Disease Model Programming

[VetLegion]

It does not pay off to allocate, say, 12 bytes. One byte (8 bites) is the smallest unit of memory you can get. Ofcourse, you can reserve, say, 30 bytes, and make a class/function to manage that memory to 5 variables, but it isnt really worth the effort.

The size of map in civ2 is roughly 100x100 tiles and that makes wonderful gameplay. And yet, managing many units on that scale can get really tedious. I am confused a little: Is a hex occupied by unit/city our "basic terrain unit"? If unit/city, can ocupy only one hex, it will limit us to not much more then 100x100 of these. If hex will be smaller (I assume so) then unit/city can ocupy more then one, or exactly as many of them as can be fitted to a bigger hex shape (radius to be decided later?). am I right?

We should keep in mind that "movement hexes" or "big ones" will be _set in stone_ when the map forms. That can make a few issues with cities: if cities will emerge from "basic terrain hexes" we will need to decide city center for uccupying purpose. If city can spread to more the one "big hex", which one is "main"?

Another issue is with terrain. If a "big hex" can consist of different smaller ones, that are actual base for terrain types, it can happen that "movement hex" has a combination of mountain and grassland hexes. How to calculate its movement points then?

I have a strong oppinion that smaller hexes are necessary. I have a fully 3D map in my "game vision" so I am confused with some of the issues I presented here. In order to have a 3D map, you need to have smaller tiles then "movement ones". Also, bear in mind that if we decide to let go of the civ2 system we risk two things: micromanagment, and loss of _strategic_ scope.

It was essential for all games in civ genre to have a well defined and simple map system. Although 3D map is my dream (Imagine actually building a dam on a river! ) I am more then willing to make a gameplay compromise.

[VetLegion]

4*8=32, sorry for typo

[amjayee]

VetLegion, unfortunately I'm not excactly sure what you are meaning. Could you clear it out? I think that the "movement hexes" will be the largest type of them; then later there could be smaller ones for graphics purposes. Were you meaning, that are we going to have larger hexes than the movement hexes, and the number of those larger hexes would be roughly the same as the number of hexes in civ2?

That has not been planned, at least yet. We were thinking of having same kinds of tiles as in civ2, but lots&lots of more, maybe 0.5 - 1 million. But now that I think of it, having larger tiles just might help us reduce the memory needed by the map significantly; Usually a large land area has roughly the same properties. We could store the properties for the large tiles, and some data of how the smaller hexes share those properties. Also we could store the tile population for the large hex, and for each smaller hex only the percentage from that population. Unsigned char would be enough for that, so only one byte! And so on. This would be more complicated, but allow much larger maps. But I don't know, just an idea.

[The Joker]

I'm sorry, Vet. But I am not sure what you mean, either.

I really like the small tiles we are using, as I think it would make the map much more versatile and realistic. And I see all kinds of ways to decrease the memory needed for it.

So the hex properties take 2 bytes. Could you (amjayee or Guildmaster) describe what these were? I remember the system you found ended up being lost with the map thread that disappeared. Anyway, even if we could save 1 or 2 bits it would be worth it. If we save a little amount a lot of places things should be workable.

------------------
"In America, first we take the sugar, then we take the power, then we take the women."
- Homer J. Simpson

GGS Website

[VetLegion]

My english is quite bad when I have to explain something complex, and even something simple sometimes.

Ok, first:
It _is_ possible to have a say, "one and a half" bytes for a variable, or 3 bytes and 4 bits for an integer. It would have to be done on application level since you cannot get less then byte from system. And I dont see real benefit. The answer to big memory demands is a smart caching system using disk. It would not be a big preformance hit to use disk I think, but it remains to be tested.

Secondly, do you people think that those 0.5 - 1 million hexes would be both:

- basic terrain tile
- basic unit movement tile
?

If I had a unit on a hex with a movement factor of one (say, settler), and I go west, would it take me 1000 left clicks on my keyboard to get back to my position?
(1000*1000=1000000 hexes, cylinder map)

If it would, well we should think about reducing the number

If it wouldnt, than we have two types of hexes (assuming we have hexes for both things, not squares):

"terrain hexes" that are small and terrain and population are based on them. there can be any number of these.

"movement hexes" that are bigger then those making the terrain, and used for strategic planning, and movement of units. There should not be more then 200 these on the largest of all maps (200 hundred would kill the game actually, more like 100).

In civ2 both hex types are same, so a square is a "basic terrain hex" and a "movement hex". There is a possibility to disband "movement hexes" and define unit speed and movement as a function of the TH but I am trying to say that it would have great gameplay impact, and at this time I am against it. I bet that civ3 will have "tiled" movement CTP2 has. I only saw it ditched in one game, and the solution is not very strong.

There is game CNW (Conquest of New world) try to see on http://www.interplay.com . Too bad the demo doesnt work on my win2000 system, but I played it on win95. It is a civ/colonization type of game that has quite a few revolutionary changes. The game is very good, but is underestimated a lot, and made very little sales. One of the things they did was loosing the "tiled" movement. If you think we should loose it too, take a look at their demo, since there is no other way of doing it. It does take some restrains away but the overall impression of the game is: far less strategic then civ2, even colonization.

So: we _cannot_ have 1000000 movement hexes! I hope I understood that one wrong. It would not only be a micromanagment hell, but it would also require a new keyboard each game

note: I am using word "tile" for both hexagon shape (our) and the rectangle shape (civ2). I really dont know any other english word for that. If "tile" has in its meaning a rectangle shape then this is wrong. But I only found some reference to "bricks" in my dictionary so I assume tile can mean "basic building element of something". So in context of map I think of "tile" as a basic building element of map without thinking of particular shape (although, the shapes that fit together are limited to square and hex)

If I havent been clear, just say so and I ll explain more.
[This message has been edited by VetLegion (edited November 22, 2000).]

[The Joker]

I think we should try to think more away from Civ2.

What we had in mind with movement was, that it should be done very differently from Civ2.

First, we are using amjayee's preplanned turns. This means that we wont be moving units directly tile by tile, but in stead choose where the unit would go, and then when we are done moving all the units we would click Start Turn, and all units would move simultaneously.

So we wont be using the keyboard for movement. In stead the mouse should be used, and we would simply click on the hex where we want the unit to go, and it would do it. So this way many tiles wont really matter.

Second we should have far less moving units than in Civ2. We will be stacking units into armies, so maybe a large civ would have only 5 moving armies at any time. This will reduce tedious unit management.

So yes, I think movement hexes should be the same as terrain hexes. And yes, there could be 1,000,000 of these on a map, since movement will be done in another way.

------------------
"In America, first we take the sugar, then we take the power, then we take the women."
- Homer J. Simpson

GGS Website

[VetLegion]

I am glad that you understand what I am saying. The fact that we couldnt disagree more is less important

I see what you mean. Ditch the movement "tiles" and make our own system of movement. I see how this could work (CNW). Well, if we manage to do it, it would have advantages.

All areas of game will benefit from small hexes: better terrain, better population, better diseases, citys, economy etc. Only thing that could (not that it necesserily will) suffer from it is the military model.

As you said about 5 armies, small hexes explicitly demand of us to abandon unit micromanagment (tactical moves), and leave it to our generals(characters).

To allow the players to do planning on the tactical level, is impossible.

Sigh, how many times have I built my economy, prepared my maps, and manipulated the diplomacy in civ2, preparing to attack. I took great pleasure when the moment came, to micromanage my units and prove my tactical genious too, more to myself then the dumb AI. I always went for victory as early as possible ... meaning it was never secure. I always had too little units, almost no reserves, no money ... It required great concentration to achieve victory, since every single unit was important.

Inluding characters in our game means we will have to have a certain "random" factor in every action they do. If player does not have direct control over his army, but his general has ... the outcome of the perhaps most crucial battles is not under players control. If I see my general losing a battle I would have won (if I had the possibility to do tactical orders) it will make me very annoyed, because of the "streched" style I play.

So while I instinctively disagree, my logic tells me it may be just the way to go. In its base, civ is a war game, the AC and other "peaceful victories" are just for people who should play simcity instead If we manage to keep war fun, while abandoning the current units/unit movement/unit micromanagment sytem it will be acceptable. But we have to design some wicked things to do to those generals who loose a battle. Hmm, a torture room instead of throne room maybe

[amjayee]

Hmm, the discussion here is interesting. I disagree slightly with both of you.

Here is my vision of the movement/tile/warfare system.

1. Units movement is tile-based, as earlier.

2. Units are not, though, moved around tile-by-tile; instead we would have a sophisticated auto-moving system. Or, if the player wants more control over his troops, he can point where the unit should go; a route appears on the map. Then player can drag-and-drop the points making up the route to change what course the unit will take.

3. Warfare will be made more fun and more realistic, thus more challenging. In civ2, warfare at its best was quite fun. But at some point it turned into rushing large amounts of units against other players. In our game, part of the challenge is, that building armies requires resources - men being one of them. So, the most vital part of warfare is also economy and resource-handling. You have to decide how much you can conquer this time, without jeopardizing your troops, or your empire. Also, war would be more like clever maneuvering of armies and striking where it is most advantageous, rather than attacking mindlessly the nearest enemy. This includes also scouting, supplying, etc. as Joker said. You would have generals and commanders, but you could have two ways of using them; either their quality affects only how their army succeeds in the battle, and you command the armies into their positions. Or, you could give warfare entirely for your commanders; you could give some rough orders, and they would try to achieve that at their best.

I agree we need to move away from civ2 system, but keep the tiled movement. Also warfare needs to be fun and rewarding.

About tactical commanding of troops, that is quite impossible in our game, and quite mindless also, because of the scope. But I'm sure strategical commanding will prove just as fun. We just need to give much of thought for the system. Also the player should be given a large amount of commands he could give to the troops, so their behavior could be adjusted for the needs of each situation.

I hope this doesn't complicate the discussion here, or make it unfriendly.
[This message has been edited by amjayee (edited November 26, 2000).]

[VetLegion]

Amjayee, you have good points there. I know you and others have thought about this more then I had, so I ll read that before I make any more "grr I hate this, but dont really have idea how to do it better" comments

Also, we kind of stole this thread from heardies disease model. This type of discussion belongs to "Abstraction, Scope, etc." thread, so lets try to hop there if discussions like this arise.

[ElmoTheElk]

I agree with you Vet. Let's go to the Abstaction, Scope, etc. tread. There you can also read my vision.

[The Joker]

I think that the small tiles would not necessarily mean that the battles were out of the player's control.

Yes, the player wont give orders to every unit in a tactical mode. But he will tell the unit what province to capture, what other army to engage etc etc.

So I don't think it would mean leaving all that to the commanders. What will be the benefit of the armies and the "new" movement system is, that the player will only have a few armies, not hundreds of units, to move around. There would still be battles. But they would be rarer (since there would be fewer units) and warfare would include other things as well, like scouting, espionage etc.

------------------
"In America, first we take the sugar, then we take the power, then we take the women."
- Homer J. Simpson

GGS Website