Can we help the devs?

[Ivan Pedroso]

There are quite a lot of discussions about balance and wishes for Dominions 3. Some are very productive and informative and others aren�t. But what would the devs actually like us to discuss and �work� on? What can we do that�ll be genuinely productive and helpful?

Names:
I like the thread about names for the standard themes. Are these sort of things anything the devs can use? Are there other areas where a naming brainstorm might be useful? National heroes? Spells? etc..

Units:
It would be cool to expand the list of mercenaries. We should all be able to come up with one or two cool mercenary Groups � would that be helpful or a waste of time?

-Pete

[Zooko]

I want to be able to write code in my favorite programming language which drives strategic AI decisions (i.e., for non-human computer opponents) and also I want to be able to write code that drives battle AI decisions. (I.e., which decides which spells to cast for both my mages and enemy mages and cetera.)

The first would be possible if Illwinter would simply open up and document the turn files, I guess. The second would require more work. For example, make a "turn file" like file for every battle turn and let me write that, and then have Dom2 read that turn file in order to decide what every unit on the battlefield does that turn...

[Gandalf Parker]

Putting it in is one thing. Putting it in so it cant be abused is another. For a PbEM game (which this primarily is) you would have to send such code to the host. That would make me nervous.

And back to the original question...
the devs have shown interest in a number of things. You can search for Kristoffer and JohanK to see.

I remember they were interested in people who are actually willing to TEST the things they kept saying about the AI's. A number of plans were developed for how the AI should decide to build castles or decide to build troops. I think one person finally decided to try and prove the troop building (play a nation as though an AI doing only the decisions in the formula) but Im not sure how far they got

[Zooko]

You mean for example, I could write a mod that runs arbitrary code in order to make its decisions, and you could run a game server, and I could send you the mod to run on your game server, and when you run it, my mod could turn out to be a Trojan Horse which steals your bank account numbers and sends hate mail to all of your friends?

I agree that this is a risk. Such risk can be mitigated in three ways:

1. Arrange to have limits on what the code can do. Ideally the only thing it can do is (a) read the turn description file, (b) use some CPU and RAM for a while, and (c) write out a turn move file. It really oughtn't be able to send mail to your friends.

2. Don't run mods from people you don't trust.

3. Insist on mods being public -- ideally open source -- so that any bad Trojan behavior is likely to be detected by people reading the source.

[Ivan Pedroso]

Gandalf Parker said:
And back to the original question...
the devs have shown interest in a number of things. You can search for Kristoffer and JohanK to see.

(Ohhhhh I've done that - been a lurker since February.)
But now that we have Dominions 3 on the horizon, it would be interesting to hear (from the devs themselves?!?) what we could help with. There are a lot of creative and talented people in the community, and who knows, someone might get a brilliant idea that perfectly captures an aspect the devs were looking fore. And I would think that everyone would be SO proud if their idea made it to the final Version.

I remember they were interested in people who are actually willing to TEST the things they kept saying about the AI's. A number of plans were developed for how the AI should decide to build castles or decide to build troops. I think one person finally decided to try and prove the troop building (play a nation as though an AI doing only the decisions in the formula) but I'm not sure how far they got

Yeah - the thread on the troop/castle-building-algorithm (thread) is an example of an idea that is backed up by a clean and scientific test.

I would like to state (perhaps more clearly than previous): This is not an "I wish this", "why don't the devs do that"... thread.
It is a reversed wish-thread: From the devs to us.

(hehehehe, if I was working on this awesome game, I wouldn't want any help - I would enjoy it too much to share But on the other hand: We do represent a resource that could be utilized.)

[Ivan Pedroso]

By the way: If this ball ever gets rolling and we generate some spin-off threads. I'll take on the responsibility of maintaining those. That is: The first post will include a clear and updated list of the best ideas in the thread. "The best" will be those ideas that (in my opinion) are well thought-out, and/or ideas that get applause from serious forumers.

[Kristoffer O]

Hmm, a difficult question. It's not like we can ask you to do coding or graphics.

Actually, what I would like to see is more mods, maps and scenarios.

I wouldn't mind playing a lord of the rings scenario. Chris Byler started one such project once IIRC, but it never was finished. We would not want to make a LotR mod ourselves as it could cause copyright problems, but I think players would like it. Much like there is a Star Trek mod in SEIV.

[Alneyan]

Kristoffer O said:
Hmm, a difficult question. It's not like we can ask you to do coding or graphics.

Why not? That looks like a marvellous concept; you get paid, you have your benevolent players do most of the work, and you can concentrate on design while leaving such mere concerns like "length needed to implement that great idea, and the difficulties thereof" to said players. Sounds like the perfect deal!

Ah, I see the problem. For the idea to work, you would have to trick the players into doing your bidding unbeknownst to them. So revealing the whole scheme might not have been such a good idea. But then, who knows with all the conspiracies going on. Hmm... *Stops digressing, and being silly*