13 Aug
2002
13 Aug
'02
2:38 p.m.
Martin v. Loewis wrote:
[...]
For the charmap codec it's mostly about performance. I don't have objections for other codecs which rely on external resources.
Please remember that we are still about error handling here, and that the normal case will be "strict", which usually results in aborting the computation.
So I don't see the performance issue even for the charmap codec.
I guess this code might be used inside a webserver that outputs XML results and that honors the Accept-Charset header from the client, so it must do encoding on the fly. So I want the code to be as fast as possible. Bye, Walter Dörwald