On Tue, Aug 6, 2013 at 1:41 AM, Kenneth Russell <[email protected]> wrote: > The Encoding spec at http://encoding.spec.whatwg.org/ seems to have > handled issues like these. Perhaps a better route would be to fold > this functionality into that spec.
Yeah, I think my preference would be at this point to expose API-only encodings there. One of those could be base64. Labels for those encodings would simply not be recognized for <form> and URL. We could even give them labels that suggest that, e.g. "api-base64". Another one I've heard requests for is true latin1 which we also use in XMLHttpRequest for various HTTP-related things. -- http://annevankesteren.nl/
