Re: %define boolean variables

[Joel E. Denny]

On Sun, 28 Jan 2007, Paul Eggert wrote:

> "Joel E. Denny" <address@hidden> writes:
> 
> > Users' grammars will be more consistent and, I believe, more intuitive if 
> > any %define variable foo is set to true always with:
> >
> >   %define foo "true"
> >
> > and is set to false always with:
> >
> >   %define foo "false"
> 
> It sounds like a fairly minor issue; I don't see much reason to worry
> about it one way or the other.  If you think it's a real win, though,
> go for it.

Thanks for sharing your opinion.

I agree it's a minor issue.  However, if it's wrong even slightly, if it 
serves no purpose, and if it's trivial to fix, why not fix it?  I feel 
that minor issues add up too quickly to an overall poor user (and 
developer) experience, so it's best to weed them out early... before 
backward compatibility demands they be kept.

So again, my plan is to remove the value-less form of %define.  I'll also 
rewrite b4_percent_define_flag_if to invoke b4_complain when the user 
specifies anything other than "true" or "false".

But I'll wait a little while in case anyone has major objections to these 
changes.  Maybe someone will point out why supporting aliases for true and 
false in Bison actually has a practical use.