Re: Do not create conditional installation directories

[Akim Demaille]

Le 18 mars 09 à 22:28, Ralf Wildenhues a écrit :

> * Akim Demaille wrote on Wed, Mar 18, 2009 at 06:31:50PM CET:
> > > > > "RW" == Ralf Wildenhues <address@hidden> writes:
> >
> > > But the question remains: how do you think we should address the bug
> > > that bothers Akim, other than by this patch and a big warning in  
> > > NEWS?
> >
> > My focus was more about an empty $(foodir), not an empty $(foo_DATA).
>
> Can you describe situations where an empty $(foodir) happens?
> Is this a bug?  Is this a regression?

My message was about make support for an empty $(foodir) (made empty  
on purpose) to mean that make install should skip all foo primaries.   
I had included the code to show that currently in such a situation,  
Automake is schizophrenic: it refuses to invoke mkdir when the  
directory name is empty, but it agrees to in "it".  When DESTDIR is  
set it turns out the mean something (yet we did not mkdir it :).

This behavior is useless, and I was claiming that making support of "$ 
(foodir) is empty means don't install anything from it" would make  
some things simpler.  But I have not a strong need for this.