Re: [Axiom-developer] Unions in Spad

[Gabriel Dos Reis]

"Bill Page" <address@hidden> writes:

| But really I think the concept of "selectors" in both Union and Record
| is at best a legacy of earlier days in programming language design. 

Curiously, I believe they almost right.  fields in
unions and records need to be properly scoped.

| It
| makes much for sense to me to define Union and Record as co-limit and
| limit in the sense of category theory. Then Union selectors are just
| injections operations and Record selectors are projection operations
| which are exported like any other function from these types. There is
| no need for any lower level language construct.

Well, if we are to be following that reasoning, there is no need for Spad
-- we just need lambda calculus. 
We should be discussing what is useful, convenient; not the bare
minimum that is necessary.

-- Gaby