Echoes of the Imperium (4X Strategy Game)

[SiENcE]

Looks great so far.

I hope you finish it!

[Lap]

Umm, why not make it so unit's don't need to "react first"? It doesn't make sense, it's not intuitive and to be honest.. it's getting boring

Turns are abstractions of long periods of time. Odds are someone is going to get there first so it does make sense.

Not being intuitive is arguable. No one is going to need to sit there comparing values of theirs and the enemy units initiative (which you can't see anyways). Players should be able to glance at their armies and get a sense for how "fast" they are. If it's really a big deal I'll just have a descriptor word like fast,average, or slow. I'm not understanding how either way can be considered more or less boring than the other though.

I've played a lot of Dominions, which uses a similar province style system with simultaneous moves that function as you suggest. It causes a lot of problems.

Example: One unit is ordered away from an enemy, the enemy was ordered to attack.
Since both units act simultaneously, both move into the same direction - nothing happens.

This is actually a hugely annoying problem. Fleeing armies can remain fleeing for a very long time, sometimes indefinitely. Most often there are three or more possible choices for an enemy armies movement and a lot of times you just have to guess. It also doesn't make sense that my pursuing army would simply continue in a totally wrong direction and not notice.

Another example: Two units of different players are given order to conquer a city tile (or whatever) - both actually move there! The result would be an encounter between both, the winner gets it all.

This is a nightmare for the combat system and no solution is at all intuitive to the player. Who would the defenders target first? Why? Would two enemies also fight amongst themselves while fighting the city's garrison?

On a side note-Allied units targeting the same province can be told to coordinate their attacks, meaning for the purposes of this turn, the fast units will wait for the slow ones and they will attack at once.

Another scenario: X attacks Y's territory, Y attack's X's territory. Both armies would simply pass each other by and switch territories. It's not that hard to solve this specific situation by checking all orders to see if direct switches are happening and then have those play out like meeting engagements, but there are many more complex situations (that I'd probably have to show a diagram for) where players would end up switching territory without any engagement.

Having armies bypass each other ends up slowing the game down quite a bit and keeping the length of turns and the length of games down is very important since I'm focusing on multiplayer.

TL;DR- Having someone move first means less technical problems, more combat, and faster games.

Go with the "gritty" UI. Definitely the best and most fitting art style!

What style of maps would you prefer to go with a gritty UI?

[ljdp]

Hmmm.
The simultaneous moves could work with a rock paper scissors format.
Player A moves rock to a tile, Player B moves paper to the same tile, Player B claims the tile.
So what If player swap positions? Well you could incorporate this into the strategy, leaving tiles empty could lead to players sneaking past.
Perhaps objects on adjacent tiles could affect, like defences.

[Lap]

There's been a lot of discussion on various forums about this. I even thought about using a Hearts of Iron style system where armies could support neighboring provinces or automatically respond to changes. It started going down a road that was getting to complex and meant a lot more army micromanagement.

Here's one of the better threads if you are interested (http://www.gamedev.net/community/forums ... 1&#3660840)

A solution was suggested a while back that seems intuitive and simple. All attacking armies are "exhausted" on finishing an offensive operation. If you sidestep an enemy army and take a territory that army cannot move next turn (reorganizing, suppressing civilian population, whatever).

I've been using this in conjunction with the initiative system and it seems to work well. Nothings written in stone yet though.

Other stuff I could use help with:

-I'm currently using MD5 to validate player passwords. This means using an outside module, which means portability is a bitch. Ideas?

-Tech trees. I've done almost no work on these, or even researching in general. It's quite a daunting task.

-What were your favorite and least favorite aspects of other 4X games?. There's plenty of time left for me to add features and avoid past mistakes.

[thelinx]

Don't mind me, just reminding you that the LÖVE party is without a project and would be happy to have something to code/art away at.

[ninwa]

It started going down a road that was getting to complex and meant a lot more army micromanagement.

I'm curious why this is considered a bad thing. Allowing more micro-management adds another skill-set a player must master to become a good player. It will allow for much more interesting games to happen. I play a lot of SC2 though, so maybe I'm biased.

[Lap]

I'm curious why this is considered a bad thing. I play a lot of SC2 though, so maybe I'm biased.

I play a lot of SC2 too...but in my case SC2 is Supreme Commander 2, which is more on the strategic end versus Starcraft 2's more tactical gameplay. I'm taking a more strategic focus. As interesting as the Hearts of Iron system is it's not intuitive and unfortunately, it does turn into a chore having to set and reset so many commands. Even defending units need to have their commands changed often. It does add slightly to tactics, but I just don't know if it's worth the trouble to both me and the players.

Don't mind me, just reminding you that the LÖVE party is without a project and would be happy to have something to code/art away at.

Well that would be a quite an honor. I'm trying to think of good modular projects that can be easily hooked into the game; preferably modules that can be easily adapted to other projects so you guys get maximum benefit from your work.

Mind Mapper/Tree Organization

Specifically, something to map out relationship between "children" and their "parents". I would personally use this to show players the tech tree and specific research/upgrade paths.

Other uses:

-A separate concept mapper/idea web program
-Showing decision trees
-Maybe something like displaying the current, past and future stages in platform games?

Mapping software

There's always a need for random maps or map editors. How specific it would be to just this game depends on how it is made.


That's all I can think of off the top of my head. Though if anyone has an interest in coding anything else to help this project I'd be glad to have some help.

[kikito]

-I'm currently using MD5 to validate player passwords. This means using an outside module, which means portability is a bitch. Ideas?

How about this?

-Tech trees. I've done almost no work on these, or even researching in general. It's quite a daunting task.

You need two different sets of models.

On one side you would have the requisites/steps/researches/whatever you want to call them of each "race" or "group" (or just one big tree if you don't have several races). Each node needs a list of predecessors (or 'parents') For example, the node for 'Hull resistance +2' will have 'Hull resistance +1' and 'Advanced foundries' as parents. It could have other things, like the cost in materials, or time to develop.

Then you need to model, separatedly, the tree that indicates where each player is. It could be as simple as a table that uses the nodes of the previous tree as keys. So when the player researches Hull resistance +2 you do something like this:

Code: Select all

  node = player.race.techTree:getNode('Hull resistance +2') -- getNode is just an example, you could get this node with other means
  player.researchedNodes[node] = true

Each of your UI buttons related with teching would have a pointer to one of the nodes. When drawing it, you would check if player.researchedNodes[node] is true, for "activating" the buttons.

-What were your favorite and least favorite aspects of other 4X games?. There's plenty of time left for me to add features and avoid past mistakes.

I like emergence on these games. Simple rules that, when combined, create more complex 'meta-rules' that are not directly programmed in the game, but are still visible. I like having several paths to victory (being very happy or very rich) instead of plain old "destroy all your opponentzzz!"

I don't like 4X games (or strategy games, for that matter) that rely on giving the player an immense number of units with slightly different characteristics. I prefer having less variance in ships, but very differentiated.

[Lap]

How about this?

Looks like exactly what I need. Thanks. ("requires bitlib" - I'll have to see if I can get this to work later)

-Tech trees. I've done almost no work on these, or even researching in general. It's quite a daunting task.

You need two different sets of models...

Storing the tech trees and all the data behind it is a cinch. No problems there. The problems are in displaying it. I can't think of a good way to dynamically generate a visual tech tree without having all sorts of weird line crossing going on. This is probably why so many games have static or semi-static images for their tech trees.

I like having several paths to victory (being very happy or very rich) instead of plain old "destroy all your opponentzzz!"

Destroying everyone will remain a viable strategy, but the other main way to win is by being elected Emperor. I can explain the vote system more later, but in short there are a lot of different ways to accumulate votes. Votes can be stolen and bargained for so players that want a more sneaky/subtle path to victory can go for that. I probably won't be including as many victory options as the Civ series. Even on their 5th game I still don't feel like they have all the victory paths balanced. If they can't do it after all these years I'm hesitant to try becoming too grandiose in this area...at least initially.

I don't like 4X games (or strategy games, for that matter) that rely on giving the player an immense number of units with slightly different characteristics. I prefer having less variance in ships, but very differentiated.

I agree. Most players only end up using a few chosen types anyways. With how easy it is to make new units I'm still going to have to contain myself from making too many though .

[kikito]

Storing the tech trees and all the data behind it is a cinch. No problems there. The problems are in displaying it. I can't think of a good way to dynamically generate a visual tech tree without having all sorts of weird line crossing going on. This is probably why so many games have static or semi-static images for their tech trees.

Oh. I see. When each node has 1 and only 1 parent, it's easy; it's a tree, and it is easy to represent.

The problem comes when one node has more than one parent. Then suddently instead of a tree you have a graph. And you want to embed it on a plane. Graphs that can be embedded on the plane like that are called Planar Graphs. It's isn't a simple subject, as this query in stackoverflow shows.

First thing is determining whether your graph is planar or not.

If it is, then you will have to draw it. I've found two libs: OGDF, Boost Graphs. But they are C++, and not small. The translation work could be significant.

If the graph isn't planar, then you will have to just assume that some lines will cross themselves and use some kind of heuristic to minimize the number of crossings.

With all this in mind, drawing it by hand will probably be simpler. There are several tools that will draw the graph automatically for you.

Edit: The "drawing a planar graph" problem is a subset of the more general Graph Drawing discipline.

Edit2: I've just found about graphdracula. It will probably be simpler to translate than Boost or OGDF, since javascript is more similar to Lua than C++.