The DO clause of LOOP has an implicit progn[1], no need to do it explicitly.
Don't be clever with FLET/LABELS, 99% of the time you're doing it out of some vague "information hiding" principle. 99% of the time you end up making it a toplevel function, since local functions just make debugging harder.
A lot of (if (= x 1) ..) and (if (= x 0)) .. type forms are usually a sign you need a readable predicate. (when (foo-p <expr>) .. body)
Array this, array that. Wrap the BOARD type already. You can use classes and structs. This is not Scheme.
Excessive use of REDUCE and LAMBDA is a sign of Schemish hacking. Again, Common Lisp is neither C nor Scheme. The code uses MAP & REDUCE half the time, then AREF, LOOP and SETF the other half. FP-purism is the root cause of code schizophrenia like this. Try to write pronounceable code -- well, relatively speaking.
--
Ohboy, looking at hyperspec I don't see this spelled out, but it works that way, at least when do is the last/only clause.
Don't be clever with FLET/LABELS, 99% of the time you're doing it out of some vague "information hiding" principle. 99% of the time you end up making it a toplevel function, since local functions just make debugging harder.
A lot of (if (= x 1) ..) and (if (= x 0)) .. type forms are usually a sign you need a readable predicate. (when (foo-p <expr>) .. body)
Array this, array that. Wrap the BOARD type already. You can use classes and structs. This is not Scheme.
Excessive use of REDUCE and LAMBDA is a sign of Schemish hacking. Again, Common Lisp is neither C nor Scheme. The code uses MAP & REDUCE half the time, then AREF, LOOP and SETF the other half. FP-purism is the root cause of code schizophrenia like this. Try to write pronounceable code -- well, relatively speaking.
--
Ohboy, looking at hyperspec I don't see this spelled out, but it works that way, at least when do is the last/only clause.