- cypherpunks-legacy - lists.cpunks.org

What's More Important?
by tcmay＠netcom.com 15 Nov '92

15 Nov '92

I try to never say on lists or newsgroups that some subject is inappropriate, or boring, has already been beaten to death, etc. Instead, I figure it's up to others to make this choice. I can always ignore threads that don't interest me. But let me confess I'm mystified as to why the subject of random number generators (and CD-ROMs filled with them) is getting so much attention while Hal Finney's remarkable post on his work toward a fully-encrypted remailer (allowing digital pseudonyms) has received no discussion whatsover! Random number generators, in software and in hardware, pop up in discussions on sci.crypt every couple of months or so. Every argument made here, every basic question, etc., has already come up several times in just past year! Furthermore, and I don't mean to sound chiding or condescending, one-time pads are fundamentally not the way to go. Yes, I know that I just applauded George Gleason's cryptodiversity idea...but I see the stark contrast between the dozens of postings on RNGs, Rander, one-time pads, and CD-ROMs and the utter lack of postings about Hal's remailer...and I am chagrined. Granted, Hal's system is hard to follow. This is another area where some diagrams would help immensely, especially drawn on a blackboard. But anonymous remailing is where cypherpunks need to be. --Tim -- .......................................................................... Timothy C. May | Crypto Anarchy: encryption, digital money, tcmay(a)netcom.com | anonymous networks, digital pseudonyms, zero 408-688-5409 | knowledge, reputations, information markets, W.A.S.T.E.: Aptos, CA | black markets, collapse of governments. Higher Power: 2^756839 | PGP Public Key: awaiting Macintosh version.

1 0

Some organizational trivia to consider
by crunch＠netcom.com 15 Nov '92

15 Nov '92

I will be "off line" for a few days now, as I have to do some work to help out an old friend. I'll be back "on line" sometime late Tuesday, and hopefull will have formulated some organizational plan to propose to the group. In the meantime, start thinking about how we can organize "on line" meetings (Or even meetings in person) to hammer out some type of plan. Someone mentioned a meeting to be taken place later, is this meeting for the Cypherpunks project? My involvement at this point will be that of an adviser, and perhaps pump in some ideas to the group. As the host of the Programmers Netsork, I am experienced at setting up organizational groups, but often these groups tend to "fizzle out" as people have other time committments, so it's important that we try and work out a mechanism for "passing the torch" when one person has to bow out for some reason. My recommendation is to set up an FTP site where one can download the latest organizational plan, people involved, who is doing what, etc, and keep this info updated on a regular basis. I don't recommend that just ONE person do this, it has to be done by several people in the event one (or more) have to drop out for some reason. In the meantime, just be thinking about the following: a) Setting up a cypherpunks FTP site to contain organizational info, latest source code, who is doing what, and a list of things that need to be done. b) Maintaining a SOurce code control mechanism whereby people can "check out" code modules and work in parallel towards a common goal. c) Provide periodic updates by posting important info to the rest of Cyberspace via UseNet or other on-line systems. Till next Tuesday, take care Y'all.... See U in Cypherspace next Tues Cheers John Draper crunc(a)netcom.com crunch(a)well.sf.ca.us

2 1

Re: Rander box and other stuff
by George A. Gleason 15 Nov '92

15 Nov '92

I have a question regarding the ideas which have been circulating about using CD ROMs as key storage. How does one go about rapidly and effectively erasing everything on a CD ROM? The reason I ask is, in the case of OTPs you want to cancel your key as it's used, to prevent accidental double-use of key; and of course in any system you want to make sure that all your crypto material can be wiped squeaky clean in the event of a barbarian invasion. -gg

1 0

cypherpunks mail list.
by Keith Basil 15 Nov '92

15 Nov '92

i'd like to be added to the list. thanks. keith

1 0

Re: Rander box and other stuff
by tcmay＠netcom.com 14 Nov '92

14 Nov '92

Eric Hollander writes: > All of these designs leave us Mac users a bit out in the cold. My powerbook > 100 has one lonely little serial port and no parallel port. Desktop macs > usually have two serial ports, and it's not easy to add more. > > For us powerbook users (and also for users of other notebooks), we want > something which is internal. It's horible to have little things dangling > off of your nice little notebook. They inevitably get pulled out, etc. Maybe I'm being crypto-dense, but why would individual Powerbook 100 users care so much about generating the bits in the volumes that a hardware-based RNG is designed to supply? For filling a CD-ROM (600 MB) with good random bits, I can see the need for a back-biased diode source (such as Tony Patti's widget), but such filling of a CD-ROM presumes a fair amount of other stuff, like the CD-ROM burners, etc. I can't imagine a PB 100 user, and I'm one of them, developing OTPs on CD-ROMs on the PowerBook. For generating PGP keys, the keyboard timing methods work quite well, as several folks have pointed out. Maybe for some of the dynamic key generation applications that will be appearing in the next couple of years such a hardware RNG will be useful (so that the bits can be generated in the absence of a human operator, and at higher rates). But I don't see them needed immediately, and certainly not on a PowerBook. (No insult to the PowerBook, it's just that I can't see Eric Hughes' and Hal Finney's remailer running on such a platform, for obvious reasons.) Am I missing something? --Tim -- .......................................................................... Timothy C. May | Crypto Anarchy: encryption, digital money, tcmay(a)netcom.com | anonymous networks, digital pseudonyms, zero 408-688-5409 | knowledge, reputations, information markets, W.A.S.T.E.: Aptos, CA | black markets, collapse of governments. Higher Power: 2^756839 | PGP Public Key: awaiting Macintosh version. -- .......................................................................... Timothy C. May | Crypto Anarchy: encryption, digital money, tcmay(a)netcom.com | anonymous networks, digital pseudonyms, zero 408-688-5409 | knowledge, reputations, information markets, W.A.S.T.E.: Aptos, CA | black markets, collapse of governments. Higher Power: 2^756839 | PGP Public Key: awaiting Macintosh version.

1 0

Re: Rander box
by George A. Gleason 14 Nov '92

14 Nov '92

Re Eric's quote of me agreeing with him that OTPs are "expensive to make relative to other forms of security." A point of clarification seems in order. I do agree that OTPs are more expensive and less convenient to use than PKSs. However, I also believe that the public interest would *best* be served by having *many* different kinds of cyphers available, including OTPs, PKSs, and various conventional cyphers, historic cyphers with relatively little current security value (for educational purposes) and so on. The main advantages of OTPs are provable absolute security and the fact that the basic technique is so straightforward that it probably could never be banned and put out of circulation. The time may come when we *need* OTPs, and we ought to have them ready beforehand, and have them in use in appropriate situations long before any crisis comes (to gain operational experience which could lead to improvements). With regard to PGP, I am not sure what the copyright status is on that one; and if there is any doubt about it, the govt could screw a lot of people to the wall on copyright-related charges if they so chose. I would like it very very much if PGP was free & clear public domain. The last thing we need is for the first warning of tyrrany to be a wave of hardware seizures on the grounds of having unauthorised copies of copyrighted material. Now I may be off base on this point, but the key here is the idea that many different kinds of cyphers, like many different varieties of plants and animals, make for a robust ecosystem which can't be wiped out by one plague. -gg

2 1

Re: Rander box and other stuff
by jordan＠imsi.com 14 Nov '92

14 Nov '92

Hey, I have a PowerBook too, but don't say you're limited to one serial port. You could use the desktop bus, and there are extenders available. I do agree that a parallel box wouldn't do me much good on my Sparc. For a while, my IPX was my "portable" computer, albeit with monitors on both coasts. But the box itself is quite lugable. If I wanted to, I could plug it in along the hallways of airports and use a palmtop with a serial connection to read mail and up/down load ... /jordan

1 0

Rander
by crunch＠netcom.com 14 Nov '92

14 Nov '92

Tom Jennings writes: >Using serial ports is probably not a good idea on PC clones. The IBM >hardware design limits a machine to two ports comfortably, four >practical maximum. Any machine that uses telecomm. (BBSs, etc) will >have to consume the serial hardware, and will interfere with whatever >you have to do. >A good choice would be a parallel port, ie. a printer port. The IBM >design allows 4 or more easily, and rarely do you see a pclone with >more than two printers attached. If there isn't a spare port >available, Fry's sells printer adapters for $19.95. Late-model printer >adapters are 8 bit in and out, and are capable of Ethernet speeds. >The parallel port uses standard busy/done and TTL levels. I already thought of that, but it would make it difficult to work on other platforms. Tom also writes: >A simple driver could simply have say 10,000 bits, and continuously >overwrite the oldest (wrap around the buffer...) This was (and probably still is) my overall intention.

1 0

re: Rander box and other stuff
by Tom.Jennings＠f111.n125.z1.FIDONET.ORG 14 Nov '92

14 Nov '92

U> From: crunch(a)netcom.com (John Draper) U> I think I have a rough description of the hardware U> serial random generator I want to build. I want to call U> it the "Rander box" for lack of a better name. Using serial ports is probably not a good idea on PC clones. The IBM hardware design limits a machine to two ports comfortably, four practical maximum. Any machine that uses telecomm. (BBSs, etc) will have to consume the serial hardware, and will interfere with whatever you have to do. A good choice would be a parallel port, ie. a printer port. The IBM design allows 4 or more easily, and rarely do you see a pclone with more than two printers attached. If there isn't a spare port available, Fry's sells printer adapters for $19.95. Late-model printer adapters are 8 bit in and out, and are capable of Ethernet speeds. The parallel port uses standard busy/done and TTL levels. (For a "universal" serial version, you could internally have the 8 bit busy/done interface drive a UART. I bet simply fixing the data rate to 9600 baud would work.) U> When switched on, and a "cr" (or some other U> character) is sent to it, random bytes will stream out U> continiously. Another "cr" will stop the byte stream. U> At least this is ONE approach. If anyone can think of a U> better way, Pse speak up. I was about to say "prolly not necessary", but others make valid points... XON/XOFF, DTR, etc are all workable... DTR is compat w/serial and parallel too. On a pclone, producing bits continuously is easier. When not in use, the driver will dump bits onto the floor. Fancy (software) drivers could spool the bits into a nice big pile ("every bit is sacred..."). A simple driver could simply have say 10,000 bits, and continuously overwrite the oldest (wrap around the buffer...) --- ReadMail * Origin: tomj(a)fidosw.fidonet.org / World Power Systems (1:125/111) -- Tom Jennings - via FidoNet node 1:125/555 UUCP - ...!uunet!hoptoad!kumr!fidogate!111!Tom.Jennings INTERNET - Tom.Jennings(a)f111.n125.z1.FIDONET.ORG

2 1

Re: Rander
by simsong＠next.cambridge.ma.us 14 Nov '92

14 Nov '92

Perhaps I'm just out of the loop, but what do you intend to do with your stream of random numbers? You going to send a tape to your friends under armed guard? Just curious. Oh, I built a hardware random number generator a few years ago for some ESP experiments I was doing. I wanted to see if computer hackers had better odds of affecting the output of a RNG than non-hackers. (You know, the hacker walks into the room and suddenly the broken piece of equipment starts working?) Turns out they can. -simson

1 0