When I started programming I quickly found that 0-based indexing was generally simpler. A lot of the time, it makes no difference one way or the other (because

When I started programming I quickly found that 0-based indexing was generally simpler. A lot of the time, it makes no difference one way or the other (because you're only using indexes to get at each item in turn, so you just memorize the boilerplate for whichever language you're using), but the moment you introduce ranges, or try to convert dimensionality - particularly with random access - or do some wraparound, or try to implement something like a ring buffer, it's all too easy to come a cropper. Either you put "1+" on every array access, which is error-prone, or you try to manage two different "spaces" (the 0-based one that's the ideal representation and the 1-based one used for indexing) and carefully convert between them... which is error-prone too. Alternatively, you could just decide to start indexing at zero. ...or you could stick with 1-based indexing, it's up to you. The right thing to do doesn't become wrong just because you don't do it ;)