Real progress.

I just commited wrtl's latest, with some tweaks.
@wrtl: make sure you do some cvs updating before you start again. After doing the CXXFLAGS thing zman asked for, I found a number of things that had to be corrected.
DTD for cockpits and the standard cockpit are committed, however, the standard cockpit doesn't actually contain the entire hud that we have. Since it isn't loaded, no problem right now.

But it is there for all to see. It's in "used it to write the DTD" shape right now.
For all, tXmlParser is ready to be used in the map parser. It will validate and load a map into the DOM type, which is what gParser does. A few changes will need to be maded to gParser to support it, I don't know the extent of the impact of those changes, so I didn't make them. If philippe can dream up a way to reduce even that without breaking the interface of tXmlParser, fine.

Otherwise, it's ready. The SAX interface it provides is thoroughly untested! THe model for use that wrtl's laying down, um, I'm not entirely sure is correct, but I don't think it's blatantly wrong either (nad it was one of my suggestions, so I'll take responsibility for it if it turns out to be wrong). ANyway, gHud inherits it, rather than creating another class and instantiating it within gHud.
So, tXmlParser will do the minimum required of a base class for the map parser, but the interface might need some tweaking. I feel like it's a good interface, but y'all check it out. It should simplify making new resources a bit, but for the DOM mode, not significant. If you use it in the untested SAX mode, then it gives you native armagetron/std data structures instead of libxml2 data structures. I suspect we'll use SAX for the cockpit, that seems like the right parser to use there.
That's it. gHud doesn't actually use the map it loads into memory yet, wrtl went to bed before that happened.
