On Fri, Jun 8, 2012 at 12:24 PM, Daurnimator
<quae@daurnimator.com> wrote:
What does foo.1.5 do?
1 . 5? or foo[1.5]?
foo[1][5]. As I said, we don't cover all possible cases on purpose, that's the only way not to end up with horrible bloatware.
The relevant question here is: how many real-life cases are there, were you need deep paths into tables indexed by floats? Are they frequent enough to justify bloating the definition of a path? In our case the answer is an obvious "no", and I'd bet 99% of users will reach the same conclusion.
Generalization has a cost; not taking it into account is expected from mathematicians, it's a fault for engineers...
