Re: XML and whitespace: lets just dump CR and LF!

Peter Murray-Rust (Peter@ursus.demon.co.uk)
Wed, 06 Aug 1997 16:12:53 GMT

In message <199708051317.XAA23619@jawa.chilli.net.au> "Rick Jelliffe" writes:
[...]
>
> I suggest that the following approach should be taken. (I think it is the only
> realistic solution, especially if we assume that 1)
> data is usually generated by applications,
Although this will be partly true, I think we still have to expect people
to use text editors for a year or two yet :-). [It's how I create most of
my XML at present :-)].

> 2) humans only check and tweak data;
Yes. XML must certainly be tweakable. So it mustn't have to have lines 1000
chars long :-)

> 3) we want operating system
> and character set independence,

critical :-)

4) line-breaking is generally done by clients
> ...so CR/LF is basically a convenience for fitting data into editors,
> not for the purposes of output.)

Yes.

>
> **A) XML applications should ignore *ALL* CR and LF as a bad joke. They should
> be entirely there for formatting the raw text into nice, eye-sized records.
> So CR and LF should never be converted to spaces. (This approach was the
> one taken by Interleaf, and I have come to appreciate it.) If you need a
> space, then start the new line with it! (Ending the previous line is difficult
> to see.)

Appeals to me :-)

>
> **B) XML applications should mandate the use of the unambiguous Unicode characters
> -- LINE SEPARATOR &#x2028;
> -- PARAGRAPH SEPARATOR &#x2029;
>
This makes sense unless someone finds a flaw in it...

P.

-- 
Peter Murray-Rust, domestic net connection
Virtual School of Molecular Sciences
http://www.vsms.nottingham.ac.uk/