January 1993 - cypherpunks-legacy

multiple aliases. It doesn't matter how many
by mjr＠netcom.com 12 Jan '93

12 Jan '93

***** Karl L. Barrus <barrus(a)tree.egr.uh.edu> ***** >I agree with Matthew that not mailing back an ack would help cut down >the flurry of mail, but it still results in all sorts of extra id's. Yes, but so what? That is, why does it matter so long as all the conversion from one to the next takes place automatically. The process strengthens your security as well as that of any respondent. True this may not be necessary, but under the circumstances (the whole point of alias servers) isn't a "lets not take chances" approach best? Lets take an extreme case (not that I'm suggesting things be implemented this way). Imagine that every time you or anyone else originates mail through an Aserver you are given a NEW ID (not just the first time, but EVERY TIME). Again, so long as a relationship is maintained between all your ID's on a given server and their corresponding ID's on the next machine down or up the line, it shouldn't matter to you at all! After a few years you could end up with hundreds or thousands of IDs. What difference would it make? You don't need to know what *any* of them are... matthew rapaport Philosopher/Programmer At Large KD6KVH mjr(a)netcom.com 70371.255(a)compuserve.com

1 0

.Sig suppression
by Hal 11 Jan '93

11 Jan '93

-----BEGIN PGP SIGNED MESSAGE----- Chael> I'm Chael> sure there are a couple, but I agree with your point that most BBS's on Chael> any mail network append an identifying "tagline" or signature. As a Chael> matter of fact, in many nets it is a requirement that your system append Chael> a tagline to all messages. Incidentally, it is preceded often by "--" Chael> on a line by itself. I'd like to hear more about systems which do this. What is the rationale for adding the system name at the end? Do these networks not use Internet-style "From:" headers, so these automatic system-wide .sigs are used for the same effect? I guess there must be gateways between these bbs's and the internet, for this issue to arise. It's too bad that these gateways don't convert the .sig info into a more conventional RFC-822 style Internet header. Hal Finney -----BEGIN PGP SIGNATURE----- Version: 2.1 iQCVAgUBK1F586gTA69YIUw3AQFNogP9FU2W3wrHTnfrZeKtrMOq4Zz5aTUN7+vv 04iMOdV975fCzqdmgR7O758qamewguYV8XHmPVloLSMwgnmbzBNs8zRZkVAwTKnB rpQqeahXPNeC1PVu/ezoiBvc26ujcN2Ga9OuGUWu9RFRvjwQ0rl51mTjHED1fJi+ 7I/TV4kT4Kk= =WeLq -----END PGP SIGNATURE----- Distribution: CYPHERPUNKS >INTERNET:CYPHERPUNKS@TOAD.COM

2 1

constructed anonymous addresses
by Karl L. Barrus 11 Jan '93

11 Jan '93

Had sort of a slow day at work, so I had the chance to crank out this: Here is a rough script which will allow you to route your mail through the various remailers. (However, extropia is not yet supported...I'm working on it). You must have the public key for each remailer on your keyring. Save the file, run it, and it will prompt for 1) mail, or 2) header. Choose 1 to pick a path and send a file, choose 2 to create a header which can be used to reach you. I intend option 2 to be used as follows: create a message to someone you want to remain anonymous from, run the script, pick 2, follow the prompts, and then instruct the script to append the header to your letter. Then your recipient will be able to follow the directions and respond to you. When the script prompts for "And then to (1-3 or address)", to end the process, type the final address you want your mail sent to. Of course, if you are creating a header, you'll want to type your own address. Further improvements include supporting extropia; and rewriting in C, perl, awk, ksh, or any language with arrays! I've tested this script using the remailers at alumni and rosebud, because the turnaround time seems faster. But there is no reason that rebma shouldn't work as well. --------8<--cut here-->8-------- #!/bin/sh # support script for anonymous remailers # allows routing a message through various remailers # NOTE: extropia remailer not supported just yet #find out which mode user wants echo "Do you want to:" echo "1. Create routing and mail a file" echo " This will allow you to specify the route your message will take," echo " and mail a file through that route." echo "2. Create remailing header" echo " This creates an appropriate remailing header, with instructions." echo " Include the generated file in one of your own messages and" echo " someone else can use it to reply to you." read choice #declare remailers mail1=hal(a)alumni.caltech.edu mail2=remailer(a)rebma.mn.org mail3=elee7h5(a)rosebud.ee.uh.edu mail4=remail(a)extropia.wimsey.com #temporary files t1=.anon1 t2=.anon2 t3=.anon3 #set up header echo "::" > $t1 echo "Encrypted: PGP" >> $t1 echo "" >> $t1 #blank out .anon3 cat /dev/null > $t3 #print menu echo "" echo "1) $mail1" echo "2) $mail2" echo "3) $mail3" #echo "4) $mail4" echo "" #get hop echo "Mail to (1-3): " read to #convert from number to address if [ $to = "1" ] then to=$mail1 elif [ $to = "2" ] then to=$mail2 else to=$mail3 fi firsthop=$to notdone=true #begin loop while [ $notdone ] do #find out remailing request echo "And then to (1-3 or address): " read rto if [ $rto = "1" ] then rto=$mail1 elif [ $rto = "2" ] then rto=$mail2 elif [ $rto = "3" ] then rto=$mail3 fi echo "::" > $t2 echo "Request-Remailing-To: $rto" >> $t2 echo "" >> $t2 # echo "remailing to $rto; encrypted for $to" pgp -ea $t2 $to 2> /dev/null cat $t1 $t2.asc >> $t3 if [ $rto = $mail1 -o $rto = $mail2 -o $rto = $mail3 ] then to=$rto else notdone="" fi done if [ $choice = "1" ] then #now include message echo "Message to include? " read msg if [ ! -f $msg ] then echo "$msg not found" exit 1 fi cat $msg >> $t3 elm -s "anonymous mail" $firsthop < $t3 else echo "Append to file: " read msg echo "--------8<--cut here-->8--------" >> $msg cat $t3 >> $msg echo "" >> $msg echo "<To reply, save everything below the \"cut here\" marks above" >> $msg echo "<into another file. Type your reply here (below the blank " >> $msg echo "<line three lines above!) and mail to $firsthop" >> $msg fi rm -rf $t1 $t1.asc $t2 $t2.asc #end of script---------------------------- /-----------------------------------\ | Karl L. Barrus | | barrus(a)tree.egr.uh.edu (NeXTMail) | | elee9sf(a)menudo.uh.edu | \-----------------------------------/

1 0

request digest format
by Richard Childers 11 Jan '93

11 Jan '93

When is this going to be converted into a Digest format ? Without wishing to appear peckish, Christmas would have been a good opportunity ... -- richard ===== -- richard childers rchilder(a)us.oracle.com 1 415 506 2411 oracle data center -- unix systems & network administration ... whatever remains, however improbable, must be the truth.

1 0

Atari PGP -is it available..
by ccat＠netcom.com 11 Jan '93

11 Jan '93

Does anyone know if there is a current version of PGP available for the Atari ST. Thanks!

1 0

Commie's
by Matt Willis 11 Jan '93

11 Jan '93

To this southern boy, it looks like the pinkos are slamming American (whoops, sorry, English) programming... :) "pgp in actual use is not recommended.", oh yeah, well our alphabet has more characters, so there... There were some interesting flaws that they pointed out... Could someone do a follow up and say what has been fixed in PGP version 2.1? If pgp cyphertext/encrypted data is predictable, I wonder how that would affect the "it'd take a Cray a year" figure? (or were the pinkos being overconfident) I actually like the Russians, but I'm just fulfilling my stereotype. peace :* +-------Matt-Willis--------------------------------+ | Matt Willis ASTMWILL(a)STETSON.BITNET | elsewhere: | Matt Willis Head of the Underground | mwill(a)mindvox.phantom.com | Matt Willis Robotech PBM List | +-------Matt-Willis--------------------------------+ "Absolutely alone in awareness of the mechanism." -Agrippa by WG

1 0

Chaining Pax and Penet.
by Hal 11 Jan '93

11 Jan '93

-----BEGIN PGP SIGNED MESSAGE----- From: Karl L. Barrus <barrus(a)tree.egr.uh.edu> > My original thinking was that once I established the id's in both > direction, when someone responded to anon.435@pax, they would be > allocated an id if they didn't have one. And since penet had by this > time seen anon.435@pax, no new id would be made, and the mail would > proceed on the me. The problem is this: when someone responds to your anonymous ID anon.435@pax, their mail _from_ Pax does not come from anon.435. Anon.435 is _your_ id. Instead, their mail from Pax comes from their own anonymous ID (possibly a newly allocated one). Then, when the mail goes to Penet, it sees this new "From" ID and allocates one of its own. The same thing happened when you sent to anon.435@pax from your system which already had a Pax ID. When the mail was forwarded from Pax to Penet, it was not marked as coming from anon.435. Instead, it was marked as coming from this already-assigned Pax ID. (I don't think you ever said what that already-assigned ID was.) Penet had not seen that ID before, so it allocated an alias for it and sent back to that ID. Penet's mail-back would _not_ go to anon.435, but rather to the Pax ID which it was replying to. Hal -----BEGIN PGP SIGNATURE----- Version: 2.1 iQCVAgUBK1F6D6gTA69YIUw3AQGMTAQAqXm7SdE6uyf+04J5GY3KU7dk7A2D7loC TeT+0UqpsSPOI+31YrJPww2h9XuwGylAZ9dqu/hPdolIzukjr+WiOKRyU34imezd iX9yYv3Ry3jCebcn9c79NY3zEQhjGh1LhqKmec5QLp3FjPB+gQZZypdaHz4GeDJF 4oDyArzKafc= =wZgY -----END PGP SIGNATURE----- Distribution: CYPHERPUNKS >INTERNET:CYPHERPUNKS@TOAD.COM

1 0

Politics of Rmailers
by norm＠netcom.com 11 Jan '93

11 Jan '93

Theodore Ts'o writes: If your mail system is broken enough that it inserts signatures without your permission, and you have no way to controlling it, it's broken. End of statement. Fix it or ditch it. I can imagine a system administrator choosing to require that all mail originating from his machine include a signature that correctly identifies the local name of the sender. I make this special point to illustrate a broader problem with remailers: They require operators of remailers to be sympathetic with the ends of the users of remailers. This obviously does not include the entire population for at least the recipient is not sympathetic. I suspect that technical solutions sought in recent mail will founder in presence of the politics of the operators of the remailers. I understand that routing your message thru at least one "friendly" remailer may be enough but if your reasons for using remailers are not sufficiently popular, then society, in some form, will pressure the friendly remailers to betray the sender without advance warning. If society polarizes into camps then there may be remailers in each camp. A remailer in one camp is unlikely to service messages from the other. Barriers then arise. I think that the technical issues are only the tip of the iceberg.

4 7

both stripping and not
by tribble＠xanadu.com 11 Jan '93

11 Jan '93

With even current remailer architectures, it's trivial to simply have different services for normal remailing (which leaves body intact) and stripping services that grundge the message arbitrarily. When we really want anonymity, for instance, we will need message rewriting services that break the correlation between authors and writing style. A friend of mine looked into that material and claims that such analysis can do a depressingly good job at figuring out what messages were written byt he same author, even if the author tries to stilt his style. dean

1 0

No Subject
by Karl L. Barrus 11 Jan '93

11 Jan '93

Hal <74076.1041(a)compuserve.com> said: >My understanding was that everyone who tried to talk to him would get >two aliases assigned automatically. Actually, what I expected to happen was this: the acknowledgement of an anonymous id on penet for the anonymous id on pax would be generated, and this ack would be sent via the chain back to me (barrus@tree) since I established the chain. That is, I had linked anon.435@pax to an5022@penet to barrus@tree. Then, upon mailing to anon.435@pax from my other account (which already has an anonymous id established, so another would not be generated), the mail would proceed to an5022@penet, which would create an id and send it back to anon.435@pax. Now I expected this ack to then be turned right around, send to an5022@penet and then on to barrus@tree. So I was expecting the creation of another anonymous id, but the acknowledgement didn't go to barrus@tree. My original thinking was that once I established the id's in both direction, when someone responded to anon.435@pax, they would be allocated an id if they didn't have one. And since penet had by this time seen anon.435@pax, no new id would be made, and the mail would proceed on the me. Anyway, that was an experiment that seems to lead to explosive anonymous id growth :-) I agree with Matthew that not mailing back an ack would help cut down the flurry of mail, but it still results in all sorts of extra id's. I was hoping the whole thing would be like a pointer: mail to id1@pax forwards to id2@penet and then on to me with no extraneous account manufactured. But since we have our own cryptographically protected remailers, we cypherpunks can make our own remailing chains (Hal's constructed anonymous addresses). This way, you can decided on the path of your outgoing mail and your return mail: create the appropriate header, and attach your message on the end. To receive responses, just send a response header with the return path encrypted along with instructions to your recipient to cut the response header into a new file, add a message to the bottom, and mail to the appropriate remailer. (Note: I've used this method successfully twice, so it isn't too hard to do). Just remember that if your recipient doesn't have pgp, don't route your mail through extropia or their message will be blocked. If I get a chance I'll work on a program that will generate the appropriate header given routing input. /-----------------------------------\ | Karl L. Barrus | | barrus(a)tree.egr.uh.edu (NeXTMail) | | elee9sf(a)menudo.uh.edu | \-----------------------------------/

1 0