Re: What to do with Xconq

[Stan Shebs <shebs at shebs dot cnchost dot com>]

Ari Rabkin wrote:
> 
> > Ari Rabkin wrote:
> >>
> >>     And the ones that do work need to be better documented.
> >
> > Sigh, more writing - doesn't anybody else want to do some of this?
> 
>     I'd be happy to!  Just noting it needs to be done.  I'll see what
> I can do.  Where do I submit finished or quasi-finished stuff?

You can send diffs against a snapshot to this list, or to me or Hans
directly and we can check in.  If you develop a taste for this, I can get
you set up to check into CVS on sourceware directly.

> >>     I have noticed that all xconq games seem alike.  This isn't good:
> >> the player should have to act and think differently in each game, not
> >> just exploit the same strategies.
> >
> > A good observation.  This is probably a side-effect of unimplemented
> > scorekeeper types, combined with lack of support code in the mplayer,
> > so every game's goal has to be to wipe out the other player.
> 
>     I suspect the prime cause is that, aside from flavor, there is little
> difference between tanks, destroyers, napoleonic armies, and starships.
> Units and games should be more individualized; I suspect that many of the
> newer engine additions will help with this.

There are a lot of unused or underused GDL features, true!

> >>     2)    Civilization is more than a war game; there is the dimension
> >> of economics, science, and city management.  Civ nations are not always
> >> at war or in (like the sides in some xconq games) perpetual alliance.
> >> Xconq is able to do science and economics to some extent with the "advanced
> >> city" code, but diplomacy is still a giant gap.
> >
> > Yeah, that's what agreements are supposed to be for, but that's something
> > that turned out to be a bigger job than I originally expected...
> 
>     Can we have that added to the "subsystem-to-finish" list?

OK.
 
>     Is the support for multiple combat models in place? If so, I'd be
> willing to have a go at writing one at some point.

Look at refs to g_combat_model in the sources; it's basically just case statements
on its value.  0 is traditional, 1 is Civ-like, leaving about 2,000,000,000 values
available... :-)

>     I've been slowly tinkering with a "modern" version of standard--it seems
> fairly playable; is there a method in-place for submitting this sort of
> thing?

Just send it to this list, either as a pointer to a file online, or in the message.
I'm sure a lot of people will be interested in giving it a try!

> > "ancient.g" is my idea of an AoE clone - it motivated me to add the
> > collecting task.  Yes, it's unfinished too... :-(
> 
>     It seems pretty good--it passed, at least for me, the "keeps player
> up late at night" test.

Wow!  But the AI didn't really do anything, right?  It doesn't have any
code to understand about foraging and such.  Interestingly, this is a
"real-time" game, in that turns timeout very quickly, so time passes
even if you don't do anything.

>     What is the hex-to-square formula, btw, if you know it offhand?

The easiest translation is to treat the square-grid map as if it were
a hex map leaning over to the right.  Then Y just maps to Y, and you
find the hex map's X by subtracting 1/2 of Y, X -> X - Y / 2.  Usually
you have to do some cleanup to compensate for the hex map having only
six neighbors to a cell, and no corner adjacencies to exploit for
things like canal/isthmus combinations.

Stan