Josh McKinney

> I see the appeal but I'm not sure I want to go down that route. I'd guess an external package containing a `Table` wouldn't see much use unless we...

> Changes on the other hand can be compared Yes. This is the entire point of bencher.

> Looks cool! I'm curious about what do we need to move forward with it? Should we reach out to them or can we just integrate it somehow? Read the...

I'd suggest that for this PR it's probably worth: - fixing up any obvious simplifications / problems I've missed - disabling the broken tests (unless there's a really obvious test...

> > Side note: there's also a [truncate](https://crates.io/crates/truncate) crate, which does various unicode split_at things, but it uses `unicode-segmentation` instead of `unicode-width` for calculating positions - there's some subtle incompatibilities...

> There's an **EXTREMELY** lengthy debate on to use / not use segmentation in #75 (ratatui's longest open issue). I don't know which side of that debate has more merit,...

I've updated the tests so that the document the current behavior even if it's wrong (as noted in the first post). This is better than panicking, so I'd suggest merge,...

> I don't quite understand why this works (haven't had a lot of time to go through it in detail). Maybe the PR description can have a high level description...

All tests passing (and corrected so that the edge cases seem like they're the right result) E.g. right aligned text that does not have enough space to render the first...

There's another possible simplification (for future consideration on this) that teases apart the span selection and span rendering. Something like (pseudo code): ``` fn render_ref(...) { for span in self.visible_spans(...)...