XOM API for Processing XML with Java

[luca <passani AT eunet.no>]

Elliotte Harold wrote:

> There is not. This is a deliberate design decision. In my opinion, 
> always processing the full DTD is the safest option for working with 
> XML. It's the 80/20 cut, I made.

I don't quite get why this is on the 20 side. It seems to be on
the 80 side to me.


> > Generally speaking the idea of making the well-functioning
> > of a software system dependent on a resource out of one's control
> > does not seem wise to me and I definitely want to avoid this situation
> > in my system.
>
>
> I agree. My response is that this design flaw is built into XML.

as someone else said, this can go in the way of leading
people to refuse XML, a technology that should have been
simple by definition, but obviously is not.


> XOM 
> faithfully models XML, including its warts and ugly parts like the 
> necessity of the external DTD subset in order to guarantee that you 
> receive the document's full and correct information content. External 
> DTD subsets can affect:
>
> 1. Default attribute values
> 2. Entity replacement text
> 3. Attribute normalization

mmmm, I don't get it. How hard could it be to have

  Builder builder = new Builder();
  builder.ignoreDtdNoMatterWhat();

?

Luca