"Martin J. Duerst" wrote:
At 17:15 00/03/14 +0100, Chris Lilley wrote:
MURATA Makoto wrote:
We are all aware of this problem. We are also aware of transcoders
which changes the charset parameter but does not rerwrite encoding
declarations.
Yes - such behaviour is clearly broken. Since a transcoder is changing many
or all the other bytes in the file, expecting it to also correctly update
the encoding declaration rather than leaving it broken is not asking too
much.
I completely disagree. A trancoder transcodes. A transcoder may know
about a few (or a lot of) encodings. It is absolutely unreasonable
to ask for a transcoder to know all kinds of data formats, and
where in that data format some encoding hints are hidden (if they are).
So, rare and unusual formats like XML are probably not worth the hassle,
then?
Making this case slightly simpler to code, at the expense of throwing away
the excellent advantage that XML has in terms of only parsing when the
encoding is correct and making it wooly and ill defined in common cases
like *transferring XML between systems* seems utterly crazy.
If we had only XML on the internet, that would be different,
but that's not at all the case (yet).
Exactly. Not only is it on the internet, but it is also everywhere else,
frequently accessed by a local file system. It's not the only format, but
equally is a very important one.
--
Chris