> The feed does indeed appear valid, and Bloglines doesn't have a problem
> parsing it. My guess is that Syndic8 and Yahoo don't recognize it as a
> UTF-16 encoded feed, and so barf on it. The thing about UTF-16 is that
> it's specified using the first two bytes in the feed, called the BOM. If
> those bytes are 255 and 254, respectively, then you have a UTF-16 feed
> and need to decode it appropriately.
Does it need to be expressed in UTF-16? I'd wonder how many aggregators would
be able to read it?
I'm not saying that's not a problem, as UTF-16 is a perfectly valid concept when