Enigma and astronomical clocks?

[JT]

How does the enigma machine compare in complexity with the 1400
centuries astronomical clocks, with their different reels of
transmission, could one use a horologi clock reencode the reels and
get a farily complex cipher from it?

http://upload.wikimedia.org/wikipedia/commons/7/7d/Gda%C5%84sk_Bazylika_Mariacka_%28zegar_astronomiczny%29.jpg
http://upload.wikimedia.org/wikipedia/commons/4/40/Schema_Orloj_en.png
http://upload.wikimedia.org/wikipedia/commons/c/c4/Strasbourg_Astronomical_clock_Equations.jpg
http://en.wikipedia.org/wiki/File:Strasbourg_Astronomical_clock_Comput_ecclesiastique.jpg

Do anyone know if there is any engineering plans or description of
horologi clocks on internet?
It would be fun to try building something using programming and
replicate the dials and turning faces
of the horologi clocks and maybe presenting a crypto using the
innerworks of them.

[JT]

On 5 Maj, 16:35, JT <jonas.thornv...@gmail.com> wrote:
> How does the enigma machine compare in complexity with the 1400
> centuries astronomical clocks, with their different reels of
> transmission, could one use a horologi clock reencode the reels and
> get a farily complex cipher from it?
>

> http://upload.wikimedia.org/wikipedia/commons/7/7d/Gda%C5%84sk_Bazyli...http://upload.wikimedia.org/wikipedia/commons/4/40/Schema_Orloj_en.pnghttp://upload.wikimedia.org/wikipedia/commons/c/c4/Strasbourg_Astrono...http://en.wikipedia.org/wiki/File:Strasbourg_Astronomical_clock_Compu...

>
> Do anyone know if there is any engineering plans or description of
> horologi clocks on internet?
> It would be fun to try building something using programming and
> replicate the dials and turning faces
> of the horologi clocks and maybe presenting a crypto using the
> innerworks of them.

If we use three gearwheels linear connected, we suppose a 256
character set on each so each have 256 connection edges, the first
reels character sets entyping order is known to everyone, however reel
2 and 3 have their characters sets order scrambled by a permutation
algorithm, the output of the gearwheels is xor together, can we assume
uniqness and that the xores is reversible?

Now if we put a 4 th wheel between them generated by a hash from the
key, let us say 10-bit, it would be feasible to have gearwheels
ranging from 3-1024 edges.
How hard would this be to break mathematicly. 2^256^1024 i just
throwed this one out?

[JT]

On 8 Maj, 11:07, JT <jonas.thornv...@gmail.com> wrote:
> On 5 Maj, 16:35, JT <jonas.thornv...@gmail.com> wrote:
>
> > How does the enigma machine compare in complexity with the 1400
> > centuries astronomical clocks, with their different reels of
> > transmission, could one use a horologi clock reencode the reels and
> > get a farily complex cipher from it?
>

> >http://upload.wikimedia.org/wikipedia/commons/7/7d/Gda%C5%84sk_Bazyli......

>
> > Do anyone know if there is any engineering plans or description of
> > horologi clocks on internet?
> > It would be fun to try building something using programming and
> > replicate the dials and turning faces
> > of the horologi clocks and maybe presenting a crypto using the
> > innerworks of them.
>
> If we use three gearwheels linear connected, we suppose a 256
> character set on each so each have 256 connection edges, the first
> reels character sets entyping order is known to everyone, however reel
> 2 and 3 have their characters sets order scrambled by a permutation
> algorithm, the output of the gearwheels is xor together, can we assume
> uniqness and that the xores is reversible?
>
> Now if we put a 4 th wheel between them generated by a hash from the
> key, let us say 10-bit, it would be feasible to have gearwheels
> ranging from 3-1024 edges.
> How hard would this be to break mathematicly. 2^256^1024 i just
> throwed this one out?

Sorry two wheels between is of course needed to make a transmission
between wheels to 3/1024 to 1021/1024 so we need transmission wheels
ranging from 3 to 1024?
Ok so first wheel is 256! positions and so are second wheel this means
256!*256! and then we have a transmission of 1024-6 positions so the
then this ancient crypto construction would have.
So the transmission have two wheels each ranging from 3-1024, i know
to little about clocks to know if this is possible? this gives
1021*1020/2 gears?
256!*(1021*1020/2)*256! states?
Of course they would need to agree upon password and know the
permutation algorithm and password to construct the displaced
character sets on gearwheel 2 and 3, while we just program it
nowadays.

So qustion is whould it work, maybe it is not that slow although the
transmission between gearwheels a bit slow, i guess that boils down to
using add upon an array?

[JT]

A typewriter wheel to turn, and two permutated wheels 256 tooths. And
the two transmission gears ranging from 3-1024 tooth.

Construction
Typewriter wheel ->permutated character set wheel 1 ->transmission
wheel 1->transmission wheel 2->permutated character set wheel 2

Function permutation wheel 1 XOR permutation wheel 2.

[JT]

Better or worse then enigma?

[JT]

256!^2*((1021*1020)/2) states?

1020 decimal digits/states?? not bad for an old clock is it, well
maybe my calculation was a bit to optimistic.

383165149009297550708020354080888465962865652661352815210998945311781584864837847756665815061273591642343952183864176800855800203828152285565854833771140606707561995322854720015394314126188431797152909993728121731456387334932029190060440803069769037284894385370462799298157754812497902941859885877996079801845472456131266055216204381323353241866536259847480322179922343120545441777383921649409428320514582972852115424909249141991336115624459320341387364388534737576356235990895705702019779933832992381956941364213022166937900792321250333637138145925362058617247133867432236819006879084688550787603195832046010489706249849768071717509193087683030732418113121513599510497376100438604677623813659562016864306687850050892467552922525665165981885558829719184470874127555913891532605651056862679451704146335443120078646579901841629893127960069992948742384063737446500489952877814273266607745830748160000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000]

Actually would be quite a fun clock to have the alphabet encoded
around clockface, the knob have 3 extra modes one for insertion of
password,encoding,decoding in the date characters is instead shown.
But howto implement a XOR by using wheels beat me, and the
permutation
seem hard to to by wheels. It is much easier with big wheels where
you
simply can attach stickers for each tooth. Have there been old
spywatches with encryption ?

[JT]

So how to solve getting 256 characters in the date, well use 8 thinn
plastics, each with an open segment and let them them rotate
sequentially and all characters can be displayed in one box.

[JT]

But a XOR doesn't work does it, have to come up with something else
unless i can come up with a wheel that show the xor value wich seem
pretty impossible, but can we get a wheel that somehow show aspects of
the two encoded wheels?
Can you make a wheel that is dependent upon two wheels with different
reciprocations?

[JT]

Ooops i did again, i was wrong. The machine is 100% working, as a
keystream generator........... Both encrypter and decrypter turn the
reels right one char at a time XOR wheel one and two, this will
generate a keystream that is used to XOR both plaintext into
ciphertext and XOR ciphertext into plaintext. Behold the amazing 1400
keystreamer made from 5 wheels, well it isn't perfect it has
geardependent period. Isn't that smart having range from 3 to 1024
better 1024 to 2048 tooths beholde the amazing keystreamer with a long
period?

I know you guys do not like the term POTP put this is as POTP they
could have made the 1400 century using 5 wheels.
I think this is stronger or is it weaker then enigma? It of course
depends alot on the mechanism that reencode the characterset upon the
two wheels by using the key. But it easy to see it does require alot
to create a strong cipher.

[JT]

It turns out a XOR is possible with gearwheels.
http://www.youtube.com/watch?v=Y2UVb7QfMqY

[JT]

Could be fun and certainly educational to use LEGO to construct the
early mechanical calculators? Probably some patents come in the way,
but maybe a lego original, construct the primitive arithmetic
operations and put together.
And why not ciphermachines out of LEGO

[JT]

Building horology clocks out of LEGO do not seem that impossible, of
course they would hardly last half a millenia, but a little plastic
with siliconcoating goes a long way, and if you could make the
gearwheels out of carbon fibers, they would probably outlast the ones
of metal.
http://www.suspart.com/news/carbon-fiber-gear-wheel-research

[JT]

What would be real fun a shaft and gear wheel emulator where you just
throw out 3D gearwheels on a sheet tell how many tooths, resize them
with mouse or fingers an outofit function to make them connect. Build
your own watch emulator ;D

[JT]

It also would be nice if the emulator did not fuck up at the of
tooths n = -1. 0.49999999999= (n/2-1)/n, because i can not stand bad
arithmetic either in Mathematica, Wolfram or in engineering.
Todays mathematicians a bit dunse i guess.

[Jeffrey Goldberg]

On 2013-05-08, JT <jonas.t...@gmail.com> wrote:

> Better or worse then enigma?

Leaving the Enigma Reflector aside for the moment, I'd say that
your scheme (which I frankly only skimmed over) is no different
than Engima except that it uses bigger numbers (bigger rotors).

The analytical attacks, and in particular the vulnerability to known
plaintext attacks, remains.

The only difference I see (again, I didn't not do more than just
skim over what you wrote) is that your machine doesn't have
reflector that Engima had. This gives you an advantage, as the
reflector made it easier to find known plaintext. The reflector
meant that no letter ever encrypted to itself, so you could line up
some suspected plaintext with the ciphertext and if no letter ever
encrypted to itself, then you would know that your candidate
plaintext remained a possibility.

Please keep in mind that Engima was broken a long time ago, using
paper and pencil and then some electro-mechaninical devices that ran
through the reduced set of possibilities that the paper and pencil
analysis generated.

If you want to develop a system based on it, you will need to fully
understand the nature of the attacks against the original.

Cheers,

-j

--
Jeffrey Goldberg http://www.goldmark.org/jeff/
I rarely read top-posted, over-quoting or HTML postings.
Reply-To address is valid.