http://qs321.pair.com?node_id=800407


in reply to Perl for hobbyists

Mazes. Mazes for my wife and I. Instead of trying to port some of my old MS-DOS C code to linux, I found Games::Maze.

I was excited, until I found how simple and predictable the mazes were. It seems that once it is boxed in, it backtracks the way it came. Making a simple change fixes that:

# Right here is where things get predictable! Mix it up. # ($c, $r, $l) = splice @queue, 0, 3; ($c, $r, $l) = splice @queue, (rand(@queue / 3) *3), 3;

Yes, my code uses more memory, but the mazes are more challenging. I guess 'more memory' is all relative, I remember how many hoops I had to jump through to use more then 64k of data in MS-DOS.

I have code that takes the Games::Maze square data and creates a png file.

Now if I could only find a good way to represent the maze, and have the output span a few dozen pages.

Replies are listed 'Best First'.
Re^2: Perl for hobbyists
by ambrus (Abbot) on Oct 10, 2009 at 11:27 UTC
Re^2: Perl for hobbyists
by moritz (Cardinal) on Oct 11, 2009 at 20:23 UTC

      Wow, nice article. I didn't care about the speed problem because when I wrote my program, computers were fast enough. I only wrote the fast algorithm eight years later in the perl rewrite. Otoh, I think my program generates better (more interesting) mazes.