Re: Why _ENV, not _G?

[GrayFace <sergroj@...>]

> On Sun, Jul 11, 2010 at 1:41 PM, GrayFace<sergroj@mail.ru>  wrote:
>    
> > IMO, after _ENV is introduced, the old _G is redundant. _ENV supersedes it.
> > So, won't it be better to keep the name _G instead of _ENV? It's shorter.
> >      
> But, it's different. _G always refers to global scope, _ENV refers to
> current 'environment'. So it would be confusing to change the meaning
> of _G.
>
>    
_G is just a field of _ENV that can be set to anything. For example, _G 
is nil after calling 'module' function (and module function is exactly a 
case when distinct access to global environment may be needed). If 
someone writes his own modules system, it's up to him to define _G. So, 
IMO, _G can't be relied on in 5.1.

--
Best regards,
Sergey Rozhenko                 mailto:sergroj@mail.ru