Re: table.insert and negative numbers

[Doug Rogers <rogers@...>]

Matthew Paul Del Buono wrote:
> ... the following line would cause the system to go into ...
> an infinite loop:
> t = {};
> table.insert(t, 2^31, 1);
> After researching, however, it appears that this is due to
> the fact that 2^31 becomes a negative number.

IMO there are two problems:

  (1) 2^31 should not become a negative number.

  (2) Inserting at any location on an empty table should not require
that much time.

The solution you posted addresses (2). I'm still worried about (1),
though. The 5.1 LRM says that 'Exponentiation works for any exponent.'
(http://www.lua.org/manual/5.1/manual.html#2.5.1), but doesn't mention
this boundary condition. I see how it happens but I think it should be a
positive number.

Doug

-- 
Innovative Concepts, Inc. www.innocon.com 703-893-2007 x220