> +--------------------------------------------------------------------------------------+
> | 私はガラス
> +--------------------------------------------------------------------------------------+
You say it displays correctly in xterm (ie. you didn't see these in your
xterm).There are HTML/XML unicode character entities, probably generated by your
mailer from your Unicode cut'n'paste.Using SQL ASCII to store UTF8 encoded data will work, but postgres won't
know that it's manipulating multibyte characters, so for instance the
length of a string will be its Byte length instead of correctly counting
the characters, collation rules will be funky, etc. And substring() may
well cut in the middle of an UTF8 multibyte char which will then screw
your application side processing...Apart from that, it'll work ;)