On 17/07/11 11:21, Olivier Jeulin wrote:
> Le 17/07/2011 00:18, Richard Salz a �crit :
>> Michael Kay <> wrote on 07/16/2011 03:32:16 PM:
>> > or it could allow "&" to represent itself if not followed
>> > by a name character.
>>
>> Then whitespace becomes significant
>> foo<bar
>> foo < bar
>> foo< bar
>> foo <bar
>
> It is already. To define a start tag, "<" can't be followed by a space "
> ", so
> foo < bar
> foo< bar
> could (and I think should) be valid XML
But this *is* valid SGML.
<!doctype text [
<!element text - - (#pcdata)>
]>
<text>
foo < bar
foo< bar
</text>
This was one of the complexities of SGML that was removed to make XML,
in order to *make it easier to write parser/validators*. There were lots
more, and I'm sure no-one involved at that stage would claim that
everything was got right. Don't forget that XML was designed for *text
documents*; any other use was either an afterthought or simply Not Yet
Invented.
///Peter
