Re: DOM and SAX parsing in Ada

[Nick Roberts <nick.roberts@acm.org>]

"Jeff C" <jcreem@yahoo.com> wrote:

> Ok..Forks are bad...but is it really "better" to create a separate XML/Ada
> thing that just starts over rather than forking all together? I suppose
> that is true if you have fundemental problems with the existing
> approach....

I don't have any fundamental problems with the approach taken by XML/Ada,
except that it doesn't have any maintenance documentation, and AdaCore don't
seem to think any is necessary (isn't that weird?).

But the DOM interface is not very suited to Ada, and neither is the SAX
interface. There are lots of little technical niggles. One of the problems
with XML/Ada 1.0 is that it doesn't support (read?) DTDs. For my own use,
although I don't need validation, I do need to be able to declare entities
in the DTD that will be correctly expanded in the document (I think this has
significance for default attribute values).

> But I guess I don't see the downside to a public fork v.s. a public
> re-write of the whole thing from scratch.

It's a question that I contemplated quite a lot myself. I agree that the
answer isn't obvious.

> My guess is ACT is just afraid to deal too closely with you because once
> you finish your OS and new Ada compiler you will be a major competitor to
> them.

I guess that's a joke, but I don't see how AdaOS or ECLAT could be
competition for AdaCore. Quite the opposite, since both will be released
under the GPL, there would be absolutely nothing stopping AdaCore (or anyone
else) from selling support for either or both, just as they sell support now
for GNAT and other GPL software. They /should/ see AdaOS and ECLAT as a
commercial opportunity.

-- 
Nick Roberts