[bitc-dev] DISCUSS: Name That Type Class!

[Jonathan S. Shapiro]

On Sun, 2008-07-20 at 16:08 +0100, David-Sarah Hopwood wrote:
> Jonathan S. Shapiro wrote:
> > We need decent names for these. My personal inclinations are:
> > 
> >   ref-types => RefType?   (definitely remove plural, shift to word caps)
> >   copy-compat => Compat?
> 
> Use 'Compatible'; the abbreviation sounds ugly.

Worth considering. My immediate thought is that "Compatible?" is long, a
pain to type, and will create indentation challenges...

Oh. And before I forget, we're either going to go to word caps
throughout or stick to lisp-style hyphenation throughout. However bad
the convention we choose may be, one convention is better than two.

It wasn't my intention to confuse this issue.

> > but I have absolutely no useful thoughts for what to call
> > top-copy-compat.
> 
> What is it intended to be useful for?

In practice, I don't expect most users to use it at all. It is needed
internally by the complete inference system. This means that we need to
be able to emit it in output, and therefore that we need to be able to
re-admit it in input. So it needs a name.

Pragmatically, it shows up in some unusual cases where an inner type
that is not mutation sensitive gets wrapped by a constructor.

Swaroop: can you send out a real-world example?


shap