Re: [Axiom-developer] Unions in Spad

[Gabriel Dos Reis]

On Wed, 11 Jul 2007, Stephen Wilson wrote:

| Gabriel Dos Reis <address@hidden> writes:
| 
| > Stephen Wilson <address@hidden> writes:
| > | I think the consensus is that `hybrid' Union types are not desirable.
| > 
| > I'm not sure I understand the arguments why they are not desirable.
| 
| Main question is _why_ they might be desirable.  Spad does not
| support them.  Aldor might in theory, but only in theory.
| 
| As such a construct appears to give no real benefit for anyone,  Im
| thinking that properly encoding the constraint is the proper move.
| 
| Do you have an argument for `hybrid' Unions?

In general, I'm of the opinion that the language should be
constrained only when we cannot come with proper semantics.
I'm not sure this case is one such case; that is why I would
like to understand the arguments.  I don't want a language
designed to express only what I can think of today.

-- Gaby