tag:blogger.com,1999:blog-10770855.post5944780798443551725..comments2024-03-28T03:20:57.393-04:00Comments on The Little Calculist: Effect masking vs. purityDave Hermanhttp://www.blogger.com/profile/00405190527081772997noreply@blogger.comBlogger4125tag:blogger.com,1999:blog-10770855.post-21310521402507378292009-05-15T05:09:00.000-04:002009-05-15T05:09:00.000-04:00You can make rolling back state be extremely simpl...You can make rolling back state be extremely simple during backtracking by using a persistent datastructure. The only mutating state left then would be a stack which points to successive 'mutations' of the persistent structure.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-10770855.post-36714000120051652022009-05-15T01:17:00.000-04:002009-05-15T01:17:00.000-04:00I've had good results using the approach from "A S...I've had good results using the approach from "A Semi-Functional Implementation of a Higher-Order Logic Programming Language" by Conal Elliott and Frank Pfenning.Andrewhttps://www.blogger.com/profile/09614914868976063174noreply@blogger.comtag:blogger.com,1999:blog-10770855.post-46555265576642821892009-05-14T23:01:00.000-04:002009-05-14T23:01:00.000-04:00It depends on the implementation of mutation. You ...It depends on the implementation of mutation. You can presumably build "masking" in, by e.g. firing a search with a copy of the current store, and discard the copy on backtrack. There's a bunch of stuff on doing this efficiently, in logic programming (dating back to CT Haynes in the 80s) and first-class stores (as duals to first-class continuations). <br /><br />MariusAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-10770855.post-4203888459746598192009-05-14T21:53:00.000-04:002009-05-14T21:53:00.000-04:00Hey Dave,
What you're talking about here was pret...Hey Dave,<br /><br />What you're talking about here was pretty much exactly my motivation for "worlds". Did I ever tell you about that stuff? If you're interested, take a look at Chapter 4 of <A HREF="http://www.vpri.org/pdf/tr2008003_experimenting.pdf" REL="nofollow">my dissertation</A>. I'd love to chat with you about this stuff at some point...<br /><br />Cheers,<br />AlexAlessandro Warthhttps://www.blogger.com/profile/05850439404872913450noreply@blogger.com