Works correctly according to the EN docs. https://3v4l.org/rgStf

$rest is false because "length is given and is negative" and "start denotes the position of this truncation or beyond".
$rest1 is an empty string (not null) because "string is equal to start characters long" (6) and "length is given and is positive".
http://php.net/manual/en/function.substr.php

But I do see some errors in the pt_BR docs, so I'm repurposing this bug for them.

Those docs still say $rest1 should be false:
- EN:    If start denotes the position of this truncation or beyond, FALSE will be returned.
- pt_BR: Se start denota uma posição além da truncagem, uma string vazia será retornada.

And the example needs adjustment:
- EN:    $rest = substr("abcdef", 4, -4);  // returns false
- pt_BR: $rest = substr("abcdef", 4, -4);  // retorna ""

Also:
-

...
Also missing a few comments: when length is given and 0/false/null, when length is omitted, returns false on failure or empty string, and the Changelog.

> Those docs still say $rest1 should be false
Backwards: they say empty string, should be false.

I posted an updated on the Php Docbook Online Editor.

But the process of getting an update out is really confusing

@thiago: I see your changes and it looks like you followed the right process: make changes, create a patch, and submit the patch for review. Someone will look over it.

This bug has been fixed in the documentation's XML sources. Since the
online and downloadable versions of the documentation need some time
to get updated, we would like to ask you to be a bit patient.

Thank you for the report, and for helping us make our documentation better.

Thanks by the patch.