Basic questions on the FO project

For topics that do not fit in another sub-forum.

Moderator: Oberlus

Post Reply
Message
Author
SgtCycle
Krill Swarm
Posts: 11
Joined: Wed Oct 22, 2003 8:02 pm
Location: Montreal

Basic questions on the FO project

#1 Post by SgtCycle »

Hi guys,

I’m new to these forums, and spent a couple of hours reading at some of your design documents and other posts. In all, there’s a lot going on, but I have the following questions to ask.

What’s special about the free orion project? It may sound silly, but I think it’s a valid question; what will this game have to offer once it’s complete? To use an IT phrasing, “what’s the vision”?

In your opinion, what will make your game fun? Perhaps another silly question, but one well worth asking, and one the designers of MoO3 should have kept in mind. They designed this very complex game, but there is no fun factor at all in it.

2 years ago, I asked that exact question to the MoO3 designers when they announced that space combat was going to be real time, “is it going to be fun?”. They said “don’t worry, it’ll work”. Yeah…right, we all know what happened. All space combat s*cked big time. Especially all the monbo jumbo with picket ring, escort ring and core…didn’t change a thing, your fleet either vaporized the other one in a volley, or yours blew up after one fire exchange

About the AI, I know this is the very beginning of the project, and very little detail is available so far, but how are you going to build/modelise it? Any of you guys ever did this before? I read some posts about not scripting it and all, which is all good, but game AI don’t think, they react to the perceived game environment. Meaning, you define parameters on how to evaluate what’s going on in the game every turn, and the AI then weights those parameters, based on its short term and long term goals, and then makes a decision. In the end, it all comes down to the decision model (the matrix) the AI is going to follow in the game.

That’s it for now, I’m looking forward to read your answers

Keep up the good work !

PowerCrazy
Creative Contributor
Posts: 383
Joined: Fri Jun 27, 2003 2:35 am
Location: Texas

#2 Post by PowerCrazy »

Aquitaine or someone will prolly add some details, BUT i'm here to tell you that we are making FO because we want to. There is no incentive other than the game itself. As such the reward of our work will be directly proportional to how fun the game is. Thus if the game sucks its because we made it suck, if its good its because we made it good. So obviously the incentive is to make the game fun, and that is why it will be fun.

What makes FO special is the fact that it is the MoO communities response to the horrible failure that was MoO3. And i can generally say that everyone is here because MoO3 did not live up to their expectations, and they hope to manifest their Ideal game in FO. Obviously though everyone can't have everything they want. So that is where our leaders come in. Most of the moderators Tyreth, Aquitaine, Nightfish and original members me, drek, etc have been pretty good at getting things done and agreeing on what our game is going to be. I see that as unique in itself, especially on the battle-field known as the internet.

Our vision can be found i believe on the website http://www.freeorion.org and basically FO will become the best of what everyone on the team has to offer.
Aquitaine is my Hero.... ;)

jbarcz1
Creative Contributor
Posts: 226
Joined: Thu Jun 26, 2003 4:33 pm
Location: Baltimore, MD

#3 Post by jbarcz1 »

One thing to keep in mind is that our definition of 'fun' may not agree with everybody else's We had an epidemic of posters in the past who thought that unless we made the game exactly like MOO2 but with better eye candy, that our game was doomed to suck because, to paraphrase: "MOO2 is a fun game and if you dont do what they did your game can't be fun".

Our goal here is to make a game that we, ourselves would enjoy playing more than we enjoy MOO3, and we're hoping that others will find it enjoyable as well.
Empire Team Lead

SgtCycle
Krill Swarm
Posts: 11
Joined: Wed Oct 22, 2003 8:02 pm
Location: Montreal

#4 Post by SgtCycle »

Absolutely ! I agree that most people have a different definition of what "fun" is in a game. Which brings me back to my original question, so far, what do you guys think will make the free orion game fun? A good AI? A cool tech-tree? Nice graphics?

You see, simple questions are actually harder to answer than complex one. People (especially professionals) tend to jump in complexity and details thinking that all of that will make the gaming experience "rich", thus fun. Yes, rich because there is so much details, but fun? That's a completely different concept/idea. Is Moo3 rich and complex? Yes. Is it fun? No.

Like a friend of mine said one day about most modern games, what do you do in the game: are you spending your time (trying to) managing it, or are you actually developping strategy and testing it in the (game) field?

Other example: in my opinion, most RTS are not fun, because your time is not spent thinking ahead, deploying, attacking, analysing, etc... It's spent clicking like a maniac, selecting a group of units, launching them forward, forgetting about them for a while and do something else. Hey, that might be someone definition of "strategy" or planning or whatever, but to me, that's not fun. That's stressful, and I have enough of that in a work day ;)

All this details and complexity analysis/criticicism lead me to this: the more complex the game, the harder it will be to actually build a decent AI. Does it mean that it's not worth it? Maybe, maybe not. One thing for sure, you should keep this in mind. When you implement a feature in a game, it might make it somehow more interesting/challenging/rich/fun to the human player, but if it's unmanageable by an AI, you're shooting yourself in the feet: the AI will s*ck.

In one of the post, someone had a good example with ship design: a human will build a fleet, send it in combat, see how it fares against the enemy fleet, and then adapt his design based on the result of the combat. If you win the fight easily, your design works. If they all die in the first turn, something is very wrong. This kind of analysis/learning experience in a game AI is VERY difficult to build if there are 36500 ways to configure a ship design. Good luck putting that into a matrix for your AI. On the other hand, If your fleet has a combined firepower of 200 in attack, 175 in defense, and your enemy has 400 in defense, you know you're attacking with 1-2 odds, which is not good. Now, that's easy to put in a matrix: don't attack with less than 3-2 odds or something.

My 2 cents for now, looking forward to read your feedback

User avatar
skdiw
Creative Contributor
Posts: 643
Joined: Mon Sep 01, 2003 2:17 am

#5 Post by skdiw »

1. Graphics
FO has too good of graphics imo. Check out the galaxy map. I say "too good" because our strategy of such high quality first run through is slowing us to a crawl. everything else has to be the same high quality or it would be weird.

2. Tech Tree
As far as I know, we only passed the orgainization of tech tree (passed mean it will be in the game. Our design process is brainstorm -> design -> features -> if passed then it will be in the game.). We are going to group them under theory, applied, and refined. As for the details, it's really still too early as we don't even have the economy and other fundamental game mechanics passed. Tech is really for v.3. We are still trying to finish v.1.

3. AI
I try to make everything very simple. There, of course, many ppl here who forgets and over-design the game, imo. Some ideas are really cool, but I frequently have to remind them to take a step back and look at how many other things you have to manage. So far, imo, FO is a repeating the design problems that moo3 faced. We had a huge discussion under brainstorm: micro vs macro thread. We have several ideas about keeping the number of clicks to a minimum so the game will be fun. Currently our main solution other than a good UI is keep things very simple so there is only one optimum way that everyone will go with. For example, I am pushing for a very simple tech upgrade game mechanic that once you research a tech to upgrade stuff, you will always want to upgrade without exceptions so you don't have to open 2 windows on every planet and select off a long list every turn. Feel free to post your anti-micro ideas on that thread. Keeping thing very simple will make the game progress along faster so FO won't be outdated when it's release (or ever will be) and makes easy AI's. We have some ideas to make MP w/ extra option for human players to dig around.
:mrgreen:

Ablaze
Creative Contributor
Posts: 314
Joined: Tue Aug 26, 2003 6:10 pm
Location: Amidst the Inferno.

#6 Post by Ablaze »

I think you are preaching to the choir Sgt. I think skidw is referring to the odd newbie who pops up and purposes that we implement some feature like each empire having a variable exchange rate depending on the strength of their economy. Most of the people around here are (IMO) focused on streamlining the UI above all else, and then on adding new features.

Personally, I look at the rather large fan base of Galactic Civilizations and am forced to admit: Not only is the simpler way easier, but people like it.
Time flies like the wind, fruit flies like bananas.

Daveybaby
Small Juggernaut
Posts: 724
Joined: Mon Sep 01, 2003 11:07 am
Location: Hastings, UK

#7 Post by Daveybaby »

IMO the most important part of the fun of FO will derive from the fact that its open source.

So if you dont like the game for reason X or Y then just change the thing.

If FO is even half successful it will probably spawn a whole slew of mods, ranging from minor tweaks to unrecognisable reworks. Look at the amount of effort people are willing to put in to mod moo3 when they cant even fix the core engine of the game.

@SgtCycle: w.r.t. learning AI, my point above is especially relevant. It probably unrealistic to expect to develop an AI which can learn to use the best ship designs - but if we as humans learn how to do it while playing the game, we can then go back and reprogram the AI to take advantage of what we've learned.

One of the problems with commercial games like moo3 is that the AI is finished before the game as a whole is. There just isnt the opportunity to iteratively develop the AI - especially with the stranglehold that publishers now have on code releases. Open source may well solve that problem.

Aquitaine
Lead Designer Emeritus
Posts: 761
Joined: Thu Jun 26, 2003 1:54 pm
Location: Austin, TX

#8 Post by Aquitaine »

Part of me wants to write a nice long post about why this is a great project and we'll make a good game.

The rational part of me says that if I do this every single time I'm challenged to, I will never get anything done.

Our vision and details about our project are posted on the announcements board and on our web site. There is nothing I can tell you that you can't learn from reading the boards. If you have a specific question about a specific system, I can tell you either what we've passed or what my thoughts are for the future -- but I'm even a little leery about that because if Tyreth or I say 'I think space combat should involve lemmings' then it will touch of a 300-post discussion on why that's the <best/worst> idea since <MOO1/MOO2/MOO3> and I should be <shot/elected galactic senate president> for suggesting it.

So if I'm close-mouthed about the future it's because I try to keep our team focused on our immediate goals; that is why we've provided a vision, because you should be able to get a sense of where we want to go without me having to say things like 'I think the Total War series can offer us a lot of ideas for space combat and that the Europa Universalis series can offer us a lot of ideas for diplomacy and Empire AI.'

So, as you see, I don't have to say things like that. :wink:

Aq
Surprise and Terror! I am greeted by the smooth and hostile face of our old enemy, the Hootmans! No... the Huge-glands, no, I remember, the Hunams!

tzlaine
Programming Lead Emeritus
Posts: 1092
Joined: Thu Jun 26, 2003 1:33 pm

#9 Post by tzlaine »

I agree with Daveybaby completely. The strength of the game is in its open-source nature. We have as much time as we like to get this project finished. Furthermore, we are not motivated by profit. The point is not to get something out the door to recoup our expenses, but to make the game we want to play. This means we won't cut corners, unless out of laziness.

This comes back to the earlier questions about what will make this game fun. Fun, art, and pornography are all extremely hard to define and trivially easy to recognize. I have a only general feeling for what it is that I like so much about this genre, but once we have a palyable game, I have faith that I will be able to spot what works and what doesn't.

@skdiw
As for the graphics slowing us to a crawl, you've got the wrong idea. There have been maybe 3 or 4 programmer-days total spent putting together the simple graphics we have now. The thing that is taking all the time is laying the foundations for the code base we will use for the rest of the project. Development is always slowest at the beginning.

SgtCycle
Krill Swarm
Posts: 11
Joined: Wed Oct 22, 2003 8:02 pm
Location: Montreal

#10 Post by SgtCycle »

Good posts guys, you're all right on many points.

Not being motivated by profit and time constraints does help to make things (here a game) that can be more enjoyable, but it does take longer, usually ;)

Being open source is also a great idea, and many games are incorporating that feature. But you have to be carefull about what you imply by open source. Obviously, the way the UI works is obviously part of the "black box"; can't be changed unless a lot of code is rewritten and recompiled. Ships statistics however are a good subject to be open source: defined in text files.

IMO though, not enough game try to use AI driven by text file. I could be wrong however, since I'm not a modder. I've never seen mods, or very few, that actually made the AI opponents play better. Usually, patches fix the AI. The only game that comes to mind is Hearts of Iron, where units have a weight, and the AI builds its unit based on their relative weight. In a way, that's the matrix I was referring to in my earlier posts.

A piece of advice on your project: the more time you spend on the initial planning and design, the less code you'll have to write, rewrite and rewrite and rewrite...

I know, I know, you probably heard that a gazillion time, but it's true. I'm an IT professional, and projects that fails, meaning overlap in cost and time, are usually due to incomplete requirements.

You should always define a goal. Always. Something like "this game is gonna rock because A, B and C". A simple phrase. Just a concept/idea. And kept it in mind during development. Anytime you're gonna spend time designing and/or coding features, always ask yourself "is this gonna fit in what we're trying to achieve?". If it doesn't, then you probably shouldn't do it.

For instance, like many gamers/programmers, I've had my share of "developing" game on the side. My last project though, I've been writing, writing and rewriting the initial design document over and over. It's about 20 pages now, for the last 3 months. Of course, I only touch this document from time to time, it's a work in progress. And I'm happy I didn't code a thing yet, because I had drastic change during this period. But I'm getting there. Slowly :wink:

In my project, my goal is to make a space wargame, with an open-architecture, that has no ressource management and no buildup phase, and where 95% of your time is spent on assembling fleets (I call them task force) and conduct operations (raid, invasion, probe, attack, etc...) against other fleets and sometimes, planets. That's why I call it a wargame, and not a space strategy or even a 4X. There is no eXploration, some eXpand, some eXploit, but definitively a lot of eXterminate :D

Why am I doing it? Probably for the same reasons as you guys: because it's fun and intellectually challenging. The part I can't wait to work on, as you might have guessed, the AI.

If you guys are interested, I'd be willing to give a hand for that part, or at least give you feedback on your design. It's always good to get feedback from people that have some distance from your work. When it's in your face, sometime you miss the simple things.

Phew, that was a long post, and not surprisingly, it went off topic a little, but hey, it's all good right ?

See...err read you soon !!!

Aquitaine
Lead Designer Emeritus
Posts: 761
Joined: Thu Jun 26, 2003 1:54 pm
Location: Austin, TX

#11 Post by Aquitaine »

Hi again--

Not to single you out, but we get a lot of people who show up, give us advice on what we should do, and then leave. Most of the advice is not even good; yours has the benefit of actually reflecting our process somewhat. :)

Re: AI files, you brought up HoI in another thread; HoI, EU2, and that whole series of games uses text files as the basis for their AI. We have several different pieces of AI -- military AI, empire AI, diplo AI -- so that will hopefully apply for some of it (possibly Empire) and not others, but I don't want to put too many words into the mouth of our AI team.

As for the rest of the advice, the best thing you can do is to join one of the teams and actually take part. Our experience with any project called 'Orion' is that you get a whole lot of people who are absolutely convined that they're experts and that 'if only QSI had listen to me MOO3 would have been better.' We've had a lot of this and we're not even to 0.1 yet, so I anticiapte that it will get quite bad. But since you're ahead of the crowd and good-intentioned, I do recommend you sign up with one of the teams, or if your interest is solely in game design, that you hang out on the game design and brainstorming boards.

The game design board will be cleaned up quite a lot before the release of v0.1 and the v0.2 design doc.

-Aq
Surprise and Terror! I am greeted by the smooth and hostile face of our old enemy, the Hootmans! No... the Huge-glands, no, I remember, the Hunams!

Daveybaby
Small Juggernaut
Posts: 724
Joined: Mon Sep 01, 2003 11:07 am
Location: Hastings, UK

#12 Post by Daveybaby »

I suppose it makes sense to load all relevant constants from a table at runtime, that is something Moo3 got right (most of the time). However its probably not worth going too mad on externalising parameters just to make the game easily moddable, since the source code will be sitting there in front of you waiting for you to change it.

SgtCycle
Krill Swarm
Posts: 11
Joined: Wed Oct 22, 2003 8:02 pm
Location: Montreal

#13 Post by SgtCycle »

Aquitaine:

Thanks for the positive comments.

Simple question then: how do I join? I'd be very interested to participate/help in the AI coding. Who's in charge of that part of the project and what is the general attitude/approach/metod for the collaboration of the coding?

Daveybaby:

Yes and no. Like you posted, you have to set limit to how much can be parametized in text files. But IMO, the more the better. Many people (testers and modders) are willing to change parameters in files, test the game, and give feedback. Very few in actually changing code, compiling, testing, and then provide feedback. In the end, it's probably a balance between the 2.

Looking forward to be more involved in this project
------------------------------------
Things are not as they seem,
nor are they otherwise
------------------------------------

Tyreth
FreeOrion Lead Emeritus
Posts: 885
Joined: Thu Jun 26, 2003 6:23 am
Location: Australia

#14 Post by Tyreth »

SgtCycle wrote:Aquitaine:

Thanks for the positive comments.

Simple question then: how do I join? I'd be very interested to participate/help in the AI coding. Who's in charge of that part of the project and what is the general attitude/approach/metod for the collaboration of the coding?
At the moment tzlaine, who is the lead programmer, is the one to contact about AI coding. There's no proper AI team at the moment for lack of volunteers, so you'd be a welcome asset. Sending tzlaine a pm is the best way.

We use cvs to manage multiple people working on the code at the same time.

Post Reply