Browsers don't seem to care about the extra hyphen or the case.

I did a test with a small cgi script, sending out various
content-type charsets and appropriately encoded body texts.

Firefox 1.04 and IE 6.0 always get it right.

I'm not sure what the preferred notation is. I guess if you leave
the v2 output as it is, and use "utf-8' for v3, we would be pretty safe.

Thanks,
Pim