Submatches vs. Recursion in ThiNG Patterns

Submatching in a pattern, +var@pattern, looks suspiciously similar to recursion in a value, var=value. In each case, we expect a value which is both named and processed, and in each case, a binding is introduced into code. The processing in the submatch case is a further pattern-match operation, and in the recursion case it is a nested evaluation.

The main difference, besides the overall context, between the two forms is the scope of the introduced binding: in...

Partial-evaluation of Monads in Scheme - A New Hope?

The other day, I noted that my generic-monad-in-Scheme implementation suffered from the flaw of not being able to type-check monad assemblies before any actions had been executed, and that in general such ahead-of-time checking is not possible without abstract interpretation of some kind, because the functional code in the right-hand-side of the bind operator can decide

Monads in Scheme - A Retraction

I don't know what crack I was smoking the other day, but my implementation of general monads not only does not have the property I claimed for it, viz. that it provides the same level of type-checking as Haskell ahead of execution-time, but it cannot provide s...

Monads in Scheme - Update

Just quickly revisiting the article the other day on monads in Scheme: I've just updated my experimental implementation to propagate monad type information both forward from the arguments to the result of the bind operation and backward from the expected result type to the required argument typ...

Trait Composition in ThiNG

Trait composition is one of the fundamental ideas in ThiNG r3/r4. It's used to assemble pattern-matching clauses in functions, to assemble definitions into a module, to (functionally-) update data-structure, to augment generic functions with new definitions, and to assemble modules into programs.

define cons [+car: [+cdr: [First: car Rest: cdr]]]
define map [+f: loop=[(cons +a +d): (cons (f a) (loop d)) Nil:Nil]]

Both these definitions use objects with more ...

Recent spikes in ThiNG

Since I last posted about ThiNG, my ideas have shifted closer to those expressed in Jay & Kesner's Pure Pattern Calculus, and I've built a few experimental spikes to gain some concrete experience with some of the language features and designs I've been thinking about - among other things, that traits and modules share a number of core characteristics, and that using monads to structure effects can be profitabl...

The Pure Pattern Calculus, and Extensible Records With Scoped Labels

Two papers very similar to the ideas I've been developing for ThiNG have popped up in the last couple of weeks:

Daan Leijen's Extensible records with scoped labels: describes an intriguing system for typing records similar to those I'm trying to define with ThiNG.
Barry Jay ...

Design Problems in ThiNG

At the moment, I've three main problems in the work I'm doing designing ThiNG: syntax, semantics, and macros. (Doesn't leave much, does it?) I'll cover the syntactic and semantic problems tonight — my thinking on macros is much too fuzzy to write down at the moment.

Syntax

I'm leaning toward a concrete syntax that, like Smalltalk and Slate, makes message-sending the only concise operation, leaving application in particular as almost second-class. Since I expect mess...

ThiNG repository available

I've been hesitant to make my code publicly available, since it's embarrassingly messy, but a couple of people have indicated they're interested in taking a closer look, so here it is:

$ darcs get http://eighty-twenty.org/~tonyg/Darcs/smalltalk-tng

The code is mostly written using Chicken (with my ^[1] of ThiNG (older articles: 1, 2), a simple normal-order pattern-matching-based evaluator with a parser built out of

eighty-twenty news