Re: Handling unrecognized or unsupported charset

> Is this mandatory or defined somewhere as a "must"? I hope no...

It will define how the behaviour should appear externally, and generally
ignoring resulting timing differences.
 
> So we are doing "inline conversion": blind ASCII-7/UTF8 assumption until
> first @charset.

As only byte order marks can precede the @charset, that has makes no
real difference in the results of a streaming versus a batch 
implementation.

Received on Friday, 16 July 2004 02:05:04 UTC