Re: [ANN] Lua 5.2.1 (work1) now available

[Enrico Colombini <erix@...>]

On 25/03/2012 19.01, Roberto Ierusalimschy wrote:
> One characterisitc of these tests is that all of them seem to be
> dominated by GC times. On one hand, that means they are not actually
> focusing on the changes in 5.2.1. On the other hand, it may mean that
> those changes have little impact on total program time. (As I said, most
> programs that use long strings as indices produce these long strings
> dynamically, and so the time to create and collect them makes indexing
> time not relevant.)

I tried to write some significant tests about string concatenation with 
short and long strings, but I realized they were not significant at all, 
for the same reason.

--
  Enrico