Re: XML, DOM and "ignorable" whitespace

keshlam@us.ibm.com wrote:

> >The long and the short is that this is a tough problem but it is also
> >disappointing from a basic developer point of view none the less.

Reworded. It makes it harder for the new developer that doesn't know all the
ins and out and leads to frustration and not new XML / DOM applications.

> "We're making mistakes as fast as we can..."

Making some good stuff as well.

> As I said: It's on the list of issues for DOM Level 3. Level 2 considered
> it but decided it couldn't be done properly without DTD support.

I agree this certainly can't be done without a DTD.

> Early filtering of the tree ("lazy initialization", as you called it) is
> certainly possible. Maybe the Load/Save chapter of DOM Level 3 can look at
> that side of the issue.

Good idea- early filtering would work great for applications that validate
the XML that don't need the nodes
but comes at the cost since the non-validated document would still end up
with the nodes
since the DOM couldn't know what text is not needed.

Thanks,
Eric Richardson

Received on Friday, 2 June 2000 12:57:45 UTC