Author here again: and thank you Tom for your feedback. I'm not the author of Screamer so am not going to defend it's runtime properties. These were not of interest to me and yes I could recode much of what I did in C++ which I am more familiar with than Prolog. I work in low latency trading, so I am happy to take any challenge from Prolog in C++. Bummer would be that I could not publish my solution. The reason Screamer is in there is to allow the problem to be expressed declaratively in terms of universal quantification. See also
> The reason Screamer is in there is to allow the problem to be expressed declaratively in terms of universal quantification.
Sure, though this particular example doesn't show its strengths very well. Your pre-analysis is so powerful that even in the case where you start with the empty board, any hand-written primitive brute force enumeration technique should be blazingly fast.
> I work in low latency trading, so I am happy to take any challenge from Prolog in C++.
I'm not saying Prolog is always faster than C++ or Common Lisp :-) I guess I'm saying that every embedding of a Prolog-like system in some other language I have seen so far was less powerful, less elegant, and much less performant than Prolog. This is only partly gloating from a Prolog fan. Another part of me would very much like to have better such systems integrated with other languages.
One drawback of Prolog, of course, is that Prolog can be a bit painful to interface with the rest of your application. Unless, of course, you write everything in Prolog...
Author here: Hello 60654, I fully understand the difference between cracking a run length encoding problem and finding weaknesses in DES. There are many tools at the disposal of a security analyst. Another would be CSP and FDR: https://www.google.com.au/url?sa=t&source=web&rct=j&url=http.... However, that article, like mine, goes to show how logic programming can very well be used to mount cryptanalytic attacks. It differs from quantitative statistical methods as these exploit different weaknesses. Qualitative vs quantitative. Most of the people with skills in the latter work in quantitative trading and hedge funds - including yours truly.
Author here: Ok thank you Tom for your feedback. "Lambda closures" is a term that is ubiquitous in C++, chiefly owing to how closures are implemented in that language. Sorry if that confused non developers. An "attack vector" is any method, approach or means by which a analyst gains an advantage against the security of a system. A more formal definition maybe found here. http://searchsecurity.techtarget.com/definition/attack-vecto... I broadly used the term to denote building up known plaintext ahead of the application of the solver.
I am happy to announce that a full Java "eco system" has been ported to IOS, the operating system powering the popular Iphone and IPad platform. This includes JIT based compilation, a graphical Swing subsystem and a development toolchain. The aim of this Java "eco system" is to deploy essentially unmodified Java applications on the IPhone and IPad - along with software written in the many other languages emerging on the Java virtual machine platform(JVM), such as Scala & Clojure. The significance is that there are many more existing, free Java applications than exist on the IPhone and IPad today. After all, the mantra of Java has always been "compile once, run anywhere."
https://chriskohlhepp.wordpress.com/reasoning-systems/specif...