Carl Mäsak

## First-class functions Taking a step in an orthogonal direction: given how absolutely awesome functions are, is there any way they could be made more awesome? [According to Christopher Strachey,...

## Continuations I have nothing in my notes about continuations. Fortunately, I've spent the last week implementing them, _again_, in Bel. I needed them for my "fastfuncs" mechanism, which is...

This is as far as I got tonight. Looking back at my progress, I think I would need two more evenings to finish up. I'll try to rise to that...

> That's a weird list of languages with tuple support. I'm not sure what those share in common wrt multiple return values? Aye, I fear I did not do that...

> If you don't explicitly ask for all the return values, you only get one. This isn't even my main beef with the feature (although that sounds like a pretty...

I haven't forgotten about this thread, but getting back into it is going to require bigger chunks of time — realistically, the next time I'll have those will be around...

[On Abstraction](https://oh4.co/site/on-abstraction.html) by Eric Scrivner: > Substituting a call to a subroutine for the contents of a subroutine builds a lexicon which can be used to reduce duplication and heterogeneity....

> ## First-class functions I just realized one thing. Here's a sentence from [cppreference](https://en.cppreference.com/w/cpp/language/operators): > When a user-defined class overloads the function call operator, `operator()`, it becomes a [FunctionObject](https://en.cppreference.com/w/cpp/named_req/FunctionObject) type....

> My mental model is this: functions abandon their role as the fundamental unit of computation, and hand it to a smaller entity: what typically is called a "basic block",...

> The removal of the basic block's power to call other functions is significant. Here is the first time we're saying "you can do many things, but you cannot do...