Xconq Maps from GIS Data

Xconq Maps from GIS Data

[D. Cooper Stevenson]

Dear Xconq Developers,

I am working to build an aggressive new image set for XCONQ. I'll just
come right out and say it: I want to build a map made from actual
Geographical Information System Images.

You can "surf" (and download) GIS data here:

  http://seamless.usgs.gov/

If your browser is not capable of viewing this data (best running
Firefox on Linux), it's the most impressive view of the Earth's surface
you've ever seen. In effect, you would be playing Xconq on a
super-realistic playing field.

See the colors? It turns out that everything is color-coded and roughly
matches Xconq's terrain system: yellow for plains, green for mountains,
light blue for sea, etc.

Additionally, it is possible to create real-time 3D views with GIS data
as shown here:

  http://grass.itc.it/demos/ortho3d.jpg

Please note that I've studied this for several days and scanned the
archives. I've not seen this asked before.

The following is a tentative plan, is it possible? Is there a better
way? 

  1) Download the GIS data and convert the image to a palatable image
format
  2) Overlay a hex grid (to scale) on the image
  3) "Break off" each of the hexes and place them into a file similar to
those found under the /lib/images directory painstakingly categorized in
order by terrain type
  4) Create a terrain file with the terrain information

I have created a complete outline here:

  http://www.gencom.us/xconq_design.html

RATIONAL for Xconq:

Xconq's engine can't be beat in my opinion. It would be a travesty to
take out of hand the hours put into all of the unit interactions,
movement tables, etc. Why re-invent the wheel?!!? Xconq is just
impressive, period. From my research, Xconq is the most mature tactical
game available.

Long term, just to throw it out there: is it possible to interface raw
GIS coordinate data (longitude, latidude, etc.) directly with Xconq's
engine?

I'm willing to perform the painstaking work if you're willing to point
me in the right direction. 


Best,


-- 
--------------------------------------------------------------
| Cooper Stevenson        | Em:  cooper@gencom.us            |
| GenCom                  | Ph:  541.924.9434                |
| "Working For IT"        | Www: http://www.gencom.us        |
--------------------------------------------------------------

Re: Xconq Maps from GIS Data

[Lincoln Peters]

On Fri, 2004-09-17 at 20:23, D. Cooper Stevenson wrote:
> Dear Xconq Developers,
> 
> I am working to build an aggressive new image set for XCONQ. I'll just
> come right out and say it: I want to build a map made from actual
> Geographical Information System Images.

Welcome to the list.  This sounds like an exciting project, and I think
we would open up countless possibilities for new games.

I suspect that Xconq's existing terrain libraries might need to be
modified (or a new, separate terrain library be written from scratch) to
accept the data you describe.  I haven't looked at it as closely as you
have (so far I've just examined the maps of my native San Francisco Bay
Area), so I have a few questions about the GIS data:

1. Does it track vegetation separately from terrain?  If so, a terrain
library based on the GIS data would do well to implement vegetation as
terrain coatings, rather than try to make cell terrain for every
combination that exists in the world.

2. Does it provide meteorological statistics, such as average rainfall,
wind patterns, etc.?  I can see that a game based based on something as
detailed as GIS data might want to include such statistics, although I'm
sure that if they're not part of the GIS data, we could find them
elsewhere?

3. Does it provide any detailed information about area, such as soil
composition or specific kinds of vegetation?  This probably would not be
important for most games based on a world map, but I could see it being
used to generate more realistic maps with the existing terrain synthesis
methods.


There are a few technical difficulties involved with such a project that
I can see already.  One thing that comes to mind is its representation
of rivers.  I'm not sure how we would translate them into Xconq maps;
depending on the nature of the game being built, it might be most
logical to represent them as border terrain (which is how it's usually
done), connection terrain (which is how it's done in bolodd3.g and
panzer.g), as cell terrain (at a very close-up scale), or maybe even as
coating terrain.  I'd be curious as to how the other Xconq developers
would handle this and why.

---
Lincoln Peters
<sampln@sbcglobal.net>

Fortune finishes the great quotations, #12

	Those who can, do.  Those who can't, write the instructions.

Re: Xconq Maps from GIS Data

[Eric McDonald]

D. Cooper Stevenson wrote:

> I am working to build an aggressive new image set for XCONQ. 

Excellent.

> The following is a tentative plan, is it possible? Is there a better
> way? 

One of the other contributors to the list, Matthew Skala, created a map 
of Antarctica from USGS data, IIRC. You may be able to get some tips, 
and maybe even a code or script (?) to help in this regard. His focus, 
though, was in using existing terrain images on a 
geographically-accurate map.

>   1) Download the GIS data and convert the image to a palatable image
> format
>   2) Overlay a hex grid (to scale) on the image
>   3) "Break off" each of the hexes and place them into a file similar to
> those found under the /lib/images directory painstakingly categorized in
> order by terrain type

When doing so, one would probably want to automatically generate the 
corresponding .imf files (found in the 'lib' directory of the sources) 
whihc actually reference the images in the .gif files.

>   4) Create a terrain file with the terrain information
> 
> I have created a complete outline here:
> 
>   http://www.gencom.us/xconq_design.html

I looked through your outline.

As far as your question regarding WCS (which I assume is what you meant 
by "Global Coordinate") data, I had not given that consideration before.

I am not sure whether Xconq should be made to take such data directly in 
.g (Xconq Game Definition) files, or whether it should be first be 
transformed to Xconq's natural coordinate system. I would probably favor 
the latter.

As far as your concerns about unit movement, and if I understand you 
correctly, one would need a translation program that bins terrain types 
  according to elevation variation, and then takes a supplied set of 
unit types and generates movement rules against these bins. The 
resulting Xconq .g output might be something like:
(table mp-to-enter
   (all-terrain-land-u* land-t* 1)
   (flat-terrain-land-u* land-type1-t* 1)
)
where 'land-type-t*' would be the bin containing the smoothest/flattest 
terrain. Or something like that....

> Xconq's engine can't be beat in my opinion. 

It is quite flexible and powerful in my opinion as well. There are still 
plenty of things that can be (and are being done) as far as improving it 
goes.

> Long term, just to throw it out there: is it possible to interface raw
> GIS coordinate data (longitude, latidude, etc.) directly with Xconq's
> engine?

If you mean whether *.g files can contain maps that are directly from 
GIS data, the answer is no, that is not currently possible. But, it is 
an area of interest. Another, possibly easier, route would be to write a 
standardized translation tool for anyone who wishes to create maps from 
such data. I think that would be quite welcomed by the community.

> I'm willing to perform the painstaking work if you're willing to point
> me in the right direction. 

Great! If Matthew is still on the list, maybe he can give you some 
useful advice. Also, if whoever created the giant world map is around 
(Stan Shebs, perhaps?), he may be able to give you some advice as well.

As far as the terrain images themselves go, there are some things to 
consider: isometric views, possibly different image sets for each of 
Xconq's different scales, tiling versus hex images (hex images being 
prefered). If you do hex images, you will likely need to apply 
rectangular regions containing a hex mask while you are extracting the 
images from the map data. If you do tiles, you only need to sample 
square regions.

Eric

Re: Xconq Maps from GIS Data

[mskala]

On Fri, 17 Sep 2004, D. Cooper Stevenson wrote:
> I am working to build an aggressive new image set for XCONQ. I'll just
> come right out and say it: I want to build a map made from actual
> Geographical Information System Images.

I've been working on making such a map for Antarctica.  Haven't touched it
in a little while because I became busy with some other things, but a
preliminary version is at http://ansuz.sooke.bc.ca/temporary/map.g.  I
have a more recent version based on better data, for which I've also made
a simple game module translating the terrain types such that you can play
the default game on it.  I have in mind to create a more elaborate
specifically-Antarctic game involving the idea from Lovecraft's fiction
that Antarctica is full of the ruins of pre-human civilization, but I set
that aside for a while because I was heavily using the background economy
features, and those seemed to be changing too rapidly.  Anyway, if people
would like to see what I've got so far I'll slap some documentation on it
and post it somewhere.

My current version (not the one currently posted) mostly draws from the
RAMP DEM, which is a digital elevation model pieced together for use in
correcting RADARSAT data.  I got it via
http://nsidc.org/data/nsidc-0082.htm - if I remember correctly one has to
fill out a form to get access, and the mini-essay I wrote explaining how
this was an educational project was no doubt entertaining reading for
whoever approves such things, but they did approve it, so fine.  I also
got some surface classification information from another US Government
source that I don't remember off the top of my head (I have it written
down somewhere or other).  The special challenge for Antarctica was coming
up with meaningful terrain types, since the continent is basically just a
sheet of ice.

Most of the data I got came as ASCII lists of points with decimal lat/long
coordinates.  I wrote some Perl scripts to translate the ASCII lat/lon
coordinates into hex-grid coordinates, take a vote of surface types within
each hex grid, and then assign the cell the majority terrain type of
points within the cell.  There were a few other subtleties which, again, I
don't remember off the top of my head but have written down to go into the
document that will eventually accompany the game module.

From a game perspective, I think one thing you'll have to face is that
there won't be a universal automated way of doing this - you will have to
customize your GIS/Xconq translation to the individual map you're
doing.  My Antactic map is probably the extreme in that line, but
the same will be true of any map.  You'll also want to fuse multiple data
sources if you want to get a really good map.

I have some background in this stuff because I worked for just over a year
total, in a couple of co-op and contract terms for a Canadian government
remote sensing lab.  If you or others want to talk about making Xconq maps
from GIS data in a serious way, I'd be interested in participating.  It's
also possible that we might even put together a paper about it that we
could publish at a conference or something, since doing it right will
actually involve doing some new research on GIS that the GIS research
community would probably find interesting.
-- 
Matthew Skala
mskala@ansuz.sooke.bc.ca                    Embrace and defend.
http://ansuz.sooke.bc.ca/