Ben Lerner

Not just overzealous attempts to make functions less observable: it had had measurable consequences for memory usage when every single function stored a copy of its docstring, especially nested closures...

Hmm. Printing annotations _might_ be possible, though still harder than I'd like in all cases (e.g. refinements or other fancy types). We still don't track docstrings through compilation yet. It's...

Can you provide a standalone test file, and see if it fails using `src/scripts/just-lex.js` or `src/scripts/just-parse.js`

The lack of a trailing newline is a red herring here; it has almost nothing to do with what's going on. The actual problem doesn't surprise me much. The two...

I think it parses, but I'm still tentatively concerned that if a _later_ expression (after that multibyte string) has an error, our srcloc info will be wrong by a character...

The jokes write themselves...: - Bravo, Brave. - I thought the blink tag was deprecated? - I thought Blink was part of Chrome? - :blinks confusedly: There is no blinking...

In addition to this, when the text box appears, it's substantially _taller_ than the status bar was, and the elastic space that pushed the other text in the status bar...

(Related to #75, I think?)

Don't we have tests for that? (And if we do, why aren't they failing...?)

I can solve one part of this: That "not a convenience constructor" is working as intended: `set` is referring to [this `set`](https://www.pyret.org/docs/latest/lists.html#%28idx._%28gentag._238%29%29) from the lists module, as exported by [`essentials2021`...