* List explicitly effects of the UTF-8 mode, but also "side effects"
* Add a new "Relationship with the locale coercion (PEP 538)" section
* Add a new "Version History" section
* PEP 540: Update based on PEP 538 experience
- CentOS 7 is a better example of problems than Alpine Linux
(default locale is C/POSIX, C.UTF-8 isn't available for coercion)
- reword the annex discussing the design trade-offs between
PEP 538's locale coercion and PEP 540's introduction of a
UTF-8 specific mode
* Tweak wording of PEP 538 comparison
* Strict mode doesn't use strict for OS data anymore: keep
surrogateesscape, explain why in a new alternative
* Define the priority between env vars and cmdline options to choose
encodings and error handlers
* add section: "It's not a bug, you must fix your locale" is not an
acceptable answer
* elaborate the "Expected mojibake and surrogate character issues"
section
* Add the "Producer-consumer model using pipes" use case
* Enable UTF-8 mode by default if the locale is POSIX
* Add Use Cases
* Add "Don't modify the encoding of the POSIX locale" alternative
* Rephase Abstract and Proposal
* Proposal: mention expected mojibake issues
* Fix PEP number: 393 => 383
* Add links
Victor Stinner
larryhastings
Nick Coghlan
Mariatta
Huang Huang
Jim Fasarakis-Hilliard
Victor Stinner
Thomas Samson