[jdom-interest] speed of DOM vs JDOM

[Jason Hunter]

> > Just be aware that someone needs to step in and write DOMOutputter
> > still.  Since it sounds like you may have a need for that class and have
> > a knowledge of DOM, perhaps it could be you.  Brett estimated 3 hours to
> > write it.
> 
> Hmmmm, I could have sworn that Kevin Regan was writing that...otherwise I
> would have jumped on that a while ago.  I've got time this weekend to give
> it a crack.

Last I heard his day job had to take priority, so the job's yours for
the taking.  I'm trying to keep TODO.txt updated with things still to do
and who's working on them.  I just put your name next to DOMOutputter. 
:-)

> > Of course, one reason it's not written is that people find ways to avoid
> > DOM.  If you could avoid ever taking your document back to DOM you'll
> > get a definite performance/memory improvement.
> 
> Of course but DOM is W3C after all and in order to leverage the standard
> most of our customers are seeing/using (for now anyway), it is much easier
> to ready our infrastructure to pass a client (GUI tool) a DOM as opposed to
> try to convince everyone to use JDOM...besides, not everyone is as smart as
> you guys ;)  Perhaps in a few months JDOM will kick enough ass that this
> won't even be an issue ;)

Best you point out that SAX isn't a W3C standard either.  It's a
commonly accepted community developed open source library.  Kind of
like...

> Perhaps many of my questions will be answered in writing the DOMOutputter.
> If you or Brett have an outline in mind for writing it (or at least a
> starting point) I'd be glad to get on it.

Well, the method signatures needed to be changed, so I just changed them
in CVS to this:

    public org.w3c.dom.Document output(Document document);
    public org.w3c.dom.Document output(Document document,
                                       String domAdapterClass);

You can look to DOMBuilder to see how we use the adapter as a way to
choose a DOM implementation.

-jh-