Re: [PATCH] Improve geometric types

At Fri, 15 Sep 2017 17:23:28 +0900 (Tokyo Standard Time), Kyotaro HORIGUCHI <horiguchi(dot)kyotaro(at)lab(dot)ntt(dot)co(dot)jp> wrote in <20170915(dot)172328(dot)97446299(dot)horiguchi(dot)kyotaro(at)lab(dot)ntt(dot)co(dot)jp>
> At Thu, 14 Sep 2017 16:19:13 -0400, Robert Haas <robertmhaas(at)gmail(dot)com> wrote in <CA+TgmobinBA7uvQifYaYGdDUoF6VTo56dvoTT6nKSpJF-Zfv5A(at)mail(dot)gmail(dot)com>
> > On Thu, Sep 14, 2017 at 3:33 AM, Kyotaro HORIGUCHI
> > <horiguchi(dot)kyotaro(at)lab(dot)ntt(dot)co(dot)jp> wrote:
> > > I recall a bit about the double-evaluation hazards. I think the
> > > functions needs a comment describing the reasons so that anyone
> > > kind won't try to merge them into a macro again.
> >
> > I think we can count on PostgreSQL developers to understand the
> > advantages of an inline function over a macro. Even if they don't,
> > the solution can't be to put a comment in every place where an inline
> > function is used explaining it. That would be very repetitive.
>
> Of course putting such a comment to all inline functions is
> silly. The point here is that many pairs of two functions with
> exactly the same shape but handle different types are defined
> side by side. Such situation seems tempting to merge them into
> single macros, as the previous author did there.
>
> So a simple one like the following would be enough.
>
> /* don't merge the following same functions with different types
> into single macros so that double evaluation won't happen */
>
> Is it still too verbose?

That being said, I'm not stick on that if Robert or others think
it as needless.

regards,

--
Kyotaro Horiguchi
NTT Open Source Software Center