Re: no return value vs. nil return value(s)

[Enrico Colombini <erix@...>]

On 12/09/2011 8.17, Dirk Laurie wrote:
> In both cases, the behaviour seems not to be planned for, but
> not sufficiently obnoxious to justify testing for it.

Speaking of testing (nothing new here):

 type(f()) --> stdin:1: bad argument #1 to 'type' (value expected)
 type(g()) --> nil

--
  Enrico