2009/8/7 gert <[email protected]>:
>
>
>
> On Aug 6, 8:44 pm, Damjan <[email protected]> wrote:
>> > Not very well.
>>
>> > We are having another argument about WSGI specification and Python 3.0
>> > at the moment on Python WEB-SIG list. The discussion seems to have
>> > exploded over night and have about 30 messages to read about it yet.
>>
>> > If some sort of resolution isn't reached this time I am going to give
>> > up and simply not support Python 3.0.
>>
>> I'd say that requiring UTF-8 (on the client side) and making most of
>> WSGI unicode is the better option. Maybe, as P.J.E says, that way WSGI
>> will loose some of it's idempotency with HTTP, but who cares... WSGI
>> is about Python applications!
>>
>> And if a request comes that can't be decoded as UTF-8 .. then a "400
>> Bad Request" error should be thrown.
>>
>> The other option, will mean that every web framework and every WSGI
>> middleware, and every programmer will have to reimplement the
>> character  conversion heuristics (and probably do it wrong).
>>
>> I don't see any problems with declaring an UTF-8 constraing for data
>> coming from the HTTP side. Less (unneeded) choice for programmers,
>> middleware and frameworks is good!
>
> What about returning binary data to a browser (png image) ? I thought
> this can only be done in latin-1 ?

Totally unrelated. The discussion is only about values in WSGI
environment and not how data is returned, which would still have to be
bytes for response content.

Graham

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"modwsgi" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/modwsgi?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to