Talk:Recurse

Arbitrary placement of blocks

I had actually considered this, but it makes parsing very difficult (and would require a seperate comment character, like Oerjan said). Actually, the second use of a function's name (in the lower-left corner) is also just to make parsing easier (since the lower border would otherwise be an acceptable line of code that would have to be chopped off after parsing). MagiMaster 18:20, 8 Oct 2006 (UTC)

I don't think it would be that difficult to parse, given a couple of rules: the (top) border of a block cannot contain spaces, and blocks must be separated by spaces. Then you can still parse top down by keeping a list of currently open blocks. --Ørjan 12:29, 9 Oct 2006 (UTC)

Hmm... I suppose not, but I'm pretty lazy when it comes to writing parsers. :) If someone else wants to write it, then arbitrary placement could be added to the specs. MagiMaster 13:43, 9 Oct 2006 (UTC)

Bigger commands?

You know. Define three vertical entry points and suddenly you have a $$$ command. --Ihope127 20:57, 9 Oct 2006 (UTC)

I actually thought about that too, but would it be worth it? It wouldn't really be too hard to parse, but it would break the otherwise strict cell-based approach (which might not be a bad thing). Right now, though, it doesn't seem like the four current entry points are used that much, since executing a function can change the direction of the cursor. What do yall think? Should a function be able to change the direction of the cursor in the calling function? It allows the definition of things like reflectors or an unconditional turn, but it makes it hard to make functions that work (the same, at least) in multiple directions. Extended to larger commands, where would the cursor exit the function? If functions couldn't affect the direction, the simple answer would be straight across. MagiMaster 02:16, 10 Oct 2006 (UTC)