From: bill payne <billp@nmol.com>
Date: Wed, 26 Aug 1998 07:46:13 -0700 (PDT)
To: j orlin grabbe <abumujahid@taliban.com
Subject: Real World thoughts
Message-ID: <35E41E6E.C88@nmol.com>
MIME-Version: 1.0
Content-Type: text/plain


Wednesday 8/26/98 7:11 AM

J Orlin Grabbe

1  Tuesday issue of the ABQ J always carries a special section on
computers.

  t 8/25/98 

  COUNTDOWN TO Y2K

  Cataclysm or inconvenience?

Of most interest to me was

  Old mainframes make sheer size of problem nearly overwhelming

  By Chris Allbritton
  The Associated Press

  NEW YORK -  The shear number of unique solutions needed for the Year
  2000 bug, not the technical aspects, are the real problem behind this
glitch
  in time. ...

for the reason of my book and translation

  Machine, Assembly, and System Programming for the IBM 360, New York:  
  Harper & Row, 1969 
                   
  Programacion en Lenguaje de Maquina, Asemblador, y de Sistemas con el 
  IBM 360, Harper & Row, 1971

In the 1970s some IBM 360/70/40xx were STILL running IBM 650 simulators
which
hosted code which companies did not want to rewrite!

The message published in the Monday ABQ J

                            	'98 Warning!
  Due to serious bugs and compatibility problems, we join many other PC 
  builders in recommending AGAINST Windows '98 at this time.

http://www.c-works.com/

makes me speculate that the hardware vendors were able to try out their
new hardware
on DOS, Win 3.xx, 95, NT but not on '98.

Since Microsoft may not have had new hardware like the AMD K6 -2 [3d],
'98 was
not fully exercised on the newer architectures.

I will send you page on Y2K.

And also the copy of

  NSA  deputy director Barbara McNamara's  wrote me a letter dated 19
August  1998.

  http://www.nsa.gov:8080/mission.html

  McNamara DENIED my fee waiver appeal for search for invoices NSA spent
on public key chips.

  From my inside dealings with NSA's R division and Sandia employees  I
learned that both had     some VERY UNFORTUNATE EXPERIENCES with public
key.

  I am sending both you and John Young copies of NcNamara's letter.

  McNamara wrote,

    This response may be construed as a denial of your appeal. 
Accordingly, you are hereby 
  advised to your right to seek judicial review of my decision pursuant
to 5 U.S.C.  552(a)(4)
  (B) in the United States District Court in the district in which you
reside, in which you have
  your principal place of business, in which the Agency's records are
situated (U.S. District
  Court of Maryland), or in the District of Columbia. 

  http://www.jya.com/hr105-37.txt 

I wonder if http://www.taliban.com/ is the 'real thing' or an
'intelligence agency'
spy plot?  Since we send

  Date: 
        Tue, 25 Aug 1998 13:43:24 -0600
  From: 
        bill payne <billp@nmol.com>
    To: 
        dpcintrn@osd.pentagon.mil, abumujahid@taliban.com

it doesn't matter too much either way.  We either get both or the same
twice.

I am reading http://www.aci.net/kalliste/

  Impeachment Watch

               The High Cost of the War on Terrorism

              Bill Clinton's $100 Million Crusade Against Ken Starr

                                        by Charles R. Smith

      August was an expensive month for the U.S. taxpayer. For example,
Bill Clinton shot $100       million on August 20, 1998 at "terrorist"
sites in Sudan and Afghanistan. The 79 Tomahawk        cruise missiles
signaled a new "war" that Bill Clinton decided to declare on terrorism
and       not to be confused with the $100 million White House "war" on
Ken Starr.  ...

Last night I was CAREFULLY STARTING to read HARCOPY OF
http://www.softwar.net/plight.html especially  

  Sensor input is processed through the A/D converter and provided as
raw 8 bit data which is     then saved in WAVformat. A program designed
to extract randomization values from the raw data    for use as a cipher
key could takevarious forms. One form would be to serially store    
non-repeating samples of data into fixed length blocks for processing,
such as 8 or 16 bytes.    The sampled values would then be summed,
XORed, left or right shift registered (LSR/RSR), least   significant bit
extracted, or any combination of math processes. The resulting byte of
data   would then represent the combination of photon values taken over
the time period of   
  the sample. Other processes could make use of various formulas, such
as lossy or LZ, to give     still other values for photons taken over
the selected time period of the sample. 

  10 CLS:DEFINT A-Y:DEFDBL Z
  20 OPEN "R",#1,"TEST.WAV",1
  30 FIELD #1, 1 AS D$:J=1:O=0
  40 Z1 = LOF(1):ZR=64:A=0:O$=""
  50 ZR=ZR+1:IF ZR>Z1 THEN GOTO 100
  60 GET #1,ZR:I$=D$
  70 GOSUB 110:IF FA=0 THEN GOTO 50
  80 E=E+1:IF E>255 THEN E=0
  90 GOTO 50
  100 CLOSE #1:STOP
  110 FA=0:SK=ASC(I$)+SK:IF SK>32510 THEN SK=SK-32510
  120 IF S$<>I$ THEN GOTO 130 ELSE RETURN
  130 FA=1:S$=I$:A=A+1:O$=O$+I$
  140 IF A=16 THEN GOTO 160
  150 RETURN
  160 A=0:B=0:O= E XOR J:K=SK MOD 256:O = K XOR O
  170 FOR N=1 TO 16:X$=MID$(O$,N,1)
  180 B=ASC(X$):O = B XOR O
  190 NEXT N:O$=""
  200 IF O=J THEN RETURN
  210 PRINT O;:J=O
  220 RETURN

I STARTED to analyze what Smith's BASIC code did. 

I am STILL THINKING ABOUT the REASONS FOR

110 FA=0:SK=ASC(I$)+SK:IF SK>32510 THEN SK=SK-32510

which contains PROBABLY ZERO-initialized variable SK.

Smith will likely NOT be indicted by BXA  

http://www.bxa.doc.gov/FOIA/Foiaintro.htm
http://www.bxa.doc.gov/FOIA/Foialib.htm

for distributing controlled code on Internet.

The statement

  The sampled values would then be summed, XORed, left or right shift
registered (LSR/RSR), least 
  significant bit extracted, or any combination of math processes.

does NOT EXACTLY TELL what Smith is doing in PCYPHER ENCRYPTION USING
LIGHT.

But Smith's APPARENT belief that terminal bits of samples taken from
continuous distributions are, we believe, uniformly distributed.

    Payne, W. H., and T. G. Lewis, Continuous Distribution
    Sampling:  Accuracy and Speed, Mathematical Software,
    ed., J. R. Rice, Academic Press:  New York (1971), pp.
    331-345. 

    Lewis, T. G., and W. H. Payne, Generalized Feedback
    Shift Register Pseudorandom Number Algorithm, Journal
    of Assn. for Computing Machinery, 21, 3 (1973): 456-
    468. 

http://www-hto.usc.edu/software/seqaln/doc/html/gfsr.3.html
http://www.friction-free-economy.com       

    Sobolewski, J. S., and W. H. Payne, Pseudonoise with
    Arbitrary Amplitude Distribution:  Part I:  Theory,
    IEEE Transactions On Computers, 21 (1972): 337-345. 
                   
    Sobolewski, J. S., and W. H. Payne, Pseudonoise with
    Arbitrary Amplitude Distribution:  Park II:  Hardware
    Implementation, IEEE Transactions on Computers, 21
    (1972): 346-352. 

http://www.mhpcc.edu/general/john.html       

HOWEVER, us the the REAL PRACTICAL WORLD http://jya.com/da/whpda.htm
might be a bit concerned
what happens if the sensor output voltage exceeds EITHER in a plus or
minus direction the INPUT range of the a/d converter.

     http://www.softwar.net/plight.html
     Collector-Emitter Voltage 30 Volt 
     Emitter-Collector Voltage 5 Volt 

If the input voltage goes SOMEWHAT HIGHER, then the terminal bit MAY BE
A 1 and not random.

On the other hand, if the input voltage goes SOMEWHAT LOWER, then the
terminal bit MAY BE A 
0 and not random.

If the input voltages goes LOTS HIGHER OR LOWER than the limits, it is
SAYONARA a/d converter!
The Real World again.

SOME NSA KG schematics Brian Snow showed me had statistical test
hardware built in to ATTEMPT to
detect non-random output.  If the output failed the test, then the KG
unit shut down.

Smith, I think, has a neat idea which needs to be looked at VERY
CAREFULLY.

Some of us continue to feel that the crypto key must be as long as the
message encrypted.  As was pointed out by Gilbert S. Vernam and Joseph
O. Mauborgne in 1918.

Pseudorandom sequences DON'T WORK VERY WELL FOR ENCRYPTION.

  Pcypher Light is designed to measure the number of photons from around
your PC (light     particles) and create a random numbers from that
light. Pcypher Light is NOT just software but   hardware as well. It is
easy to install and use.  We give you SOFTWAR's advanced technology    
Light Probe, a highly sensitive phototransistor, which the software
samples at thousands of     times per second. Think of it as a
microphone for light. In fact all you have to do is plug it   directly
into the MIC input of any PC sound card. That is it. Run the PC Windows
software and   start making REAL randomized keys. NOT A PSEUDO KEY
GENERATOR.

Two reasons are

1  There should always be TWO KEYs.  The 'real key' and the key you
claim IN A LEGAL-SENSE is the 'real key' in case you get CAUGHT.

     A  Nuke them
     B  Send flowers

If you send A be sure to have a B key and a receipt for the flowers.

2  Pseudorandom sequences MAY make it difficult to find the SECOND key.

So let's hope these crypto messes 

http://www.aci.net/kalliste/speccoll.htm   
http://caq.com/cryptogate  
http://jya.com/whpfiles.htm  
http://www.qainfo.se/~lb/crypto_ag.htm

gets SETTLED SO THAT WE CAN ALL DEBATE TECHNICAL ISSUES and PUBLISH our
results in RECOGNIZED SCIENTIFIC JOURNALS and POST on the Infobahn too.

I want to revise my book for Windows and the 80C32 too.

Best
bill
http://www.apcatalog.com/cgi-bin/AP?ISBN=0125475705&LOCATION=US&FORM=FORM2         
http://jya.com/f86/whpf86.htm