Thoughts on R7RS

benjisimon.blogspot.com

64 points·gnosis··9 comments

Odd how he dislikes both not having a way to create disjoint types and the way provided to create disjoint types (define-record-type).

Indeed, and what could be a truer reflection of the Scheme community than this ;-)

LISP and Buddha-nature... ;)

The problem, I think, is that he wants a low level tool for generating disjoint types (not satisfied by define-record-type, which is not required to be disjoint from vectors or conses or what have you). This would give you the ability to have enums, ML style discriminated unions, simple records, etc. without necessarily requiring that each and every one of them be in the small language.

That said from a practical point of view some sort of record type is long overdue.

As of the current draft, define-record-type is disjoint. They're "distinct from all existing types" and explicitly mentioned as being disjoint in the notes.

Ah! Well, that certainly improves things; either I misread the draft or that language wasn't there when I looked.

Still going to stick with R5RS. do we really need to revise Latin?

My understanding from superficial observation is that they are trying to fix the (perceived) mess that was R6RS, and thus unify the community.

That would be the case. Some people were unhappy with R6RS not keeping Scheme a minimal language anymore.

Now they're making one "small" Scheme and one "big" Scheme. That's a good thing because they can actively decide what to put in the big version (full numeric tower, e.g.) and what to put in the small language (e.g. a standardized module system).

Reply on news.ycombinator.com