Re: another effc++ spurious warning?

["Gabriel Dos Reis" <gdr at acm dot org>]

You Wrote Giovanni Bajo
> Gabriel Dos Reis wrote:
>
>> Those rules were designed for fairly old-fashioned code that
>> mostly look like C code, i.e. verbose and a pain to evolve.
>> As you pointed out, they don't make much sense for modern
>> C++.  I would suggest to just remove that check.
>
> Well, you would suggest to remove -Weffc++ altogether (and I even might
> agree,
> but still).

I believe I have already made that suggestion, several times :-)

> I think it still have its uses for certain kind of code, and
> for
> beginners. In that case, we still want this check. I don't think it makes
> sense
> to disable a single warning of -Weffc++ unless it's proven totally
> useless.

The rules -Weffc++ is trying to implement makes sense only
  (1) in a given specific context
  (2) most important, you do have the actual machinery to do
      *true* semantics analysis.

we don't have (2), we have no clue about (1).

And I'm not surprised by the result.