Page 2 of 3

Re: 2017 Symposium on Cryptologic History

PostPosted: Mon Oct 30, 2017 1:10 am
by Mr lowe
Great presentation Dave

Re: 2017 Symposium on Cryptologic History

PostPosted: Mon Oct 30, 2017 3:39 am
by Jarlve
Very good, actual and clear presentation doranchak!

This is really the kind of information I want people to see when they google for the 340.

All questions answered perfectly, well done.

Re: 2017 Symposium on Cryptologic History

PostPosted: Mon Oct 30, 2017 9:04 am
by Largo
Thank you very much for uploading the video! The presentation was really good and very informative. I especially liked the nice animations.

Re: 2017 Symposium on Cryptologic History

PostPosted: Mon Oct 30, 2017 1:21 pm
by doranchak
smokie treats wrote:I really liked the presentation. You covered the significant observations. I think that it was wise to not discuss hypotheses for the observations, especially the P15/19 repeats. You didn't have a lot of time, and without the hypotheses maybe someone else will think up some fresh ideas for a cipher. Good job.

Thanks. Yes, it would have been too much to include all the various hypotheses. The talk already had too much detail; I was already afraid of boring everyone with too many numbers and other technical info. Hopefully someone else will indeed think up some fresh ideas!

Re: 2017 Symposium on Cryptologic History

PostPosted: Mon Oct 30, 2017 1:22 pm
by doranchak
traveller1st wrote:Very good & interesting. So much easier than having to read stuff. Glad my repeating pattern shape thingy was of use for something. At least I think it was one of mine. :lol:

Thanks. Yes, the "trav patterns" were pretty useful! :) Sorry they didn't turn out to be significant features - but they were fun to analyze.

Re: 2017 Symposium on Cryptologic History

PostPosted: Mon Oct 30, 2017 1:24 pm
by doranchak
Mr lowe wrote:Great presentation Dave

Thank you very much!
Jarlve wrote:Very good, actual and clear presentation doranchak!
This is really the kind of information I want people to see when they google for the 340.
All questions answered perfectly, well done.

Thanks Jarlve. You and others here did a lot of the work in finding and analyzing the features I talked out in the presentation, and I am very grateful to you!

Re: 2017 Symposium on Cryptologic History

PostPosted: Mon Oct 30, 2017 1:26 pm
by doranchak
Largo wrote:Thank you very much for uploading the video! The presentation was really good and very informative. I especially liked the nice animations.

Thanks - the animations were fun to work on. It was much easier to show certain ideas instead of trying to explain them in detail.

Re: 2017 Symposium on Cryptologic History

PostPosted: Tue Oct 31, 2017 12:29 pm
by Tahoe27
Great job. You are very well-spoken. All of your experience speaking is showing! I must say, your added humor is a nice touch. ;)

Re: 2017 Symposium on Cryptologic History

PostPosted: Tue Oct 31, 2017 12:30 pm
by Tahoe27
Hey, you mention that diagonal shuffle using 19 characters. (Sorry if I'm not using the right wording). Did you try that with 17?

Re: 2017 Symposium on Cryptologic History

PostPosted: Tue Oct 31, 2017 8:33 pm
by doranchak
Tahoe27 wrote:Great job. You are very well-spoken. All of your experience speaking is showing! I must say, your added humor is a nice touch. ;)

Really nice of you to say - thanks!
Tahoe27 wrote:Hey, you mention that diagonal shuffle using 19 characters. (Sorry if I'm not using the right wording). Did you try that with 17?

Yes, we've tried other periods; 17 doesn't stand out as much as period 19 when it comes to those repeating bigrams. And I think someone's tried all period values in an automatic solver such as AZdecrypt without any luck.

"17" does have some significance in another area (besides being the width of the cipher grid). It has to do with the length of ciphertext segments that have no repeating symbols in them. So imagine starting at the top left of the cipher and counting how far you get until you hit a symbol you've already seen. Write down the length of the segment (excluding the repeating symbol). Then repeat from the 2nd position of the cipher. Then the 3rd. And so on.

Eventually you'll have a pile of numbers, each representing a length of a segment that has no repeating symbols. You'll find that in the pile of numbers, there are more 17s than any other number. That's a pretty interesting feature that Jarlve discovered.