GeezerVison

These are not ASCII bytes, though?

Not sure I understand, did I make an error in the result?

ASCII encoding uses only the bytes up to 127, and it can’t encode characters such as é at all. In some places of the API – like Buffer.from(…, 'ascii') – Node just treats ascii like latin-1 for performance reasons, where é maps to 0xe9. That doesn’t always work, though; for example:

> Buffer.from('tést', 'ascii').toString('ascii')
'tist'

If you want to use this example, I’d suggest going with something like Latin-1 bytes, and use Buffer.from(…, 'latin-1').

Ah okay, wasn't aware of that. I'll change it to latin-1.

Why remove this example? I think I can see where this is coming from, but it might be better to just replace ASCII with ASCII/UTF-8 in the comment?

My reasoning was that it's more consistent with the other examples in the doc, and also shows the difference between the two encodings that will affect the output ('test' is the same in ascii and utf8 but not 'tést'). That may or may not be important

Just changing the text to ASCII/UTF-8 is still an improvement but I think this example is clearer.