Talk:Solitaire (cipher)

From Wikipedia, the free encyclopedia

This article is part of WikiProject Cryptography, an attempt to build a comprehensive and detailed guide to cryptography on Wikipedia. If you would like to participate, you can choose to edit the article attached to this page, or visit the project page, where you can join the project and see a list of open tasks.

I'd like to add a lot more to this page, including a Wikicode version, but I'm not sure about the copyright issues. There are several implementations available from Bruce's web page, and most of the are licensed using the GPL. I've also personally ported the C version to PHP and submitted it back to Bruce, but I never got a response back from him and it's never been added to that page.

Can I just go ahead and create the Wikicode and put it in the article?

--Jachim69 06:08, Apr 8, 2005 (UTC)

Good work on starting this page, and please do add more, if you get the chance. I'm not a lawyer, but (as I understand it), as long as your Wikicode isn't a "derived work" of some GPL code, then it should be OK. I would think that if you understand how the algorithm works, and are simply describing it in pseudocode, there shouldn't be any problem. — Matt Crypto 00:08, 18 Apr 2005 (UTC)

Bruce has sort-of lost interest in Solitaire because of the bias in the output. The best known hand cipher is currently RC4-52.

Also I personally can't see the point of Wikicode when Python is equally readable but has an interpreter, but it's up to you. — ciphergoth 08:17, 2005 Apr 18 (UTC)

What about discarding every second letter? This get rid of the "consecutive identical letters" bias. Of course it make the process twice slower, there may be better ways. Are there other bias ? - 12:10, August 21, 2005 140.77.13.62

I don't think this proposal would lead to a strong or efficient hand cipher. The lessons of Memo to the Amateur Cipher Designer apply to one applying "tweaks" to an existing cipher as much as to one designing a new cipher from scratch. — ciphergoth 15:16, August 21, 2005 (UTC)

Unless I'm misinterpreting the problem, discarding alternate letters would only halve the bias.

No, it would have a stronger effect than that, something more like the square but I don't have time to work it out just now — ciphergoth 06:20, August 25, 2005 (UTC)

Regarding the reversibility problem, suppose one changed the first two steps to:

1. Move the A joker one card down. If the joker is the bottom card, move it to the top.
2. Move the B joker two cards down. If the joker is the bottom card, move it below the top card. If the joker is the second card from the bottom, move it to the top.

That is, basically treat the deck as having an invisible "dummy" card between the top and bottom, much like a dummy node is used in a circular doubly-linked list to eliminate special cases. Wouldn't this make the operation fully reversible? Or would it screw up the algorithm in other ways? — Xaonon (Talk) 03:50, August 25, 2005 (UTC)

Yes, apparently this was what Bruce meant to propose in the first place but he got it wrong. However, because of the bias he didn't pursue it; the bias is present in this modified version also. — ciphergoth 06:20, August 25, 2005 (UTC)

--bsjeep 09:50, Aug 17, 2006 (PST)

The 2 decks vs. 4 decks discussion in the main article is confusing, and after reading the detailed writeup on the Author's website, it appears that the example in the article is only using 1 deck. Can someone confirm and clarify the main article...