Path: news.net.uni-c.dk!newsfeeds.net.uni-c.dk!newsfeed1.uni2.dk!news.get2net.dk!not-for-mail From: "Jim Lucas" Newsgroups: comp.ai.neural-nets,comp.lang.apl,comp.lang.awk,comp.lang.beta,comp.lang.cobol,comp.lang.dylan,comp.lang.forth References: <3AABD82E.C4D4E5CB@Home.Com> Subject: Re: Einstein's Riddle Lines: 51 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.00.2919.6600 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6600 Message-ID: <4rVq6.559$ou4.12675@news.get2net.dk> Date: Mon, 12 Mar 2001 01:58:00 +0100 NNTP-Posting-Host: 194.234.11.94 X-Complaints-To: abuse@danbbs.dk X-Trace: news.get2net.dk 984359744 194.234.11.94 (Mon, 12 Mar 2001 02:15:44 MET) NNTP-Posting-Date: Mon, 12 Mar 2001 02:15:44 MET Organization: DanBBS Xref: news.net.uni-c.dk comp.ai.neural-nets:67522 comp.lang.apl:29361 comp.lang.awk:17087 comp.lang.beta:12735 comp.lang.cobol:102474 comp.lang.dylan:24144 comp.lang.forth:78492 "David Ness" wrote ... > Steve Graham wrote: > > > > Who would be interested in using his/her brain > > (and his computer) to solve the following? > ...this isn't much of a puzzle,... > > ...it does strike me as being _very_ important > to draw the line at only those [puzzles] of > particular interest or of the highest quality. > Mediochre puzzles are of _no_ interest to me, > and do not accomplish what the last round > associated with the `New Scientist' did, namely > expose lots of different and contrasted > technique. You appear to be making an interesting claim, i.e., that the range of techniques for solving this puzzle computationally is limited (and also too simple to be interesting?), regardless of the language. Certainly, simple enumeration of all the combinations followed by selection of the (presumably) only one that satisfies all the constraints seems easy enough to program, though I'd still be curious to see how the constraints are specified programmatically in different languages. But I would hope that other, more clever solutions are also possible, and perhaps there might be variation among languages as to which are the simplest or most efficient. Personally, I thought the New Scientist puzzle was ridiculously simple, at least in APL and after one realized that it was nothing more than simple pattern matching. As I see it, the constraint matching in this "Einstein" puzzle is much more complex than the New Scientist problem, *especially* if one wants to aim for efficiency instead of brute force. By the way, for you the problem may seem simple because it's one you're familiar with and because you have honed your skills in solving that particular type of puzzle, but 1) *most* people *do* have difficulty with that kind of problem (although the quoted 98% figure may be a bit high) and 2) problems that are "simple" for the human mind to solve are not always simple to program. Consider, e.g., parsing English sentences, or identifying the "brown" regions in a painting. I, for one, would like to see people submit solutions to this one, and see how -- or if -- they vary. And if it's really as simple as you imply, perhaps you could prove it by whipping off a program and being the first to submit. Or if you're claiming -- as I think you are -- that this particular puzzle is an unnecessarily simple example of it's genre, pose us a more difficult version, then give us a program that will solve both... or preferably *any* problem of the same type. /Jim Lucas