R. David Murray added the comment:
My understanding is that iso-8859-1 is the standard, and really means "binary".
Turning it into unicode happens later. See the WSGI PEPs for more information.
--
nosy: +r.david.murray
resolution: -> invalid
stage: -> committed/rejected
status: open
New submission from Matt Tierney:
Changed the headers string charset to utf-8 since a header field may contain
non-iso-8859-1 characters. For instance, a Location header field may contain
Cyrillic characters.
--
components: Library (Lib)
files: tierney.patch
keywords: patch
messages: 1
