To unicode or not to unicode
rridge at csclub.uwaterloo.ca
Sat Feb 21 23:07:35 CET 2009
Ross Ridge (Sat, 21 Feb 2009 14:52:09 -0500)
> Except in practice unlike Python, many newsreaders don't assume ASCII.
Thorsten Kampe <thorsten at thorstenkampe.de> wrote:
>They assume ASCII - unless you declare your charset (the exception being
>Outlook Express and a few Windows newsreaders). Everything else is
No, it's an assumption like the way Python by default assumes ASCII.
>> The original article displayed fine for me. Google Groups displays it
>> correctly too:
>Your understanding of the principles of Unicode is as least as non-
>existant as the OP's.
The link demonstrates that Google Groups doesn't assume ASCII like
Python does. Since popular newsreaders like Google Groups and Outlook
Express can display the message correctly without the MIME headers,
but your obscure one can't, there's a much stronger case to made that
it's your newsreader that's broken.
>> I could just as easily argue that assuming ISO 8859-1 is the defacto
>> standard, and that its your newsreader that's broken.
>There is no "standard" in regard to guessing (this is what you call
>"assuming"). The need for explicit declaration of an encoding is exactly
>the same in Python as in any Usenet article.
No, many newsreaders don't assume ASCII by default like Python.
>> The reality however is that RFC 1036 is the only standard for Usenet
>> messages, defacto or otherwise, and so there's nothing wrong with
>> anyone's newsreader.
>The reality is that all non-broken newsreaders use MIME headers to
>declare and interpret the charset being used.
Since RFC 1036 doesn't require MIME headers a reader that doesn't generate
them is by definition not broken.
l/ // Ross Ridge -- The Great HTMU
[oo][oo] rridge at csclub.uwaterloo.ca
More information about the Python-list