December 2003 - cypherpunks-legacy

Notes on getting a business on the Internet
by hugh＠ecotone.toad.com 17 Dec '03

17 Dec '03

At the April San Francisco Bay Area Cypherpunks meeting I gave a talk on what it takes to get a small business on the Internet to provide some service. Most of the talk was on very technical gnunk dealing with telecom equipment & lines as well as finding and starting up IP service. I have been asked many times for the notes, lists and tables I used in the talk, so after (not enough) editing and some expansion so that they would make some sense out of the context of the meeting here they are. Do note the new section on Frame Relay, as I got updated on the changes in Frame Relay after the April meeting. ||ugh Daniel hugh(a)toad.com # # notes.on.internet -- A list of questions and resources. # by ||ugh Daniel <hugh(a)toad.com> 1994/04/09 # # RCSID $Id: notes.on.internet,v 1.6 1994/05/18 15:48:08 hugh Exp $ # # Copyright 1994 by Hugh Daniel # Permission is granted for non-profit net distribution. # Comments, questions, permission, consulting or correspondence works # best with me over the net, use hugh(a)toad.com or hugh(a)xanadu.com. I # I can also be reached via telephone at +1 415 473 0669 or by snail # mail via: # Hugh Daniel, 210 Clayton Street, San Francisco, California 94117-1914 Getting_your_biz_on_the_Internet If you are thinking of doing that bright new network business idea of yours, here are a few questions to lead you through some of the maze that is the networking world of today. Mostly this note focuses on one of the middle parts of the problem, getting your self on the net in such a way that you can provide that great new network service. There are many ways to get on today's Internet, and things are changing fast, what was a good idea or product last month is often the slow dumb way to do things this month. Keep asking questions, watch out for new services and tariff's as they might be much better or cheaper that what you are planing right now, even if you got it out of these notes! Note that you might be able to buy the networking services you need from a vendor currently on the net, and this note does not deal with that. You can likely also find someone, maybe in the role of a consultant, who can help you do much of this work, this can be a very valuable service. After you are on the net there is much left to do, and hopefully there are a few useful pointers here to get you going in the right direction. This file was built out of my notes for a talk given in April 1994 about how to get a small business on the Internet and providing a service. Remember that its usefulness and your mileage will vary. Good luck! * Contents Getting Hooked Up Get your idea and business structure together first Decide whom your first target market should be What sort of connection do you/they need to each other? Which IP provider do you want to buy from? Questions to ask prospective IP providers Netiquette Provide information about your service via the net itself! Resources An example of comparing IP providers Frame Relay Acronym's Phone number de-construction Leased line providers in the SF Bay Area SF Bay Area IP providers Some National IP Providers Useful Books and Readings Government Service Providers Datacom Hardware Beam Cast & Broad Cast systems PacBell ISDN Questionnaire Online Resources UseNet NetNews Groups Telecommunications speeds * Getting Hooked Up ** Get your idea and business structure together first If you do not know what you want to do, then you will be lost from the get-go. Write up your idea, see if others get the idea, ask them if they would buy (don't stop if they say no, but this gives you an idea who your market is and is not). Your biz idea needs to be clear to yourself, as you will need to explain it to many folks along the way (like the folks who are working for you trying to do it!). Get your business act together, in the USA this can seem (and likely is) a huge waste of time and effert, but having a structure that you can work with as things grow is important. It just might keep the government off your back as well. ** Decide whom your first target market should be Do you have the software?, networking protocols? Do they have the money to pay? Etc. ** What sort of connection do you/they need to each other? Look at whom you wish to serve with what. Where are they? How are you going to reach them? What do they already know how to do? How far can you push them (new tech.) before they burn out? Will they buy faster modems/ISDN/FRAD's to talk to you? ** Which IP provider do you want to buy from? Ok, you have thought about things and decided that you need to put your shiny new service on the Internet your self. Now you have to decide on whom to purchase your IP feed from, and that can be tricky to say the least. Here are some questions to ask your self, and then some more to ask of IP providers, phone company etc. *** Questions to ask yourself first: Do you know enough to run your own Box (Un*x or what ever) at as commercial service? Do you know enough about networks, firewalls, telecom and the like to do set this up? Do you understand overbooking? (It's standard in networking) What sites are best for a POP? Is it near a RBOC CO? What city's have the largest local area coverage? What types services might you be providing, what data rates. Interactive? Batch? Client/Server? FrontEnd/BackEnd? Do you understand Through put vs. instantaneous transfer rate. (100K per hour vs. Latency, say 56K=10ms or T1=3ms,) Do you know how to set up a secure, stable, central POP? Do you need a Service Provider or a Network IP Provider. Some things can be done just fine at a service (they run a machine for you on the Internet). Are you willing to learn enough to do all this? Know you major wire telecom methods: Modem 300bps thorough 28.8k bps dial-up or full time Leased line 56k bps or T1 ISDN 2B+D 64k, 112k, 128k, 142k Frame Relay 56, 128, 348, 512, T1 (Pacific Bell) Beyond common `current' Needs... T3/DS3, FDDI, ATM/SONET, X.25, BeamCast Now go shopping for a: IP Provider Wire (Leased Line, Microwave, etc.) POP (secure room) judge on: contract Can you do what you biz needs under this contract? price Is the products value to your biz worth the cost? support: Structure (Do they have an organization that can handle both executive and technical problems so that you have little or no down time?) Do they have skilled People Do they provide Guarantees? (a guarantee says that if they screw up you get something beyond "sorry", the service is backed up with more then the sales persons word.) Before calling around for hardware, leased lines and IP do your homework, read up in NetNews, some of the books and poke around some of the better WWW and FTP servers educating your self. (See later in this file for some useful references.) You will ask better questions. *** Questions to ask prospective IP providers: --- Wiring questions: Where is your nearest POP to my site? Do you have a 24 hour staffed trouble hot line? What is your trouble reporting and tracking system? How far is it from my wall plug to first point where a single point failure will NOT interrupt my service from you? Will you provide both a logical and physical a map of your IP interconnect to other the IP networks? --- IP questions: Do you provide the leased line, DSU/CSU, and/or IP router? (Some will sell/lease these to you for extra $$) Do you require that I have a certain DSU/CSU or Modem? Do you require that my gateway router be from some vendor? How many routes will you advertise to the net for me? Can your system route the MBONE to our router? Do you have any restrictions on how I use my IP feed? Are you a CIX member? Do you have a WWW/FTP or gopher site? Do you have email to your NOC, sales or executive employs? --- Contract questions: What service guarantees do you provide? Under what circumstances will you pull my IP feed? Will you email/www/snailmail a copy of the IP service contract now (so you can pick nits in it and maybe ask them to CHANGE it for you!)? What is your policy on overbooking of bandwidth, when will you put in more bandwidth to me if my line if way too overbooked? How much is the install price of <56K, T1, etc.> service? How much is the monthly price of <56K, T1, etc.> service? Do you have alternative payment plans (pay ahead, pay install costs over X months, etc.)? Which (if any) PUC tariff is this service offered under? Do more homework, is there a better tariff that your PUC forces your phone company to sell but they did not tell you about? Build a spread sheet of the possible, make your choices. Sign your contracts for POP space, leases lines, IP etc. Make sure you have more then one person who can get into the POP to fix things, 24hours a day even! Remember now that physical security is the first step to a secure and well run system. Time to Buy things for your POP: VOM Tester DSU/CSU Router Service Box (UNIX, terminal servers, etc.) UPS Remote control Rack Fire Extinguisher. Lots of cables Labeler gun Register your self with the DNS. Getting things up can take days, allow time for it. Get your service machine up. Do your first security pass BEFORE your net connection comes up. Now is the time to install crypto protocol services. Do a backup BEFORE you get on the net... Get your connection up, trace wire problems from the ends in. Get IP packets flowing, routing. Test your IP throughput, know what its parameters are as your line may never be this idle again. Get your email working first, you need these aliases: postmaster info and your users will likely want: admin support Get your WWW working (or advertise else how on the net). Get listed on the various WWW Biz pages: http://www.eit.com/demos/storefronts.html http://tns-www.lcs.mit.edu/commerce.html Get listed in the IBP (Internet Biz Pages) (info(a)msen.com) Do backups. Tune into Net Culture, there are some netnews groups with cultural norms, read up on them now: news.announce.important news.announce.newusers Constantly test your service, as it's better you find a problem that to have a user report it. Use it your self, in the same way your users would. Make every customer very happy... Join the EFF. Sell your service, provide value and reap profits from a job well done. Use Free Strong Cryptology! ** Get your your self on the net, learn its ways (Netiquette) Get on a Usenet site and read the postings in news.announce.newusers, it is full of useful information on good Netiquette. You can do this by buying a shell or other type of account that gives you access to the net for email and netnews, downloading etc. before your biz is on the net. ** Provide information about your service via the net itself! Check out the Usenet biz.* groups, this is where you can do business on the Usenet, like advertise your services and/or products. Check out the various WWW servers that have lists of commercial internet WWW sites, see if you can get yours added. Places currently doing such things are eit.com, the GNN (ora.com) and msen.com. Check out the lists of lists and as the -owner's of lists that are related to your biz if it's ok to post an informative message about your service. Ask (politely!)that a reference to your biz be added to the lists FAQ if any. * Resources Here are some random notes, tables and lists that might help you find things you will need. Most of this is targeted to the San Francisco Bay Area, as that is where I have been doing consulting on these topics, though the basic ideas are useful no matter where you are doing telecom. ** An example of comparing IP providers Below is some info that I collected in the process of setting up a site to go in the internet. First there is an example of collecting and comparing some of the data gathered. There there are several sections of useful information that follow. -------- Example Internet Service (IP/TCP) Q&A array ------------------------- Quest/Whom |TLG Sprint BARRNet Alter Netcom ANS PSI CERFnet ---------------+--------------------------------------------------------------- Called | on date | 02/29 02/29 02/29 02/29 never never never status | yes message message message responded on |*1 Mar-03 Mar-03 Mar-07 | Nearest POP |MV SJ SJ SJ | Whos Equipment | Near DSU/CSU |cust either cust netcom Far DSU/CSU |cust either? alter netcom Near Router |cust either cust netcom Far Router |TLG Sprint alter netcom Whos LeasedLine|Cust Sprint Alter cust | Service | 24hour HotLine |Nope Yes Yes Yes | SellDownStream |Yes Yes No No | 56K Install |$1500 $ 750 $ 0 $1995 56K Monthly |$ 325 $1000 $ 795 $ 400 | T1 Install |$1600 $1000 $5000 $6000 T1 Monthly |$ 800 $2700 $1250 $1000 | Totals: |--------------------------------------------------------------- | *1 == All TLG info was on the tlg WWW/FTP server and it took only minutes to get all the info needed except leased line info, which had to come from GTE & PacBell in my example. -------- DSU/CSU line drivers (Modem like things) ---------------------------- Speeds Company Model Price ------------------------------------------------------------------------------ T1 Tylink ONS-150 $1200 each 56K Motorola $500 56K BAT $250 -------- IP Routers ---------------------------------------------------------- Company Product Price ------------------------------------------------------------------------------ Livingston IRX Cisco (any) (high end, with quality) -------- Terminal Servers ---------------------------------------------------- Company Product ------------------------------------------------------------------------------ Livingston 2 & 2e Cisco (any) (high end, with quality) -------- Example Leased (digital) lines prices ----------------------------- Quest TLG Sprint BarNet AlterNet ------------------------------------------------------------------------------ Nearest POP MV ?? ?? ?? Mileage 14 PacBell ADN0 56K Install $1240 Monthly $184.10 Delay 17wkd ADN1 T1 Install $2648 Monthly $794.18 Delay 1wkd POP 444 Mileage 36? MFS ADN 56K Install $675.00 Monthly $327.15 ADN1 T1 Install $813.50 Monthly $1038.51 POP 55 S.Market SJ Mileage MFS ADN1 T1 Install $1140.30 Monthly $455.31 Sprint POP SanJose ADN0 56k Install $935 Monthly $327.87 ADN1 T1 Install $1699 Monthly $1074 -------- Leased Lines PacBell leased line pricing (new for 1994 it seems!): $6 per mile. Node (each end point) $50.05 Some credits might be given back to you via some CPUC Mandated rebates. ** Frame Relay This section is totally new to these notes, and is likely to be new even to many folks who have been doing leased line and other methods of getting IP moved about, as PacBell has just gotten a new tariff on Frame Relay services, making them much more useful (and affordable, all this as of early 1994). Frame Relay used to be cost effective only if you had many sites that were far (over 50 miles) apart from each other. Now it seems that if you have three sites, most any distance from each other other then Frame Relay is likely for you. Quickly, Frame Relay is a system that lets you have a digital line (it's really a leased line) more directly into the phone company's digital transmission cloud then a leased line. You pay for a single line into each site you want to wire ONCE, and then pay for routing to each other site in your group of Frame Relay drops. You get one bill for N sites, and it is simpler for the RBOC to deal with too (both on a billing and on a technical level). You will need to get a Frame Relay capable router (Frame Relay is its own protocol) sometimes called a FRAD, these are about $2000, but some of the standard IP routers already know how to deal with Frame Relay. Here are the current prices that I think PacBell is charging as of 1994/05. There are three parts to the charges, #1 is the leased line you have to rent from your site to PacBell, though it seems they do NOT change per mile for this. #2 is the cost of Frame Relay its self, and #3 is the cost of them adding more entry to their routing tables. Note that #3 is very small unless you want to have many sites DIRECTLY talking to many other sites, if you go though a few hubs you can keep the cost down. As of right now there are no time user or per packet charges for this service. Service Costs: Startup Monthly --------------------------------------------------- #1 Local Access Line (your leased line into the data cloud) 56Kbps $ 620.00 $ 50.05 1.536Mpbs $1,324.00 $ 162.59 #2 Network Port Connections (Frame Relay its self) 56Kbps $ 375.00 $ 75.00 128Kbps $ 375.00 $ 150.00 384Kbps $ 375.00 $ 400.00 1.536Mbps $ 375.00 $ 500.00 #3 Data Link Connection Identifiers (routing table entry) 1 $ ---.-- (No Charge) 2-6 $ 15.00 7-11 $ 10.00 12+ $ 5.00 Lets do a quick example. I have 4 sites I want to hookup on my own little net. I want T1 between Site A and B, 384Kbps between site B and C, and 56Kbps between Site B and D. I want the high speed ports to talk directly to each other, but the low speed port to just talk to the B hub site. So I get: Service Units @ Startup Units @ Monthly ------------------------------------------------------------------------- #1 3 T1 leased lines $1,324.00 $3,972.00 $162.59 $ 487.77 1 56K leased line $ 620.00 $ 620.00 $ 50.05 $ 50.05 #2 2 T1 Frame Relay Ports $ 375.00 $ 750.00 $500.00 $1,000.00 1 384Kbps FR Port $ 375.00 $ 375.00 $400.00 $ 400.00 1 56Kbps FR Port $ 375.00 $ 375.00 $ 75.00 $ 75.00 #3 1 3 DLCI Ports (site B talk directly to A,C,D) $ 15.00 $ 15.00 2 2 DLCI Ports (sites A & C talk to A,B,C) $ 15.00 $ 15.00 1 1 DLCI Port (site D talks only to B) $ 0.00 $ 0.00 -------------- Totals --------------------------------------------------- $6,092.00 $2,042.82 You will have to add in the cost of a DSU/CSU (now called a TSU) at each site. If you want to make a change (say run the 384Kbps line up to T1) just making the change costs $30.00. You will want to compare such these numbers to the cost and pain of installing 4 separate leased lines, more DSU/CSU's etc. ** Acronym's ADN = Advanced Digital Network ASCII = American Standard Code for Information Interchange ATM = Asynchronous Mode Transfer CCITT = Consultive Committee for International Telephone and Telegraph CDS = Circuit Digital Service (CDS 56) CIX = Commercial Internet eXchange CO = Central Office (phone company switching building near you) CPE = Customer Premise Equipment CPUC = California Public Utilities Commission CSU = Customer Service Unit (now allays DSU/CSU) Cust = Customer DAML = Digital Access Multi Line (Field T1 Phone Mux unit) DDS = Digital Data Service DLCI = Data Link Connection Identifiers (FR routing table entry) DS0 = Data service grade 0 (56kbps line) DS1 = Data service grade 1 (1.544mbps line) DS3 = Data service grade 3 (45mbps line) DSU = Digital Service Unit (now always DSU/CSU) E1 = A 3.088 mbit/sec channel? FCC = Federal Communications Commission FDDI = Fiber Distrubited Digital Interface FRAD = Frame Relay Access Device? FTP = File Transfer Protocol (runs on top of TCP) HDLC = High-Level Data Link Control IP = Internetworking Protocol (TCP/IP) ISDN = Integrated Services Digital Network Kbps = Kilo (base 2) bits per second (x 1,024) LAN = Local Area Network (1-2 kilometers) LATA = Local Area TA? (Baby Bell monopoly region) M&P = Methods and Practice (Standard Operating Procedures) MAN = Metro Area Network (10's of kilometers) Mbps = Mega Bits Per Second (base 2) (x 1,024,000) MBONE = Multicast BackBone (Protocol suite on top of IP, also a Net) MPO = Minimum POint of entry (the RBOC wire closet in the basement) MTBF = Mean Time Before Failure (or Between Failures) NNX = N=2-9 X=0-9 == Old pre +1 exchanges NOC = Network Operations Center NPA = Numbering Plan Area == Area Code NXX = N=2-9 X=0-9 == Modern (post +1) Exchange PAN = Planetary Area Networks (1000's of kilometers) POP = Point of Presents POTS = Plain Old Telephone Service PPS = Public Packet Switched (PacBells term for there X.25 network) PRI = Primary Rate Interface (T1 bulk feed for 24 ISDN B channels) PUC = Public Utilities Commission PVC = Permanent Virtual Circuit RBOC = Regional Bell Operating Company's (Baby Bells) RFC = Request For Comment (Internet Tech standards) SAN = System Area Networks (10's of light hours) SDS = Switched Digital Service (SDS 56) SMDS = Switched Multi-MegaBit Digital Service SNA = Systems Networking Architecture (IBM WAN technology) SONET = Synchronous Optical NETwork? T1 = A 1.544 mbit/sec channel T3 = A 45 mbit/sec channel TCP = Transmission Control Protocol (TCP/IP) TDM = Time Division Multiplexing TSU = Terminal Service Unit (a DSU/CSU) UDP = User Datagram Protocol (a protocol on top of IP) V.35 = A almost never used specification for 48kbit modems V.35 = Serial Line software/hardware protocol (something like RC232C) WAN = Wide Area Network (100's of kilometers) WDM = Wavelength Division Multiplexing WWW = World Wide Web (One-Way HyperText protocol suite) X.25 = Old packet switching system baud = discrete signaling events/sec (used wrong, you want to use bps) bps = Bits Per Second gbit = Billion (giga) Bits Per Second kbit = Thousand (kilo) Bits Per Second mbit = Million (mega) Bits Per Second sec = Second, 1/60 of a minute wkd = Workings Days (Monday through Friday) ** Phone number de-construction POTS, Plain Old Telephone Service Phone numbers in the USA have these components: +<country code> <area code> <exchange> <line> or if you live at the phone company: +<??> <NPA> <NXX> <line> an example: +1 415 555 1212 remember other country's have different systems (bogus British example): +44 20 22 2121 The idea behind the plus (+) format is that there were a lot of formats for phone numbers that were different even if the phone number format was the same, folks just like doing their own, and this was confusing. The plus format is simple, a plus followed by the country code followed by the phone number. So to dial a random phone number you look at the +??? country code, if you are not in that country then you will need to dial a international access code (different depending on where you are) and then everything after the +. An example: If I am in the USA and have to call +44 22 234 3213, then I would have to dial 011 (connects me into the international phone net) and then the country code 44, and then the phone number 22 234 3213 and if I wanted to be really fancy I could dial a # to tell the phone company that I was done with the number (no more digits). ** Leased line providers in the SF Bay Area Metropolitan Fiber Systems (+1 415 362 3300) <> Pac Bell (+1 800 974 2355) <> GTE (+1 800 487 5000) <> ** SF Bay Area IP providers SprintNet (+1 415 357 5500) <> BarrNet (+1 415 725 1790) <info(a)barrnet.net> AlterNet (+1 703 204 8000) <alternet-info(a)uunet.uu.net> Netcom (+1 408 554 8649) <info(a)netcom.com> ScruzNet (+1 408 457 5050) <info(a)scruz.net> NorthBayNet (+1 415 472 1600) <info(a)nbn.com> Internex (+1 415 473 3060) <internex(a)internex.net> Wombat Internet Guild (+1 415 462 8800) <> ** Some National IP Providers AlterNet Service area: US and international Contact: Alternet Sales Voice: 800-4UUNET3, 703-204-8000 email: alternet-info(a)uunet.uu.net FTP more info: ftp.uu.net:~info/alternet ANS CO+RE Systems, Inc. Service area: US and international Contact: Inside Sales Voice: 800-456-8267, 313-663-7610 email: info(a)ans.net FTP more info: ftp.ans.net:/pub BARRNet Service area: Northern & Central California Contact: R.J. Goldberg Voice: 415-723-7003, 415-322-0602 email: info(a)barrnet.net FTP more info: ftp.barrnet.net CERFnet Service area: California and International Contact: Sales Manager Voice: 800-876-2373, 619-455-3900 email: help(a)cerf.net FTP more info: nic.cerf.net Global Enterprise Services Service area: US and International Contact: Marketing Dept Voice: 800-35-TIGER email: market(a)jvnc.net FTP more info: N/A InterNex Information Services Services inc. Service_area: Palo Alto/San Jose California Voice: (+1 415 473 3060) Services: IP over ISDN, WWW homing, MSEN Service area: Michigan Voice: +1 313 998 4562 Fax: +1 313 998 4563 Snail: 320 Miller Ave. Ann Arbor MI 48103 WWW more info: www.msen.com NETCOM On-Line Communication Services, Inc. Service area: Nationwide Contact: Desirree Madison Voice: 408-554-8649 x2603 email: info(a)netcom.com FTP more info: ftp.netcom.com PSINet Service area: US and International Contact: Inside Sales Voice: 800-827-7482, 703-620-6651 email: info(a)psi.com FTP more info: ftp.psi.com, cd ~ SprintLink Service area: US Contact: Bob Doyle Voice: 703-904-2167 email: bdoyle(a)icm1.icp.net FTP more info: N/A The Little Garden Service area: Northern California Contact: Tom Jennings (admin(a)tlg.org) Voice: +1 415 487 1902 email: info(a)tlg.org FTP more info: ftp.tlg.org WWW more info: www.tlg.org ** Useful Books and Readings "Connecting to the Internet" by Susan Estrada published by O'Reilly & Associates, Inc. ISBN 1-56592-061-9 "The Online User's Encyclopedia: Bulletin Boards and Beyond" by Bernard Aboba published by Addison-Wesley ISBN 0-201-62214-9 "Practical Internetworking with TCP/IP and UNIX" by John S. Quarterman and Smoot Carl-Mitchell published by Addison-Wesley, Reading, MA (1993) "DNS and BIND" by Paul Albitz & Gricket Liu published by O'Reilly & Accosiates ISBN 1-565692-101-4 ** Government Much of the telecommunications system of the USA is still controlled by government, so many of the services of your RBOC are what they have been told to give you, which is often different from what the RBOC's would like to sell you and even more different from what you want. There are two bodys that control most of the telecom in the USA, the FCC (Federal Communications Commission) and your states PUC (Public Utility's Commission, here in California its the CPUC). Note that it is always interesting to go and READ the tariffs, as often you will find great deals in them the the RBOC's don't advertise... CPUC: General Information +1 415 703 1282, +1 800 848 5580 Complaints +1 415 703 1170 CPUC Tariffs: CPUC B5 ADN lines CPUC B9 High Capacity (T1 and the like) CPUC A18 Frame Relay? CPUC 175 More T1 stuff FCC: FCC-128 Cheaper leased line rates for out of state traffic ** Service Providers Here is a list of company's who provide various services on the internet: shell accounts, www pages homing, email, tele-conferencing and what ever the market can bare. Access InfoSystems Voice: (+1 707 442 1034) Email: info(a)commnuity.net Service_Area: Solano County (NE of the SF Bay area) Services: Internet Access, UseNet, IP, WWW pages Maillist, MUDs, etc. CCnet Service_Area: Contra Costa County (E of the SF Bay area) Voice: +1 510 988 0680 Email: info(a)ccnet.com The Well Services: Unix Shell Accounts, UUCP, Internet access, conferencing, UseNet Service_Area: Local to SF, Berkeley, Marin. Nationally via CPS network. Voice: +1 415 332 4335 Email: info(a)well.com Netcom Msen Internex See above. ** Datacom Hardware Capella Networking Voice: +1 415 591 3400 Service: Sells telecom equipmemt. Morning Star Technologies Inc. Voice: (+1 614 451 1883) (+1 800 558 7827) Online: www.morningstar.com or ftp.morningstar.com Service: Livingston Enterprises Voice: +1 800 458 9966 Service: Manufactures network routers and terminal servers. Cisco Systems Voice: +1 415 326 1941 or +1 800 553 6387 FAX: +1 415 326 1989 Online: http://www.cisco.com Service: Manufactures network routers and terminal servers. ** Beam Cast & Broad Cast systems More work needs to be done on private Beam and Broad cast systems. The state of the art (or at least it's use by us Internetworkers) is maybe a decade behind our use of leased lines, and we NEED this stuff. Here are some buzzwords to get you dreaming... Packet Radio IR Laser Mircowave Spread Spectrum There are some cool .8W SS Raido modems that can do 1.3Mbit over short distances (~5 miles) I have herd rumors of from Cylink?) Cylink corp. Voice: +1 408 735 5800 ** PacBell ISDN Questionare From: Rob Rustad <info-isdn(a)pacbell.com> Newsgroups: ba.internet Subject: PACIFIC BELL - ISDN FOR CALIFORNIA INTERNET USERS Date: 29 Apr 1994 00:08:44 GMT Organization: Pacific Bell Lines: 108 Distribution: world Message-ID: <2ppj6c$isv(a)gw.PacBell.COM> NNTP-Posting-Host: rjrusta.srv.pacbell.com Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-UserAgent: Nuntius v1.1.1d27 X-XXMessage-ID: <A9E590CAE00432F4(a)rjrusta.srv.pacbell.com> X-XXDate: Thu, 28 Apr 94 16:11:54 GMT PACIFIC BELL - ISDN FOR CALIFORNIA INTERNET USERS Hi. To better serve you, we at Pacific Bell are attempting to forecast demand for ISDN services within the Internet community of California. We would like to accelerate deployment of our digital services, but we need help in determining the level of future interest and need for those services among our California customers. Our current plan is to offer digital services to virtually all of our customers by 1997, but we would like to provide them sooner where a need exists. We would like to hear directly from you regarding your needs for higher speed digital access (digital 56kbps and above) to the Internet and other on- line services. The most important information for us is the area code and prefix of the telephone number(s) for which you probably would want ISDN services. Please help us by filling out any or all of the form below and e-mailing it to isdn-info(a)pacbell.com. ----------------------------------------------------------------------- -------------- To: isdn-info(a)pacbell.com Subject: ISDN Services Yes, I would like Pacific Bell to be able to offer me ISDN. Area Code: xxx Prefix (first three digits of your seven digit telephone number): xxx I would like to be able to communicate digitally with (IP Provider, other On- line Service, my company LAN, other):______________ Time frame - (ASAP, within 6 months, within 2 years): ____________ Other: ----------------------------------------------------------------------- -------------- The information you provide will be strictly confidential and you will not be contacted by Pacific Bell unless you indicate in your message that we may contact you. For additional help: ! isdn-info(a)pacbell.com ! 510-277-1037 - BBS with ISDN Information. ! Pacific Bell Gopher Server is available with ISDN information. ! 800-995-0346 - ISDN Availability Hotline (automated audio response) ! 800-662-0735 - ISDN Telemarketing (ordering information - please bear in mind, these people have no information on "pending" products like Home ISDN) Pacific Bell ISDN Basics: Pacific Bell has two ISDN Basic Rate Services to choose from -- SDS ISDN, Centrex ISDN. A third, Home ISDN, is a proposed product scheduled for availability on 8/2/94. All three provide circuit switched end-to-end digital connectivity for customers at speeds up to 112 kbps. Currently, over 60% of California has access to ISDN services. SDS ISDN Monthly Service: $26.85* Installation: $70.75** Usage: usage is billed at regular business voice rates on a per B-Channel basis * this price will be lowered to $22.85 on 8/2/94. ** a $150 installation fee is waived for a 2 year service commitment. Centrex ISDN Monthly Service: $31.65 Installation $225.00*** Usage: usage is flat-rated within the Centrex account. Outside the Centrex usage is billed at regular business rates. *** Centrex Establishment Charges apply for new systems $200. Home ISDN(Proposed)**** Monthly Service: $22.95 Installation $40.00 Usage: billed at regular business rates Mon-Fri 8am-5pm. All other time, zone 1 and 2 usage is flat-rated. **** proposed tariff pending CPUC approval, sales effective date estimated to be 8/2/94. ISDN Terminal Equipment: prices for ISDN terminal equipment have been dropping over the last year. It is currently possible to purchase a PC card, NT1 and power supply for under $500 that allows you to communicate at 112kbps uncompressed. For purposes of communicating over the Internet, it will be important to coordinate with your IP provider. ** Network Resources Here are some good WWW and FTP sites to explore on the net, looking though these sites can give you an idea of what networking is turning into. www.msen.com Good site to explore, lots of info www.tlg.org Good site to explore, lots of info ftp.tlg.org (same as the www.tlg.org site) ftp.netcom.com List of IP services /pub/miperrey/??? www.eit.com List of commercial services ftp.internic.net RFC's, policies and procedures files www.cisco.com Lots of good info on networking gw.pacbell.com Gopher site with digital telecom info. www.pacbell.com WWW Interface to the gopher site. *** UseNet NetNews Groups ba.internet Questions and announcement about accessing and using the internet in one of it's hottest hotbeds, the San Francisco Bays Areas comp.dcom.* Computer Data Telecom, each of the groups in this sub-tree deals with a different aspect of digital telecom. comp.dcom.telecom comp.dcom.telecom.tech General telecom questions, answers and talk on issues from the cheapest long distance service to what billing software interfaces to what phone switches. comp.dcom.modems Every stupid question (and some *real* good ones) you ever wanted asked and answered about modems. alt.dcom.telecom alt.dcom.telecom.ip These two groups are much like the comp.dcom.* groups, though a lot more free from and with information on telecom from parts of the world outside of the USA. biz.comp.services biz.comp.hardware The biz groups are where you can talk about services (ie. shamelessly advertise your service!). A good place to find products to use (hardware) or places that might do some of the work for you (services). alt.internet.services A great place to find out about what services are out there on the net already. Lots of talk about how services work, what are the different access methods, how to stop or start flame wars... comp.internet.* This is where internet protocols, usage and even its future are gone over an over and over... alt.security As Vernor Vinge said, it's "the net of a thousand lies...". Cyberspace needs to be secure, and here is one forum that goes into security of the net and the machines on it. ** Telecommunications speeds The first time I gave a talk based on these notes I spent close to a third of the time going over the various physical communications methods, modems, leased lines, fiber, beam and broad cast systems, and then the data standards you can run over these media. It seems that folks are very interested how how the basic telecom happens, but get very confused over all the possibilities. To attempt to clear things up a little here is a table of many of the various current methods of digital data telecom: Bits/Second Nick Names Wire Method Telecom Protocol ------------------------------------------------------------------- 0 Direct Current, little interesting data transmission... 50 50-600 are very old 75 110 134 134.5 150 200 300 POTS Modem Bell 103, V.21 600 1,200 1200 baud POTS Modem Bell 212A, V.22 1,800 2,400 2400 baud POTS Modem V.22bis 4,800 4.8 Kbps POTS Modem V.32, V.32bis 7,200 POTS Modem V.32bis 9,600 9.6 Kbaud POTS Modem V.32, V.32bis 12,000 POTS Modem V.32bis 14,400 14.4 Kbaud POTS Modem V.32bis 16,000 ISDN D channel used for signaling 19,200 A usefull benchmark 28,800 POTS Modem V.34, (aka. V.FAST) 38,400 56,000 Leased Line AND0, DS0 56,000 Frame Relay 56,000 ISDN PacBell's idea of 64k? 64,000 POTS Digital POTS voice line 64,000 ISDN (1 B channel) 112,000 ISDN 2 B channels minus a D channel 128,000 Frame Relay 128,000 ISDN (Fully both B channels) 238,000 AppleTalk (A useful benchmark) 384,000 Frame Relay 512,000 Frame Relay 1,017,000 SMDS 1,536,000 1.536 Mbps Frame Relay 1,544,000 1.544 Mbps Leased Line T1, ADN0, DS0 1,544,000 Frame Relay 3,088,000 Leased Line E1 ??? (uncommon)(2.048mbit?) 4,000,000 SMDS 10,000,000 10 Mega bit EtherNet (A useful benchmark) 10,000,000 SMDS 16,000,000 SMDS 25,000,000 SMDS 34,000,000 SMDS 45,000,000 Fiber|Coax T3, D3 60,000,000 Fiber|Coax? ATM ??? (soon?) 100,000,000 Fiber FDDI ??? (LAN) 135,000,000 3DS3 Microwave Standard 155,000,000 155Mb/s OC3c ATM 600,000,000 Fiber ATM ??? (someday?) 622,000,000 OC12c ATM 1,000,000,000 1 Gigabit Fiber... Future networks... 3,400,000,000 3.4x10^8 bps Fiber High Speed AT&T Trunks 20x10^12 20 Terabits Fiber Theoretical limit (ie. a guess) POTS is a two wire standard (a copper pair) Digital POTS, a voice line that gets digitised at the CO 8000 times a second with 8 bit samples. Leased lines are 4 wire standards (two copper pairs) Frame Relay is a leased line with a different telecom protocol on it that can run at many different speeds AppleTalk is a twisted pair LAN EtherNet is a Coax or twisted pair LAN standard FDDI is a Fiber optics (glass wire) standard ATM really is a protocol and not really a wire standard SONET SMDS Protocol levels: wire What sort of electricity/photons are used. telecom What the bits look like on the wire software What is done with the bits (IP, SNA etc.) ** Emacs Outline-Mode Note that the format here is GNU Emacs Outline mode, if you have emacs you can use outline mode on this document and it might be easyer to navigate. ;;; ;;; ;;;

1 0

quantum Computing
by geoffw＠nexsys.net 17 Dec '03

17 Dec '03

this term keeps poping up recently. Can anybody give me a pointer to where I can find out more info? Someone said that it is nonsense, "quantum computers?, Isn't that something out of a carlos casteneda novel?" I'm just trying to find out the real deal. FTP sites and any papers articles would be appreciated. G

1 0

Re: quantum Computing
by juola＠bruno.cs.colorado.edu 17 Dec '03

17 Dec '03

Rick Busdiecker writes: > Not true. What that means is that a polynomial time solution exists > for an NFA. The only part has not been shown. Mike McNally responds: >While we're being picky, I'll point out that (unless I'm wrong of >course) it's not really an NFA, but a non-deterministic Turing >machine (an "NTM"?) that's the automaton at issue here. and I then gibbered : That is correct. As a matter of fact, it's an easy theorem that an NFA has the same computing capacity as a DFA; it is not known whether this theorem holds for more powerful machines, and is in fact the heart of the P ?= NP conjecture. Whups. Teach me to post before eating breakfast.... Ignore what I just said above. - kitten

1 0

No Subject
by helmer＠interlock.svo.com 17 Dec '03

17 Dec '03

>From @interlock:ehelmer@ponder.csci.unt.edu Wed May 18 10:58:30 1994 Received: from interlock.svo.dfw.ibm.com by schwing.svo.dfw.ibm.com (AIX 3.2/UCB 5.64/4.03) id AA29020; Wed, 18 May 1994 10:58:28 -0500 Received: from ponder (ponder.csci.unt.edu) by interlock.svo.dfw.ibm.com with SMTP id AA11505 (InterLock SMTP Gateway 1.1 for <helmer(a)interlock.svo.com>); Thu, 19 May 1994 10:45:44 -0500 Received: by ponder (5.61/1.36) id AA03144; Wed, 18 May 94 11:01:46 -0500 Date: Wed, 18 May 94 11:01:46 -0500 From: @interlock:ehelmer@ponder.csci.unt.edu (Eric Helmer) Message-Id: <9405181601.AA03144@ponder> To: helmer(a)schwing.svo.dfw.ibm.com Status: R >From alt.cyberpunk Wed May 18 10:55:26 1994 Path: news.unt.edu!cs.utexas.edu!usc!nic-nac.CSU.net!clstac!achen1 From: achen1(a)csupomona.edu (ALAN CHEN) Newsgroups: alt.cyberpunk Subject: This could ruin your day. Date: 15 May 94 23:45:03 PST Organization: Cal Poly U. Pomona Lines: 237 Message-ID: <1994May15.234503.1@clstac> NNTP-Posting-Host: vmsa.is.csupomona.edu This is a document thats been kicking around our local system at Cal Poly for the last two days; I think its important enough to ask all of you to read it. Please take to time to do so as it is an issue that affects the future of the net as we know it. Alan Chen (Pre) Occupations: |"Never underestimate the power of human stupidity -Aerospace eng. undergrad | -L. Long (RAH) -RPG player |Yes, I'm an ARO. No, I'm not transferring! -Dreamer |Email: achen1(a)csupomona.edu GE -p+ c++ l(+) u- e+(*) m+(*) s !n+(-) h(-) f+ !g w-(--) t+ r++ y? -------------------------------- Subject: Metered Usage of the Internet: JSN Please forgive the mass mailing, but I feel this is a subject which is of great importance to anyone who benefits from the bountiful resources of the Internet. A very bad storm is brooding on the horizon. In the future, you might have to pay a charge for every E-mail message you send or receive, every Usenet article you read, every kilobyte of data you transfer with ftp, every hypertext link you follow with NCSA Mosaic or Gopher... Hopefully this frightens you as much as it does me. But it will happen, unless YOU do something about it. Please read the attached, fill out the requested info, and mail it back to mike(a)essential.org. It also wouldn't hurt to forward a copy of this to everyone you know on the Internet. Thanks for your support. Craig Smith, <bcs(a)cs.tamu.eduor <craig(a)stat.tamu.edu Texas A&M University, Dept. of Computer Science 205 HRBB, 862-2084 (CPSC). [PGP2 Public Key Available on Request] --- TAXPAYER ASSETS PROJECT - INFORMATION POLICY NOTE May 7, 1994 - Request for signatures for a letter to NSF opposing metered pricing of Internet usage - Please repost this request freely The letter will be sent to Steve Wolff, the Director of Networking and Communications for NSF. The purpose of the letter is to express a number of user concerns about the future of Internet pricing. NSF recently announced that is awarding five key contracts to telephone companies to operate four Internet "Network Access Points" (NAPs), and an NSF funded very high speed backbone (vBNS). There have been a number of indications that the telephone companies operating the NAPs will seek permission from NSF to price NAPs services according to some measure of Internet usage. The vBNS is expected to act as a testbed for new Internet pricing and accounting schemes. The letter expresses the view that metered pricing of Internet usage should be avoided, and that NSF should ensure that the free flow of information through Internet listserves and file server sites is preserved and enhanced. Jamie Love, Taxpayer Assets Project (love(a)essential.org; but unable to answer mail until May 15). Until then, direct inquires to Michael Ward. If you are willing to sign the letter, send the following information to Mike Ward of the Taxpayer Assets Project (mike(a)essential.org, fax: 202/234-5176; voice: 202/387-8030; P.O. Box 19367, Washington, DC 20036): Names: ___________________________ Title: ___________________________ (Optional) Affiliation: ____________________________________ (for purposes of identification only) Address: ______________________________________ City; St, Zip ________________________________ Email Address: _____________________________________ Voice: __________________________________ for verification) The letter follows: Steve Wolff Director Division of Networking and Communications National Science Foundation 1800 G Street Washington, DC 20550 Dear Steve: It is our understanding that the National Science Foundation (NSF) and other federal agencies are developing a new architecture for the Internet that will utilize four new Network Access Points (NAPs), which have been described as the new "cloverleaves" for the Internet. You have indicated that NSF is awarding contracts for four NAPs, which will be operated by telephone companies (Pac Bell, S.F.; Ameritech, Chicago; Sprint, NY; and MFS, Washington, DC). We further understand that NSF has selected MCI to operate its new very high speed backbone (vBNS) facility. There is broad public interest in the outcome of the negotiations between NSF and the companies that will operate the NAPs and vBNS. We are writing to ask that NSF consider the following objectives in its negotiations with these five firms: PRICING. We are concerned about the future pricing systems for Internet access and usage. Many users pay fixed rates for Internet connections, often based upon the bandwidth of the connection, and do not pay for network usage, such as the transfer of data using email, ftp, Gopher or Mosaic. It has been widely reported on certain Internet discussion groups, such as com-priv, that the operators of the NAPs are contemplating a system of usage based pricing. We are very concerned about any movement toward usage based pricing on the Internet, and we are particularly concerned about the future of the Internet Listserves, which allow broad democratic discourse on a wide range of issues. We believe that the continued existence and enhancement of the Internet discussion groups and distribution lists is so important that any pricing scheme for the NAPs that would endanger or restrict their use should be rejected by the NSF. It is important for NSF to recognize that the Internet is more than a network for scientific researchers or commercial transactions. It represents the most important new effort to expand democracy into a wide range of human endeavors. The open communication and the free flow of information have make government and private organizations more accountable, and allowed citizens to organize and debate the widest range of matters. Federal policy should be directed at expanding public access to the Internet, and it should reject efforts to introduce pricing schemes for Internet usage that would mimic commercial telephone networks or expensive private network services such as MCI mail. To put this into perspective, NSF officials must consider how any pricing mechanisms will change the economics of hosting an Internet electronic mail discussion groups and distribution lists. Many of these discussion groups and lists are very large, such as Humanist, GIS-L, CNI-Copyright, PACS-L, CPSR-Announce or Com-Priv. It is not unusual for a popular Internet discussion group to have several thousand members, and send out more than 100,000 email messages per day. These discussion groups and distribution lists are the backbones of democratic discourse on the Internet, and it is doubtful that they would survive if metered pricing of electronic mail is introduced on the Internet. Usage based pricing would also introduce a wide range of problems regarding the use of ftp, gopher and mosaic servers, since it conceivable that the persons who provide "free" information on servers would be asked to pay the costs of "sending" data to persons who request data. This would vastly increase the costs of operating a server site, and would likely eliminate many sources of data now "published" for free. We are also concerned about the types of accounting mechanisms which may be developed or deployed to facilitate usage based pricing schemes., which raise a number of concerns about personal privacy. Few Internet users are anxious to see a new system of "surveillance" that will allow the government or private data vendors to monitor and track individual usage of Information obtained from Internet listserves or fileserves. ANTI-COMPETITIVE PRACTICES We are also concerned about the potential for anti- competitive behavior by the firms that operate the NAPs. Since 1991 there have been a number of criticisms of ANS pricing practices, and concerns about issues such as price discrimination or preferential treatment are likely to become more important as the firms operating the NAPs become competitors of firms that must connect to the NAPs. We are particularly concerned about the announcements by PAC-Bell and Ameritech that they will enter the retail market for Internet services, since both firms were selected by NSF to operate NAPs. It is essential that the contracts signed by NSF include the strongest possible measures to insure that the operators of the NAPs do not unfairly discriminate against unaffiliated companies. Recommendations: As the Internet moves from the realm of the research community to a more vital part of the nation's information infrastructure, the NSF must ensure that its decisions reflect the needs and values of a much larger community. 1. The NSF contracts with the NAPs operators will include clauses that determine how the NAP services will be priced. It is important that NSF disclose and receive comment on all pricing proposals before they become final. NSF should create an online discussion list to facilitate public dialog on the pricing proposals, and NSF should identify its criteria for selecting a particular pricing mechanism, addressing the issue of how the pricing system will impact the Internet's role in facilitating democratic debate. 2. NSF should create a consumer advisory board which would include a broad cross section of consumer interests, including independent network service providers (NSPs), publishers of Internet discussion groups and distribution lists, academic networks, librarians, citizen groups and individual users. This advisory board should review a number of policy questions related to the operation of the Internet, including questions such as the NAP pricing, NAP operator disclosure of financial, technical and operational data, systems of Internet accounting which are being tested on the vBNS and other topics. 3. NSF should solicit public comment, though an online discussion group, of the types of safeguards against anticompetitive behavior by the NAPs which should be addressed in the NSF/NAPs contracts, and on issues such as NAPs pricing and Internet accounting systems. --------------------------------------------------------------------- TAP-INFO is an Internet Distribution List provided by the Taxpayer Assets Project (TAP). TAP was founded by Ralph Nader to monitor the management of government property, including information systems and data, government funded R&D, spectrum allocation and other government assets. TAP-INFO reports on TAP activities relating to federal information policy. tap-info is archived at ftp.cpsr.org; gopher.cpsr.org and wais.cpsr.org Subscription requests to tap-info to listserver(a)essential.org with the message: subscribe tap-info your name --------------------------------------------------------------------- Taxpayer Assets Project; P.O. Box 19367, Washington, DC 20036 v. 202/387-8030; f. 202/234-5176; internet: tap(a)essential.org ---------------------------------------------------------------------

1 0

PGP 2.5: Mini-review
by schirado＠lab.cc.wmich.edu 17 Dec '03

17 Dec '03

Having not seen any activity on the list for the last week or so (I hope everyone's busy writing code!), I figured I'd simultaneously check to see if the list still existed, and share some interesting excerpts from NEWFOR25.DOC, from the PGP 2.5 MIT-legit package. PGP 2.5 is apparently still written by Phil Zimmermann - at least, it purports to be - which in itself is a considerable relief to those of us who had no idea who was responsible. The source code is also available, as before, and I'm sure programmers the world over are even now poring through it in minute detail, looking for backdoors and such. I also expect we'll be hearing from them relatively soon, to tell us of the presence or absence of any suspicious code. Not being a programmer myself, I can only comment on a few aspects. First, there is this: [...] >[An] RSAREF limitation is that it cannot cope with keys longer than >1024 bits. PGP now prints a reasonably polite error message in such a >case. I recall someone mentioning at one point that increasing the size of a key beyond 1024 bits did not justify the increased computing time, but I do not recall the reason why. I believe the reasoning was not that it offered no additional security, but rather, that it was already difficult enough to crack 1K keys, and if you're really that worried about security, you should be tightening up in other areas, such as deciding who to trust and who not to, deciding what information to enter into the computer and what to keep in your head, or maybe making a homemade TEMPEST shield. :) I'd still like to see the math explained a little better, though. Also, has anyone found those references to elliptic-curve crypto? The original article is _An Implementation of Elliptic Curve Cryptosystems Over F-2-155_ , IEEE Journal on Selected Areas in Communications, Vol. 11, #5, June 1993 (page 804). (Schneier mentions that Next Computer's Fast Elliptic Encryption, FEE, uses elliptic curves, and is patented by R E Crandell, USP# 5,159,632,27 October 1992.) Also, look for works by Neal Koblitz. >Printed keyIDs have been incresed to 32 bits, as there were enough keys >out there that 24-bit keyIDs were no longer sufficiently unique. The >previous 24-bit keyID is the LAST 6 digits of an 8-digit 32-bit keyID. >For example, what was printed as A966DD now appears as C7A966DD. So even though the keyservers only have 5,000 or so registered users, there are enough people out there using PGP and NOT registering their keys with the servers that this extra bit of coding was necessary? Hmm. 24 bits gives us 16,777,216 unique ID's. 32 bits gives us 4,294,967,296. Are there really over 17 million PGP'ers out there, or is my math-impaired brain missing something painfully obvious? >PGP now enables clearsig by default. If you sign and ascii-armor a >text file, and do not encrypt it, it is clearsigned unless you ask >for this not to be done. Which would seem to indicate that PGP is mainly being used for e-mail! Goody! >[...] > >PGP now wipes temp files (and files wiped with pgp -w) with pseudo-random >data in an attempt to force disk compressors to overwrite as much data as >possible. > >[...] > >The normal help files (pgp -h) are pgp.hlp or <language>.hlp, such as >fr.hlp. Now, there is a separate help file for pgp -k, called pgpkey.hlp, >or <language>key.hlp. No file is provided by default; PGP will use >its one-page internal help by default, but you can create such a file >at your site. > >PGP used to get confused if you had a keyring containing signatures from >you, but not your public key. (PGP can't use the signatures in this case. >Only signatures from keys in the keyring are counted.) PGP still can't use >the signatures, but prints better warning messages. Also, adding a key on >your secret key ring to your public keyring now asks if the key should be >considered ultimately-trusted. Prviously, you had to run pgp -ke to force >this check, which was non-obvious. > >[...] > >On Unix, PGP now figures out the resolution of the system clock at run >time for the purpose of computing the amount of entropy in keystroke >timings. This means that on many Unix machines, less typing should be >required to generate keys. (SunOS and Linux especially.) > >The small prime table used in generating keys has been enlarged, which >should speed up key generation somewhat. > >There was a bug in PGP 2.3a (and, in fact in 2.4 and dating back to 1.0!) >when generating primes 2 bits over a multiple of the unit size (16 bits >on PC's, 32 bits on most larger computers), if the processor doesn't deal >with expressions like "1<<32" by producing a result of 1. In practice, >that corresponds to a key size of 64*x+4 bits. > >Code changes: > >At the request of Windows programmers, the PSTR() macro used to translate >string has been renamed to LANG(). > >The random-number code has been *thoroughly* cleaned up. So has the >IDEA code and the MD5 code. The MD5 code was developed from scratch and >is available for public use. So, all in all, PGP 2.5 would seem to be more than just a possible conspiracy by MIT/RSA/et. al., and more than just minor bug fixes that most people wouldn't care about. With the possible exceptions of the size limitations on keys, and whatever arcane pieces have been hacked out of the RSA code to comply with whatever demands they may have made, PGP 2.5 appears to be a legitimate upgrade, with more than a few bugfixes, both major and minor, as well as the all-important improved security (as far as can be seen). Comments? -- schirado(a)lab.cc.wmich.edu [O|o]bjectivist, Evil Capitalist(tm;-), s..O).... You hit the smurf! --More-- male, lesbian, polyamorous, @.../.".. You destroy the smurf! --More-- reader, atheist, Discordian, $$*...].. You feel cynical! free and natural sovereign individual the Frog Farm: e-mail frog-farm-request(a)blizzard.lcs.mit.edu (PGP available)

1 0

Re: quantum Computing
by gtoal＠an-teallach.com 17 Dec '03

17 Dec '03

this term keeps poping up recently. Can anybody give me a pointer to where I can find out more info? Someone said that it is nonsense, "quantum computers?, Isn't that something out of a carlos casteneda novel?" I'm just trying to find out the real deal. It's purest bullshit: there are a class of mathematically difficult problems called "NP-Complete". These problems are all equivalent to one another in difficulty, ie if you can solve one you can solve them all (that's where the complete part comes is - it's NP-complete if you can prove that equivalence to another NP-complete problem). The "NP" part is "Non-deterministic, polynomial time". What that means is that there is a solution possible in polynomial time (rather than exponential time) *ONLY* on a *NON-DETERMINISTIC* machine. And that's the fun part, because a non-deterministic machine is one that *guesses* the correct path every time it has a choice to make. It's like trying to guess a 3-bit number, and saying "Is the first bit a 1?" Yes! "Is the second bit a 0?" Yes! "Is the third bit a 0?" Yes! Clearly, in real life, this doesn't happen. However, in fairy-tale land (or quantum physics as it's called) such things *can* happen - because one interpretation of the Einstein-Podolsky-Rosen thought experiment is that every time you make a choice based on the outcome of a quantum event, you fork off a pair of universes! In one universe you make one choice; in the other universe you made the other choice. Consequently if you loose a computer on such a problem, in *one* of the many many universes it generates, it'll find the right answer in polynomial time. The basis of quantum computing as a means to crack NP-complete problems therefore reduces to finding which of these universes found the answer and comminicating that answer to all the other universes. (Of course, you don't have to do this part, but the 99.9999999999999999999999999999999% of experimenters in all the universes that didn't find the result are not going to believe the method words too well...) Basically, it's a theoretical result with no application in the real world, and if ever anything happens that makes it mappable to the real world we'll have been subjected to such a major upheaval in the way the universe works that no-one will give a damn any more about such trivial things as encryption because we'll all effectively have turned into magicians :-) G

5 10

Re: So PGP2.5 is becoming clearing...
by jamiel＠sybase.com 17 Dec '03

17 Dec '03

At 3:43 AM 05/18/94 -0700, John E. Kreznar wrote: >You would have to consent to be a national person (United States of >America or Canada) in order to have it. You wouldn't want to give up >your freeman status to do that, would you? Where do you live? Just curious. -j

1 0

Re: PGP 2.5: Mini-review
by gtoal＠an-teallach.com 17 Dec '03

17 Dec '03

: From: Schirado <schirado(a)lab.cc.wmich.edu> : >Printed keyIDs have been incresed to 32 bits, as there were enough keys : >out there that 24-bit keyIDs were no longer sufficiently unique. The : >previous 24-bit keyID is the LAST 6 digits of an 8-digit 32-bit keyID. : >For example, what was printed as A966DD now appears as C7A966DD. : So even though the keyservers only have 5,000 or so registered users, : there are enough people out there using PGP and NOT registering their : keys with the servers that this extra bit of coding was necessary? Hmm. : 24 bits gives us 16,777,216 unique ID's. 32 bits gives us 4,294,967,296. : Are there really over 17 million PGP'ers out there, or is my math-impaired : brain missing something painfully obvious? It's the old "birthday paradox" game. If you're generating numbers at random within a certain range, how many numbers do you have to generate before you have a probability >= 0.5 of generating two the same? Do it first for range = 0..2^24-1 and then for range = 0..2^32-1 ... G

1 0

Re: quantum Computing
by juola＠bruno.cs.colorado.edu 17 Dec '03

17 Dec '03

Rick Busdiecker writes: > Not true. What that means is that a polynomial time solution exists > for an NFA. The only part has not been shown. Mike McNally responds: >While we're being picky, I'll point out that (unless I'm wrong of >course) it's not really an NFA, but a non-deterministic Turing >machine (an "NTM"?) that's the automaton at issue here. That is correct. As a matter of fact, it's an easy theorem that an NFA has the same computing capacity as a DFA; it is not known whether this theorem holds for more powerful machines, and is in fact the heart of the P ?= NP conjecture.

2 1

Re: PGP 2.5 / PGP 2.6
by pcw＠access.digex.net 17 Dec '03

17 Dec '03

>On Wed, 18 May 1994, Dave Otto wrote: > >MIT's press release on 2.6 also mentions bugs in 2.5 that have been >pointed out to them. I'm all for liberating 2.5 from MIT but we need to >fix the bugs in it that they fixed in (or are fixing in) 2.6. If this is done successfully, it might be a good idea to name it PGP 2.7 for obvious reasons.

1 0