Date: Wed, 5 Dec 2001 01:33:03 -0500 To: 6046-contest@theory.lcs.mit.edu From: Bob Hearn Subject: Programming contest update Hello there contest entrants! The contest site has been updated. The C sample code is finally there (sorry!), the javadoc documentation is there for the Java code, and there are a few useful links to other sites. Please let me know if you have any problems building, running, or understanding any of the sample code, and especially let me know if you find any bugs! Cheers, Bob -- Date: Thu, 6 Dec 2001 09:39:08 -0500 To: 6046-contest@theory.lcs.mit.edu From: Bob Hearn Subject: New test files Hello there puzzle hackers, I've added some more test files (test_files2.tar) to the contest site. Note that there is a subtlety for the pentomino shapes. Consider the following piece arrangement: *********************** *** . . . . . . . . *** *** 3 . . . 2 2 1 . *** *** 3 3 3 3 2 1 1 1 *** *** . . . . 2 2 1 . *** In a physical puzzle, you could move pieces 1 and 2 up 1 unit simultaneously. But we have defined a move as a motion of a single piece. So you don't need to worry about that kind of motion - the sample code doesn't allow it, and you don't need to either. If you come up with any particularly nice test files yourselves, I'd appreciate a copy. Good luck! Bob -- Date: Thu, 6 Dec 2001 15:31:55 -0500 To: 6046-contest@theory.lcs.mit.edu From: Bob Hearn Subject: Prize announcement First prize for the programming contest is $100, plus a one-of-a-kind, custom-made puzzle. William Waite, a talented puzzle designer and craftsman living in Denmark, will translate a puzzle generated by the winning program into elegant physical form. You get to take your winning entry home with you! Good luck, Bob -- Date: Thu, 6 Dec 2001 17:36:30 -0500 To: 6046-contest@theory.lcs.mit.edu From: Bob Hearn Subject: Suggested code modification Hi there puzzle programmers, In regards to the nature of first prize (a puzzle made from one of your solutions), I might suggest you make the following change to your programs. The sample code prints out each solution that is more difficult than the previously most difficult. I would suggest changing this > to a >=. Then whatever the most difficult puzzle found, there will often be several found of that difficulty, and we can choose among them based on a subjective aesthetic evaluation when deciding which one will actually get made. The extra time required to print the additional solutions will be of minimal impact, I think. Thanks, Bob -- Date: Fri, 7 Dec 2001 09:56:05 -0500 To: 6046-contest@theory.lcs.mit.edu From: Bob Hearn Subject: Fwd: programming contest 6.046 Ashwini is looking for programming partners. Any takers? - Bob > To: rah@ai.mit.edu > Subject: programming contest 6.046 > Date: Fri, 07 Dec 2001 09:51:26 -0500 > From: Ashwini K Agrawal > > Hi, > I entered the programming contest for the class, and I wanted to find > people to work with. None of my friends are gonna enter the contest, > so would it be possible to have an email sent out to the contest list > where I could say that I am a student interested in working with a group > on the project? maybe then someone else can email me? > Otherwise I can work on it by myself. Thanks a lot :) > ashwini -- Date: Fri, 7 Dec 2001 17:43:50 -0500 To: 6046-contest@theory.lcs.mit.edu From: Bob Hearn Subject: Contest site update I've updated the programming contest site a bit. The only really new thing is an archive of mail sent to the contest mailing list. Those who joined recently should take a look. Bob -- Date: Fri, 7 Dec 2001 23:21:45 -0500 To: Eric TszLeung Ho From: Bob Hearn Subject: Re: Contest site update Cc: 6046-contest@theory.lcs.mit.edu Yes. I meant to document that; thanks for pointing it out. Likewise, you can assume that the largest box is at most 12x12 (MAXBOXDIM in inputspec.h). - Bob > In the demo code, the largest pieces being supported is at most 6x6. > (According to piece.h) > Can we assume the largest piece for all contest input is at most 6x6? > > > Eric --