[
Lists Home |
Date Index |
Thread Index
]
From: "Champion, Mike" <Mike.Champion@SoftwareAG-USA.com>
> I'm out of my depth here, but this argument doesn't smell right to me. I
> thought we concluded in the massive Blueberry thread a few months back
that
> #x85 probably should have been included in the S production in the first
> place, and wasn't mainly because of a lack of mainframe expertise among
the
> members of the original WG. (PLEASE set me straight if there was a better
> reason, I don't claim to know much about mainframes myself except that
they
> generate a good bit of the revenue that pays my salary <grin>).
Actually, at the time that XML 1.0 was created, the control characters 80-9F
were undefined in Unicode. Actually, I believe they still are (I am writing
this away from an internet connection): Unicode only says what control
characters they are by default, but a particular application can map them to
private semantics.
At the time XML was developed, I called for whitespace and separator rules
to be determined by a rational basis, rather than ad hocery. I believe
there are only two rational bases: either
* all visible whitespace should be allowed as sepchars (with the exception
of NBSP for obvious reasons--it should be an error) and all new-lining
characters should be newlines, or
* parsing considerations should rule, in which case only <127 characters
should be used (e.g. so that XML in most ASCII family encodings can be
parsed without regard to the specific encoding)
The XML WG adopted the second, which is not my preferred choice, but at
least is rational. The XML 1.1 suggested rules seem a retreat from
rationality into lazy committee thinking of piling on special cases.
Cheers
Rick Jelliffe
|
