Re: [boost] Re: Comments on the thread library

From: Greg Colvin (gcolvin_at_[hidden])
Date: 2001-04-03 10:15:58

From: Peter Dimov <pdimov_at_[hidden]>
> From: "David Abrahams" <abrahams_at_[hidden]>
> 
> > From: "Matthew Austern" <austern_at_[hidden]>
> >
> > > > As with string, the needs and wants of mutex users vary a great deal. If
> > > > we don't have a generative system, we can't meet these varied needs.
> > >
> > > I don't agree with that last sentence.  If you need more than one class,
> > > it does not follow that the only solution is to have a generative system.
> > > Maybe the right solution is just to have more than one class.
> >
> > We could avoid the problem using a layered approach: if you have all the
> > classes, sticking a generative interface on top is simple.
> 
> Generative programming does not mean "stick a generative interface on top."
> If you have all the classes, a generative interface is simply not necessary.

It doesn't mean that, but it can do that.  And when is a generative
interface every strictly necessary?  You can always give a name to
every possible combination of features.

> Generative programming, if I understand it correctly, means "look at several
> concrete examples of the solution space, separate the orthogonal features,
> generalize."

Which is one good way of designing any set of classes.