[Noisebridge-discuss] Why do so many crackpots go for "crypto"?

Tue Jun 15 22:41:36 UTC 2010

Sorry to post so many times but I thought this may be helpful to those
trying to crack this nut.  Especially the use of tr, which I figured
out only after writing a perl program to do it - doh!

Here's the frequency graph I generated.

A = 0.101587, **********************************************
B = 0.014286, ******
C = 0.009524, ****
D = 0.017460, ********
E = 0.055556, *************************
F = 0.173016, ********************************************************************************
G = 0.061905, ****************************
H = 0.006349, **
I = 0.053968, ************************
J = 0.074603, **********************************
K = 0.009524, ****
L = 0.004762, **
M = 0.014286, ******
N = 0.025397, ***********
O = 0.046032, *********************
P = 0.017460, ********
Q = 0.128571, ***********************************************************
R = 0.003175, *
S = 0.022222, **********
T = 0.031746, **************
U = 0.007937, ***
V = 0.061905, ****************************
W = 0.026984, ************
X = 0.007937, ***
Y = 0.009524, ****
Z = 0.014286, ******

If this is English, it seems very likely F maps to e.

However, a straight frequency mapping is unhelpful, as I guessed,
since it's not a big enough sample to work:

$ tr FQAJGVEIOTWNSDPZMBCYKXUHLR etaoinshrdlcumwfgypbvkjxqz < 26_symbols.txt
netwananetyamalkahnlaogdasoni
etioriahcaoyeetipetetioi
thzihdehetbacsovunftreta
tgdesefwedetvsesnetimantah
unsetpialtrlsacoyoetbacso
ehacrwoeenalsciedoenetaniaera
etbacsoetraetpheteqsvtearpsa
mevnidoetrtjahoiusnmcaoeyo
marusdnhltaoeihxtetunsoehe
etilscoihjasetirxffahetyah
etbcsoetragietsehefldretcgen
neheltaajahdehetwtonoyoettjoen
etioitahenetirkctahihlhefsoi
etraueaieierrmsdlhksav
gdasocaetiweckoltaemsesoe
eteihaoetraoadunsoetuaoeibegt
netnqeseretilhaonxfnfeetuneas
etyarlehathopialuneasetunto
netradeyhettomrwtgtetrmad
tvurnftretaoyoettjoetigrndo
etehnespndeh
etinznlonetbacsoetraeheegrddoe
rwaokadonwrsei
etmauiwiuxtcehetiqsnla
etswai
dmiettoetswih

And actually, FQYANEJ isn't the longest gram, FQYANEJFQ is, it appears
twice:

$ grep FQYANEJFQ 26_symbols.txt
FQYANEJFQOAFQCIFQFLEKQFAOCEA
FQGVRVWJVFQYANEJFQOAFIFFMOTTJF

It has an interesting structure, too, because FQYANEJ appears several
more times by itself, suggesting that FQ might be a suffix of some
kind, as well as the beginning of the same word.  This is reinforced
by the second line, which ends in FQFQGJG, suggesting the end and
beginning of a word.  Let's try two-letter suffixes.

$ grep -i '^es.....es$' /usr/share/dict/words
escalates
escapades
escaroles
estimates
estranges
estuaries

None of those seem to work, tried them, and they're too weird for this
text (except maybe estimates; he might have been saying something
about our skill level being beyond it).

Could it be -et?

$ grep -i '^et.....et$' /usr/share/dict/words
$

Nope.  How about -en?

$ grep -i '^en.....en$' /usr/share/dict/words
enlighten

Tempting, but enlight isn't a word.

Based on this I'm guessing the "suffix" I saw is actually the
beginning of another word.  Of course, the second letter (Q) is less
sure than F=e.  It could easily be t, a, o, i, or the next few, as
suggested by this frequency graph:

[*] http://en.wikipedia.org/wiki/File:English_letter_frequency_%28frequency%29.svg

Also, notice that FQEPAG is on a line by itself (second to last
line)...  if it is one word, it is likely to be one of these:

$ grep -i '^et....$' /usr/share/dict/words
Etna's
Eton's
Etta's
etched
etcher
etches
ethics
ethnic

I also toyed with the idea that FQYANEJ = ethical.

$ tr FQYANEJ ethical < 26_symbols.txt
VetPiViVetBiDiWXiIVWilMTialVG
etGlOGiIcilBeetGCetetGlG
tIRGITeIethicalKSVZtOeti
tMTeaeZPeTetKaeaVetGDiVtiI
SVaetCGiWtOWaiclBlethical
eIicOPleeViWacGeTleVetiVGieOi
ethicaletOietCIeteLaKteiOCai
DeKVGTletOtUiIlGSaVDcileBl
DiOSaTVIWtileGIHtetSValeIe
etGWaclGIUiaetGOHZZiIetBiI
ethcaletOiMGetaeIeZWTOetcMeV
VeIeWtiiUiITeIetPtlVlBlettUleV
etGlGtiIeVetGOXctiIGIWIeZalG
etOiSeiGeGeOODaTWIXaiK
MTialcietGPecXlWtieDaeale
eteGIiletOiliTSValetSileGheMt
VetVLeaeOetGWIilVHZVZeetSVeia
etBiOWeIitIlCGiWSVeiaetSVtl
VetOiTeBIettlDOPtMtetODiT
tKSOVZtOetilBlettUletGMOVTl
eteIVeaCVTeI
etGVRVWlVethicaletOieIeeMOTTle
OPilXiTlVPOaeG
etDiSGPGSHtceIetGLaVWi
etaPiG
TDGettletaPGI

Hmm, that doesn't seem too promising... the last line might contain
"settle", but in that case, the second line is nonsense:

$ tr FQYANEJG ethicals < 26_symbols.txt | head -2 | tail -1
etslOsiIcilBeetsCetetsls

Also, it's VERY likely he mentions encrypt, and the first letters seem
plausible, so let's try that:

$ tr FQYANEJ encrypt < 26_symbols.txt
VenPrVrVenBrDrWXrIVWrtMTrptVG
enGtOGrIyrtBeenGCenenGtG
nIRGITeIencryptKSVZnOenr
nMTepeZPeTenKpepVenGDrVnrI
SVpenCGrWnOWprytBtencrypt
eIryOPteeVrWpyGeTteVenrVGreOr
encryptenOrenCIeneLpKnerOCpr
DeKVGTtenOnUrItGSpVDyrteBt
DrOSpTVIWnrteGIHnenSVpteIe
enGWpytGIUrpenGOHZZrIenBrI
encyptenOrMGenpeIeZWTOenyMeV
^^^^^^^^
VeIeWnrrUrITeIenPntVtBtennUteV
enGtGnrIeVenGOXynrIGIWIeZptG
enOrSerGeGeOODpTWIXprK
MTrptyrenGPeyXtWnreDpepte
eneGIrtenOrtrTSVptenSrteGceMn
VenVLepeOenGWIrtVHZVZeenSVerp
enBrOWeIrnItCGrWSVerpenSVnt
VenOrTeBIenntDOPnMnenODrT
nKSOVZnOenrtBtennUtenGMOVTt
eneIVepCVTeI
enGVRVWtVencryptenOreIeeMOTTte
OPrtXrTtVPOpeG
enDrSGPGSHnyeIenGLpVWr
enpPrG
TDGenntenpPGI
   ^^^^^^^

Well, either this has some errors in the original, or that can't be
the case.

Oh, incidentally, while FQYANEJ appears several times, and FQYANEJFQ
twice, there's also an appearance of FQYNEJFQ at the beginning of a
line.  I've double-checked the source and it's not a transcription
error.  I don't think that pattern is likely to be a coincidence.

Perhaps the missing symbol, A, is a single-letter word (a/I) that is
left out once.

Another interesting thing, is that twelve lines begin with FQ!

$ grep ^FQ 26_symbols.txt
FQGJOGAINAJBFFQGCFQFQGJG
FQYANEJFQOAFQCIFQFLEKQFAOCEA
FQGWENJGIUAEFQGOHZZAIFQBAI
FQYNEJFQOAMGFQEFIFZWTOFQNMFV
FQGJGQAIFVFQGOXNQAIGIWIFZEJG
FQOASFAGFGFOODETWIXEAK
FQFGIAJFQOAJATSVEJFQSAJFGYFMQ
FQBAOWFIAQIJCGAWSVFAEFQSVQJ
FQFIVFECVTFI
FQGVRVWJVFQYANEJFQOAFIFFMOTTJF
FQDASGPGSHQNFIFQGLEVWA
FQEPAG

If we assume that words aren't broken across lines, then FQ has to be
a VERY common way to begin a word, or is a word unto itself.

Possible explanations:

1) it's a very common two-letter word in English...
   in, an, so, if, is, it, possibly others.

2) it's Arabic, and translates to "al" (the, maybe of...)

3) it's some extremely common prefix in English, but what?
   in?  un?  Nothing seems that common.

Frankly, I'm not sure this is English!  After all, what are the
chances a random text of this length (650ish characters) would use Q
and Z?  But maybe he did that just to exercise the alphabet.

Frankly, given the preponderance of "FQ" at the beginning of the line,
and the diversity of things that follow it, I'm beginning to think it
might be Arabic.

Okay, that's enough for me for now, I'm taking a break. ;-)

HTH,
Travis
-- 
A Weapon of Mass Construction
My emails do not have attachments; it's a digital signature that your mail
program doesn't understand. | http://www.subspacefield.org/~travis/ 
If you are a spammer, please email john at subspacefield.org to get blacklisted.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://lists.noisebridge.net/pipermail/noisebridge-discuss/attachments/20100615/cb35d1ff/attachment-0003.sig>