Re: well-formedness error

[Julian Reschke <julian.reschke@xxxxxx>]

Paul Hoffman / IMC wrote:

> At 10:27 AM -0700 6/18/04, Tim Bray wrote:
>
> > You know, we could specify that Atom MUST always be encoded in UTF-8 
> > and/or that the root element must be <Atøm>.  Then, we'd have 
> > belt-and-suspenders safety in the face of the most deranged encoding 
> > breakage.  No, that's probably not a serious suggestion. -Tim
>
>
> Disclaimer: I'm an interop person, not a developer.
>
> Given the number of edge cases this thread has brought out (many of 
> which have bitten developers over the years), I think that mandating 
> UTF-8 is a reasonable serious suggestion. It would eliminate all the 
> edge conditions by saying "if you create something using other than 
> UTF-8, I assure you I will not be able to figure it out". That should 
> cause folks on the creation side to fall into place quickly.

How about modifying the proposal to say "either UTF-8 or UTF-16"?

Reasons:

1) AFAIK, for many languages, UTF-16 will be more efficient than UTF-8

2) XML parsers are REQUIRED to support just these two encodings -- so 
anybody serving XML content in some other encoding is already risking 
that the recipient will not be able to decode it.

Best regards, Julian


--
<green/>bytes GmbH -- http://www.greenbytes.de -- tel:+492512807760