Re: [ANN] Lua 5.3.0 (final) now available

[Jorge <xxopxe@...>]

On 14/01/15 16:13, tomas wrote:
> > And if a
> > application has some other meaning for nil besides "absence of value",
> > perhaps something from the application domain, then I think said
> > application is badly designed. In that sense, in application domain,
> > nil is not a "perfectly valid return"
> In SQL, NULL is that kind of value.  And the mapping of NULL to nil is 
> quite natural, although sometimes the programmer have to take care 
> (for instance, not to make a query that could result in NULL at the 
> first column of a valid row).

This is a case of bad design I was thinking of. A NULL from a database 
query is not a nil in the language, but a singleton (say local 
DBNULL={}). Tough my thinking on this is more on the arrays-with-holes 
line (but as Roberto said, they're related).

Jorge