used for encryption. Maybe that's a better idea. Let's all use a government accessable key for signing, making it as long as possible to reduce identity theft. And use homebrew, super encryption for the actual secrecy. Let GAK be our CPA.

1 0

N.Y. judge rules for Planned Parenthood in web site case
by Vipul Ved Prakash 17 Dec '03

17 Dec '03

*** N.Y. judge rules for Planned Parenthood in web site case A federal judge has temporarily stopped an abortion foe from identifying his web site as belonging to Planned Parenthood and selling an anti-abortion book from that home page. The temporary restraining order, issued Wednesday, stops Richard Bucci from identifying his home page as "Planned Parenthood's Home Page" and using "plannedparenthood.com" as a domain name. Planned Parenthood, which has its own web site, sued Bucci for trademark infringement. The next hearing is slated for Feb. 20. For the full text story, see http://www.merc.com/stories/cgi/story.cgi?id=1426121-b7e -- Vipul Ved Prakash | - Electronic Security & Crypto vipul(a)pobox.com | - Web Objects 91 11 2233328 | - PERL Development 198 Madhuban IP Extension | - Linux & Open Systems Delhi, INDIA 110 092 | - Networked Multimedia

1 0

Re: alt.cypherpunks
by John Pearson 17 Dec '03

17 Dec '03

> Hey, folks, > > I thought all you had to do to create an alt.group was to have a news admin > somewhere just make a group; send out a control message of somekind. I > thought that alt groups didn't have to go through a charter, or voting > process, or anything else. They just happened. > > Then, everyone just has to tell their local news admin that they went to > see it, or they can wait until their news admin sees it flying by, and adds > it to the available groups list at his own discretion. > > It's not like we want comp.cypherpunks or something, with a voting process, > right? > It's not quite that simple. While a "Big 7" group requires CFD, charter and voting, once you pass tthose hurdles you're virtually guaranteed good propagation. alt has no "set" creartion procedure, but your group will not receive good propagation unless the powers that be "approve" of the idea. In this case, the people you have to convince are the news admins who must permit your control message to be honoured; time was most sites ran on autopilot, but joke and badly-named groups mean that most news admins now drop most newgroup messages on the floor. These people hang out in alt.config, and you should propose the group's creation there, followed by a couple of weeks of discussion, hopefully featuring many more supportive messages than non-supportive ones. After that, a newgroup message stands a much better chance. > Cheers, > Bob Hettinga > > > Hip hip, John P. john(a)huiac.apana.org.au

1 0

Re: Excerpt on SPAM from Edupage, 11 February 1997
by Sean Roach 17 Dec '03

17 Dec '03

At 09:23 PM 2/12/97 -0800, jim bell wrote: ... >I decided long ago (okay, well, many months ago) that the "solution" is to >invent a mechanism to allow spammers/advertisers to include a small amount >of ecash as a gift with every spam. I figure that if USnail junk-mailers >are willing to pay $0.32 for postage and probably $0.50 for production, >printing, and labelling costs, all for no guarantees of results, they should >even more happy to pay, say, 10 cents to each recipient. At that rate, an average >person would probably receive enough "spam" to pay for his Internet >account, quite analogous to the way advertiser-supported TV is presented to >the public for no explicit charge. First of all, they don't spend 32 cents per post. They get one of those bulk-mail permits which allows them to send those at a significant savings. The whole package complete with the cheap "Yes, I'll try your service, and send me the free gizmo"/"No, I don't need any more services, but I could use the free gizmo" stickers probably costs well within 50 cents. TV is not a gift per say, though your origional posts, involving a cookie, suggested something closer to TV. Currently, advertiser subsidized services, like hotmail, Yahoo, and similair, are closer in that you get the gift with only the minor distraction of the advertisement. A more functional system might be to include a cryptographic "key" of only a few bytes in the e-mail post that opens some online service of relative value. Along the lines of a hidden, no URL FTP site with all the best downloads. This would not be dependant upon e-cash being in place. Would require that the customer read at least part of the message in order to get the instructions on how to use the key. Prevent the user from just handing out the key to whoever asked as it would probably have a limited lifespan. And give the sender some location from which to conduct polling, headcounts, etc.

1 0

Re: Cypherpunks Distributed Remailer (fwd)
by E. Allen Smith 17 Dec '03

17 Dec '03

From: IN%"ravage(a)EINSTEIN.ssz.com" "Jim Choate" 12-FEB-1997 23:19:40.35 To: IN%"cypherpunks(a)toad.com" CC: Subj: RE: Cypherpunks Distributed Remailer (fwd) Received: from toad.com by mbcl.rutgers.edu (PMDF #12194) id <01IFCGV6E7PC94GG0G(a)mbcl.rutgers.edu>; Wed, 12 Feb 1997 23:18 EDT Received: (from majordom@localhost) by toad.com (8.7.5/8.7.3) id RAA13474; Wed, 12 Feb 1997 17:59:02 -0800 (PST) Date: Wed, 12 Feb 1997 18:59:05 -0600 (CST) From: Jim Choate <ravage(a)EINSTEIN.ssz.com> Subject: RE: Cypherpunks Distributed Remailer (fwd) Sender: owner-cypherpunks(a)toad.com To: cypherpunks(a)toad.com Message-id: <199702130159.RAA13474(a)toad.com> X-Envelope-to: EALLENSMITH Precedence: bulk Forwarded message: > Date: Wed, 12 Feb 1997 18:47 EDT > From: "E. Allen Smith" <EALLENSMITH(a)ocelot.rutgers.edu> > Subject: Re: Cypherpunks Distributed Remailer (fwd) > > Someone with more experience correct me if I'm wrong (I'm just > starting to learn about mailing list management), but isn't saving > at least the first bounce from a given address good in order to check > what address is the one that's actually bouncing? Admittedly, this only > applies to badly-formed bounces. The rest can just have the address > noted. Hmmm, I can see saving the address causing the bounce in some sort of array that is indexed to the majordomo subscription list for a given remailer. It would not cost anything but overhead to save the entire message. I guess my motivation was not that I cared their mailbox was full or it was a invalid address but that I wouldn't be able to deliver at this time. I just don't see any reason that it helps me to know why they can't receive mail only that they can't. [re: bi-directional news-mail gateways] >> >Want to volunteer? >> >> Give me some idea of how much programming knowledge, time, etcetera >> is required, please, before I can answer. I'd be renting space on some other >> computer (e.g., an account on cyberpass.net) to do it in; the only higher-level >> programming language (i.e., beyond Applesoft Basic) that I know anything of >> is Perl (and I don't know that all that well); and I am somewhat busy with >> other stuff. In other words, it depends. >Hmmm, not shure how much programming it would take - never done it myself. >I think Perl would be an acceptable language. I certainly hope to do the >majority of the scripts on the CDR in Perl myself. Perhaps there is somebody >who knows the fine points of how INN cache's messages. Redbeard and I had a >short discussion about this issue today relating to tracking messages to >reduce mail-loops. I will discuss it with him and see what happens. So far as I can tell, there are two types of news2mail gateways in common usage: newsgate, which is in C, and the standard LISTSERV bit.listserv.* gatewaying. I don't know C and don't have that much reason to learn it (perl makes a lot more sense to me); therefore, I'd need to locate a perl version or let someone else take care of the actual programming end of it, with me just handling what the programming person didn't. Hmm... I wonder if Lance would be available for the programming? I'd probably be using an account on cyberpass.net for the gateway in any event. -Allen

1 0

distributed mailing list architecture (fwd)
by Jim Choate 17 Dec '03

17 Dec '03

Forwarded message: > Date: Fri, 7 Feb 1997 20:11:15 GMT > From: Adam Back <aba(a)dcs.ex.ac.uk> > Subject: distributed mailing list architecture > Igor Chudov <ichudov(a)algebra.com> writes: > > I'd suggest a simplier solution: to connect each server with a couple, > > or maybe three, other servers. This scheme is rather robust, does not > > consume too much CPU time and bandwidth, and is easy to implement. > > I'm not sure what the architecture you are suggesting is, but this is > what I suggest as the simplest to set up. I envision it to look like a fishnet. > Have one main majordomo. There should be no "one main" anything on this project. > You subscribe to the main majordomo request address, and it forwards > your subscription request to a random mail-exploder. > > You unsubscribe to the main majordomo request address, and it forwards > your subscription to all the mail-exploders request addresses > (unsubscribe traffic is low anyway, keeping track of who is subscribed > where at the main major domo doesn't seem worth it). > > Each person who wishes to run an exploder is subscribed (manually) to > the main majordomo. > > You submit articles to the main majordomo, and it sends copies of the > articles to it's subscribers (the mail-exploders). > > The mail-exploders send mail to the address on their subscriber lists. > > (John Gilmore suggested this architecture, as a simpler alternative). Simpler? Hardly. With the remailer chain proposal we could have a working remailer network up in a couple of more days. What you propose will take weeks to write and debug code and then follow it through. No, this approach is neither simple or efficient. What I call a fishnet toplogy is much cleaner. Each node only has to filter messages where it was in the source chain. Easy to do with procmail, if you look at the first example on the man page, eg how to dump to dev/null, you find a perfect example of what procmail needs to do. It does not need to know who else is on the network (as in a star) and its traffic consists of one copy of each incoming instead of n copies from each of the 'star' remailers plus the outgoing from the local subscribers. A easy protocol can be arranged so that if the downstream remailer breaks it subscribes to the next one in the chain, again easy to do with just a few lines of code... If (bounce count of next site expires) { subscribe backup feed } If (next site returns) { unsubscribe backup feed } As to the actual architecture: A --> D ^ \ ^ \ / v / v C <-- B <-- E ^ | v N (mail-news gateway) Again, several nice features are: * Can be gotten up from a stock majordom/procmail install in a trivial manner * Site only has to filter mail from itself * It only has to keep up with two sites in the network instead of many (support for limited scope and anonymity) * Each site gets a critical feed from only 1 other site so traffic of duplicates is a minimum * Except for sites on the boundary of the net every site has two places to send its output, adding to robustness, if we allow the edges to wrap (eg torus) then no sites have less than 2 other sites to send traffic to. * Supports various levels of encryption or authentication on a link by link basis, not forcing all members to submint to a general architecture * Scales easily, something the star model does not do since hub machines must grow at a rate dependant on the total number of remailers in the chain this equates to mullah and lots of it should the network take off * The actual number of machines in a given chain is flexible, I would guess the number should match a plane tiling figure, triangle (n=3), squares (n=4), or hexagons (n=6) * Traffic analysis and sync floods are harder to impliment in this architecture than a star or hub based network * Supports 'little' machines and operators who are not guru's or have deep pockets Jim Choate CyberTects ravage(a)ssz.com

1 0