On Aug 23, 2004, at 7:47 PM, Tom Lane wrote:
>> Oh, are you sure that you got my UTF-8 data? Because it came back in
>> your reply all mangled.
>
> I deliberately left it half-MIME-ified so that the hex byte values
> would
> be visible. One of the variables we have not controlled for is whether
> what you posted is the same as what other people received --- Tim
> Allen's point about the claimed encoding of the message is very
> relevant
> here. Lots of mail software will try to translate encodings.
Yes, my mail client changed the UTF-8 to EUC-KR (I'm going to find out
how to make it send everything in UTF-8). But my examples were working
as I described in the terminal with psql (and I do have my term set to
use utf-8), and the problem was solved by initdb --locale=C.
And now that I know to just use C, I won't worry about it again.
Cheers,
David