FreeOrion

This is my first post here and first I'd like to express my deep appreciation for somebody embarking on a project like this. It's been in my mind since I began to understand what a disappointment MOO III was to become.
Actually, I started having misgivings as early as it was announced that the game was to center on so called "macromanagement", several months before release. Then I lurked quite a lot on the MOO III boards and saw how the situation deteriorated sharply after release.

Personally, I have only played MOO II, but it was a game that I absolutely loved playing. I understand that this effort will not use the same "races" or storyline etc because of copyright issues. Allright, I guess I'll have to live with that.

Anyway, ad res! What I cannot understand is why it has been decided to use so called "starlanes"? If I understand this concept correctly, starlanes are sort of roads between star systems that you will have to stay on in order to travel from system A to system B.
Provided I'm right I must protest against this idea most profoundly.

1) The concept apparently serves no game or strategic function. The idea of funneling avenues of approach towards certain points do not deepen or contribute to the strategic challenge of interstellar maneuver, on the contrary, it simplifies it.
If the enemy in principle can strike anywhere on the other hand, you'll have to make decisions on what is vital defending and forces you to try and control the "inner lines" (to speak clausewitzian) in order to coordinate and decide where to counterattack. Now that is strategy!

2) Starlanes is an entirely superficial concept. Such things simply do not exist in physical reality and will only weigh heavily on your readiness to 'believe' the game.
I think that space shall be open for all to go anywhere, only restricted by your level of civilization and technology.

What was wrong with the way interstellar transit was conducted in MOO II?

Please, I beg you to reconsider this!


Respectfully,
Gunsan

#1) I moved this thread to the brainstorming forum where it belongs.

#2) I cannot take the time to dispute your entire argument, except to say that it has already been done. Unfortunatley, most of it was on the old forums; you can see a design thread on starlanes, I think, in the 'game design' archive board, but even that is not the whole discussion.

The feature has already been passed. Someone may be able to respond entirely to your argument if they have the time, but I cannot right now. :-/

What I can say is that your second argument, the 'do not exist in physical reality' point, has no bearing on any decision we make. What things are like in Real Life affects us little or not at all; this is a game design and AI decision. The majority of the discussion, if I recall, was centered on the question of strategic bottlenecks (e.g. if you have no starlanes and can always fly anywhere to the limit of your fuel tanks, then one planet is just as vulnerable as the next; your 'front line' is not a front line, but a hodgepodge) and you cannot really draw a heterogenous star map using this method. With star lanes, it's much more of a map; there is defined strategy; the AI can more easily defend important points and has metrics to figure out what an important point is (e.g. a star system with many habitable worlds with many star lanes that lead to enemy systems).

That's just a bit of what we talked about.

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!

Okay, thanks. We could probably sit here all night discussing frontlines in space, which of course does not exist, just as they do not exist in naval warfare, but what would be the point? As you say, it's obviously a passed issue and you haven't got the time. I'm happy to see there's at least one rational argument behind this starlane thing though (computer AI).

Hmm, maybe this isn't the game for me after all? Just wondering, if this place exist, could there be any other projects trying to do the same thing, that is creating a Master of Orion IV some other place on the net?
Would be interesting to check them out.

I highly reccommend that you read the design document before passing judgement on anything. I hate to say 'I don't have time for you,' but put yourself in my place: if I stopped and re-argued every single thing we've argued already every time a new person showed up and question our decision, we would never get anything done.

-Aquitaine

_________________
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!

You should really read the thread which discusses starlanes.

As I recall another major proponent was that they created a sort of terrain, and strategic choke points.

_________________
Time flies like the wind, fruit flies like bananas.

Don't worry Gunsan, I hear you. I was very much against starlanes and its mechanisms, but some starlanes will add to strategy. There are some advantages to having starlanes. In addition, you will get the option to mod the game or to turn starlanes off if you really wanted. Right now, I think we will have few short starlanes interconnecting systems with few starlanes in the connected system as default; most systems are not connected.

_________________

From the v0.2 design doc:

• Starlanes: A UI element will be provided for the following user selections: frequency of long starlanes (default few); frequency of medium starlanes (default some); frequency of short starlanes (default several); the choices will be none, few, some, average, many, and very many. The UI will not permit the user to select ‘none’ for all types of starlanes. A composite value that is an average of the aggregate values will be displayed as a summary descriptor (e.g. ‘mostly average starlanes,’ ‘all sorts of starlanes,’ ‘a mixed amount of short and long starlanes,’ ‘Nightfish starlanes’)

_________________
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!

Addendum to the last:

Obviously, the game could be modded to support no starlanes; however, this involves a significant degree of AI programming, and the 'official' game engine will we written with starlanes in mind. But anyone is free to change it, as always!

-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!

Gunsan, dont forget that starlanes are just the 'express routes'. You will still be able to go off-road - it will just take you longer (or take bigger engines or somesuch) to get there.

- Good!



- So if I understand you correctly, there will be an option to mod the game although the parameters for the AI will be set, since of course, modding, at least in my experience, and programming AI are very different things. Personally, I'd even accept a half-wit AI rather than starlanes, but each to his own.

Looked through a few pages of the past thread on starlanes. Indeed there seems to have been a fair amount of controversy over the issue in the past. I respect the decision to build the game around starlanes of course, but let me just point out that features that repeatedly attract controversy also tend to influence the reception of the finished product.

Another thing about starlanes. A dimension apart from conquest in MOO II, although underplayed, I feel was exploration. Remember how you started up an entirely new game with a universe full of beckoning pretty stars and mystery? You could estimate what systems were best to settle but there was no way to be sure. Virgin territory. Strange inexplicable things lurked out there and even reaching the nearest stars was a gigantic undertaking demanding every effort of your fledging spacefaring civilization. The GCN symbolizing rumours and scientific hints filtering through about intelligent life, far away.
With starlanes, however, you'll get a universe already criss-crossed by interstellar highways. There is nothing to do but stay on the freeway till you reach the next interstellar McDonalds drive-through. Nothing to see here, move along. See what I'm driving at?
Apart from being platantly absurd, starlanes risk creating a constricted feel of a place where the scene is set and everything has been done.

Of course, I'm digressing a bit since after all, it has been decided that the starlane feature (among other things) should be moddable, which I think is an encouraging and magnanimous resolve, catering to all tastes. Just hope you'll keep it that way, or better yet, why not make it a game option to simply turn it off at game start?

It might hamper the AI, but provided it won't make the game break I can accept that. I'm not demanding any hefty amount of work put into it just to satisfy certain personal whims.

- Nice to know there are other starlane sceptics on the forum.
Yes, I agree that starlanes will add to strategy in some ways. Like if you were fighting a ground war in mountaineous areas where control of certain mountain passes is paramount.
However, space is the cosmic ocean. A very different setting. There was some talk about frontlines in previous posts. Let me just say thus: a frontline is an abstract concept. In a way there are frontlines in a space, but they are not defined by trenches and solid obstacles but by technological capabilities, logistics and the risk of enemy countermoves. Generally, you need a string of systems to support and supply your spacefleet for example and conversively, relatively sheltered base areas exist inside an empire.
It's just that I don't see the reason for starlanes to define what's behind the frontline. This should be regulated by other factors in the game.



- If you feel like it, please share your opinions! I'm quite dead against it but there could be things I've overlooked, naturally.



- Indeed, this is the good thing.

Realism argument: "Space is the cosmic ocean, in which we can't build boats, can't swim, would take hundreds of years to get to another bit of land if we could and die within 30 seconds if we fall in."

Let's not do realism arguments. They just aren't fun.

Gameplay argument: you seem to be saying that you'd rather impose restricted movement by some kind of range limitation or supply and logistics requirement. Lot's of talking about this kind of idea has gone on in the forum already, the upshot of which is that no one has yet managed to come up with a workable model that isn't more horrible than starlanes, which is why even those who aren't that keen on starlanes have given up and accepted that's the way we're going.

After all, nobody complains about all those unrealistic "starlanes" between the squares in Civ or SMAC (I believe they call them "edges" in those games).

And of course, this is an open source project, if the starlanes thing turns out not to work, we can always change it, or you can, but for the moment, let's run with it.

The AI argument:

With no starlanes, the only limitation is range, like it is in MOO2. Yes, the concept of frontlines is artifical, but it is essential for the AI; after all, the player knows to garrison the most important worlds. The player knows which worlds are threats. Even without starlanes, the player can guess where the enemy might strike, but it essentially lowers or even removes the heterogenous nature of the map; it makes all systems equally vulnerable in the late-game.

In this situation, it is very difficult to have an AI that can come close to matching a human player. It is difficult to teach an AI what an 'important' star system is outside of the value of the worlds in it. With starlanes, though, you actually have geography, in a sense; there may only be three 'entrances' to your Empire, and so the AI sees this as a border, as a front line; it can also plan an actual attack strategy and say things like 'I am going to attack his mining world, and to do that I must get past points A and B, so I will garrison point B with a small force and move on to point A through starlane X.'

Much of what I know about the challenges of AI programming does not come from my being an AI programmer (I'm not) but from watching the development of EU2, HoI, and Victoria, all of which have enormous challenges with AI. The distinctions they use tend to be along the lines of 'the eastern front' or 'guard the northern coast of France.' The more we can distinguish the map, the more the AI can work with it.

This was actually not my reason for initially supporting starlanes; it's one of the few things I like about MOO3 even as a player, but as I learn more about what our AI challenges will be, it helps me keep in mind that the computer has to play the same game as we do.

_________________
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!

Why is it so difficult to tell the AI which worlds are important? Especially if so much of the game is numeric, it should be "relatively" easy. Let's say a world is Medium, Rich and Toxic (using MOOII standards) and another is Huge, Poor, Terran. Each one of those becomes a value that you use math to calculate. Add to that number something for buildings and structures and you've got a nice figure. I'm not sure if Leaders are involved, but a leader adds value to a planet as well.

Just not sure on that we can't teach them to value planets without starlanes.

Because a poor toxic and alltogether hellish world might be on such an important strategic position, that the AI has to defend it.