XOM API for Processing XML with Java

[Elliotte Rusty Harold <elharo AT metalab.unc.edu>]

At 7:18 PM -0800 2/19/04, Brian Sam-Bodden wrote:
> Rusty,
>    I haven't done a pound per pound comparison of
> marshalling versus parsing. But in most of our
> scenarios the object is already in memory so there is
> no marshalling involved and we get it by reference.

If it's already in memory, why do you need to serialize and reparse 
it at all? Wouldn't you have the unserialized XOM object handy?

>  The
> problem is that to get it to some of these caches it
> needs to be serializable. Even if it gets to go through
> a round of marshalling/demarshalling to get to the
> cache it is still better than for each client that
> accesses the object to have to parse the XML again.

Again I ask you, how do you know that? What evidence do you have that 
marshalling/unmarshalling is faster than serializing/parsing? You're 
making an assumption here that I've seen no evidence for. Object 
serialization in Java has a long history of being dog slow. if you 
want to claim it's faster than XML parsing you're going to have to 
measure it.
--

  Elliotte Rusty Harold
  elharo AT metalab.unc.edu
  Effective XML (Addison-Wesley, 2003)
  http://www.cafeconleche.org/books/effectivexml
  http://www.amazon.com/exec/obidos/ISBN%3D0321150406/ref%3Dnosim/cafeaulaitA