Re: changes in 'require'

[Edgar Toernig <froese@...>]

Roberto Ierusalimschy wrote:
>
> 1) Following Diego's suggestion, require "a.b" would first require "a".

IMHO that would be wrong.  That assumes the "a" is the core of "a.b".
Shouldn't it be that "a" is the complete package and "a.b" a subset?
How is "a" supposed to require anything else in "a", i.e. "a.c"?
You instantly get cyclic references - IMHO a bad idea.

If you really want some automatism, you could load "a.common" or "a.core"
or something like that on a require"a.b".  But is it really necessary?
"a.b" could do that by itself ...

Just my 2 cent, ET.