Carl Mäsak

When I re-read this issue, I feel what's missing from the OP is that control flow is an aspect of _code generation_. (That's also why I just closed #252.) Getting...

Note that even the conservatism suggested in https://github.com/masak/007/issues/145#issuecomment-232278144 is not going to suffice. Why? Because blocks with parameters code-generate as separate "units" of callable code in the general case, because...

Just driving by to point out that `Symbol` (#312) would be a really good fit for constructing labels at the bytecode level.

> * The above is mainly meant for the innards of a 007 compiler. But as an added bonus, macro authors could `use cg;` inside their module, and just make...

After several years of waiting for the other shoe to drop, I now think that the control API should be phrased in terms of continuations. Hoping to be able to...

Maybe I'm not really contradicting 2021-me when I say that it feels like we're essentially talking about algebraic effects here. Maybe the connection is this: if you have continuations and...

Reminder: even after the move, just like Travis did (#422), GitHub Actions should run the tests with all the feature flags switched _on_.

By the way, @raiph, this is very close to what I once referred to as [an Algoloid that uses macros to define itself](https://github.com/masak/007/issues/294#issuecomment-393669644). In a way, it's nothing exciting. Just...

Indeed. I ran the above, and: ``` false none none ``` It's a bug.

In a way, I was right six-seven years ago when I said to @moritz++ that ASTs (Qtrees) close over their environment. The only refinement I would make to that today...