Re: [jamboost] Disallow free property in build request

From: David Abrahams (david.abrahams_at_[hidden])
Date: 2002-07-16 06:52:24

----- Original Message -----
From: "Vladimir Prus" <ghost_at_[hidden]>
To: <jamboost_at_[hidden]>
Sent: Tuesday, July 16, 2002 7:46 AM
Subject: Re: [jamboost] Disallow free property in build request

> David Abrahams wrote:
> > From: "Vladimir Prus" <ghost_at_[hidden]>
> >
> > > > I know we discussed this and it made sense to me at the time to
provide
> > > > motivation, but now it just seems obvious. They're non-propagated
by
> > > > definition, so what else is there to say?
> > >
> > > I don't agree. You know, user doesn't care about non-propagated
features
> >
> > at
> >
> > > all. He types
> > >
> > > bjam release define=FOOBAR
> > >
> > > in the command line and asks what targets will "define=FOOBAR"
affect.
> > > If you say that define is non-propagated feature, he would ask why it
is
> >
> > so.
> >
> > So then you'd like me to write a rationale about why define is a
> > non-propagated feature?
>
> I'd rather you answer the original question. Or you think the user will
not
> wonder?

I don't think the user will wonder why we have a rule that a features with
the "non-propagated" attribute are not propagated.

Different individual features will have their own reasons for not
propagating. I could discuss "define" specifically, but that's not really a
motivation for why non-propagated features work the way they do...

not-trying-to-be-difficult-ly y'rs,
dave