Re: Explain?

[Ben Sunshine-Hill <sneftel@...>]

Hehe, good point with the order of execution WRT locals.

I think the real moral of this discussion is: "I'm glad I'm not the
one who has to write the compiler." Condolences to Luiz, Roberto, and
co. ;-)

Ben

On Thu, 28 Oct 2004 17:24:32 -0500, Rici Lake <lua@ricilake.net> wrote:
> 
> On 28-Oct-04, at 3:59 PM, Ben Sunshine-Hill wrote:
> 
> > which could conceivably cause inefficiency in the case of
> > upvalues. My main point was simply that the use of a variable in an
> > upvalue does not represent a "hidden" extension of the scope, only an
> > additional consideration when closing the scope.
> 
> Sure, you could do it, for example, by reordering stack assignments
> within a scope to reflect the order of deletion.
> 
> My main point was that dead local detection is not as simple as
> it looks at first sight (not that it is impossible), and I think
> our subsequent discussion demonstrates that.
> 
> By the way, your example does not actually present any difficulties,
> since the right-hand side of a local statement is evaluated before the
> local scope starts:
> 
>  do
>   2 --> local a [= 2]
>   a + 3 / toss out a / --> local b [= a + 3]
>   dosomething(b)
>   b + 1 / toss out b / --> local c [= b + 1]
>   doSomething(c)
>  end
> 
> So a, b and c could all share a stack slot.
> 
> However, adding doSomething(a) after doSomething(b) would create the
> issue. In that case c and a could share a slot, but b would need to be
> in a lower slot. All doable, but...
> 
> :)
> 
> R,
> 
>