Hello GGS

[Azail]

Hello everyone, I'll make this as short as possible.

I've been reading up on GGS, I like what you are trying to accomplish here and I'd like to help make it a reality.

My coding abilities:
C++ (4 yrs., before that VB, before that BASIC )
DirectX/OpenGL (not my specialty )
AI coding ( ...ahh my specialty)

Things done wrong so far (IMHO)
1. - developing MP game before SP
There's a reason why Sid and the boys developed the SP game first: it allows for a much faster developing cycle. There's no need to log on to test something out, for example. I could list more but I'll leave it as is for you to think about and I'll discuss it more if someone wants to.

2. - There is no clear origin point of development
Every game project I've worked on (and completed) started at the beginning, and I think you should too. For example, think of Civ... the first thing you see is an intro movie
then a logo splash
then the main menu(s)
then you see the map and unit(s)
In other words, develop the game in the order you expect to see it in the final version. This will also speed up development of the game and allow you to reach a first demo rather quickly (even if all you can do is move a single unit in that first demo).

Thanks for reading & I hope to be coding some GGS soon.

[Leland]

I'd like to add...

3. - We drive away new potential programmers with pitchforks and torches.

In other words, we are quite horrible at organizing our programming efforts. We're lazy, unproductive and probably a majority of your questions regarding programming can be answered with "nah, we haven't done that" or "I'm busy". I am not sure if this has become obvious to you quite yet, but I hope that you will not be scared away by our utter lack of motivation, and have a little bit more determination than the rest of us.

Anyway, I disagree with the two problems you perceive.

SP vs. MP: I don't think this is an error as much as it is a design choice. Personally, I am not interested in a game written around restrictions of AIs, nor do I think that starting up a hotseat game to test stuff that's being developed would slow us down at all. Single player mode on the other hand makes it pretty much impossible to test stuff like diplomacy or game balance because you have to write rather complicated AIs before you can sit down and play.

Point of origin: Actually, we do have a more or less a clear image of where to start. It's been discussed in chats and the mailing list (which you might want to join), and I think we are in agreement that what needs to be done right now is some sort of mock map and a chat program (part of the diplomatic interface). The real problem is, quite bluntly, that nobody bothers to get up their asses and do something about it.

[Azail]

Originally posted by Leland
I'd like to add...

3. - We drive away new potential programmers with pitchforks and torches.

In other words, we are quite horrible at organizing our programming efforts. We're lazy, unproductive and probably a majority of your questions regarding programming can be answered with "nah, we haven't done that" or "I'm busy". I am not sure if this has become obvious to you quite yet, but I hope that you will not be scared away by our utter lack of motivation, and have a little bit more determination than the rest of us.

I was thinking of joining several months back, when the only thing that seemed to be happening here was a game of Diplomacy. That's what kept me from joining before. Then there was a sudden upheaval of activity and I decided the time was right.

Originally posted by Leland
SP vs. MP: I don't think this is an error as much as it is a design choice. Personally, I am not interested in a game written around restrictions of AIs, nor do I think that starting up a hotseat game to test stuff that's being developed would slow us down at all. Single player mode on the other hand makes it pretty much impossible to test stuff like diplomacy or game balance because you have to write rather complicated AIs before you can sit down and play.

You're right, it is a design choice, an error in design choice IMO. MP games will have to be tested with several people, some of which are good strategists and some are bad, and this is but a single test (game). It's totally unbalanced each and every time due to the fact that you can't be sure you'll play the same people every time. SP on the other hand, allows you to run multiple tests (several people each play their own game) and it's always against the same opponent (the AI).

Diplomacy and game balance should actually be some of the last things you test. Diplomacy requires so many things be finished before it can be an accurate test (military and economy at the very least, but IMO you really need nearly the entire game done). Game Balance is the last thing to test... simply put, you can't balance a game until it's finished.

As for writing complicated AI's... this is simply not true. In a few hours I could have a good FSM written which would perform well enough for testing purposes. Half of the time (if not more) spent developing the AI would be thinking of the different possibilites to add to it's list of choices (and I've pretty much got those figured out already, from years of playing Civ).

Originally posted by Leland
Point of origin: Actually, we do have a more or less a clear image of where to start. It's been discussed in chats and the mailing list (which you might want to join), and I think we are in agreement that what needs to be done right now is some sort of mock map and a chat program (part of the diplomatic interface). The real problem is, quite bluntly, that nobody bothers to get up their asses and do something about it.

I still think the beginning is the best place to start, but at least you do have a starting point. From what I saw on the website the map seems to be fairly well on it's way, though I don't know how old the text is.

[Leland]

Originally posted by Azail

You're right, it is a design choice, an error in design choice IMO. MP games will have to be tested with several people, some of which are good strategists and some are bad, and this is but a single test (game). It's totally unbalanced each and every time due to the fact that you can't be sure you'll play the same people every time. SP on the other hand, allows you to run multiple tests (several people each play their own game) and it's always against the same opponent (the AI).

I think the ability to have both good and bad strategists in a single game without screwing everything would be a feature, not a bug. And it seems to allow much greater diversity than an AI opponent. In any case, what do you think would be the problem with the multiplayer first approach, isn't it possible to develop AI opponents simultaneously?

Diplomacy and game balance should actually be some of the last things you test. Diplomacy requires so many things be finished before it can be an accurate test (military and economy at the very least, but IMO you really need nearly the entire game done). Game Balance is the last thing to test... simply put, you can't balance a game until it's finished.

Ah, I mistakenly assumed that these were the kind of things that you thought would make the development cycle shorter. But if you were talking about simpler testing like seeing just how some particular features work out, isn't it equally simple to kick off a 2-player hotseat game than to start a single player game? Where else do you need the AI if not for game balance (after all, the game would be rather boring without opponents...)?

I still think the beginning is the best place to start, but at least you do have a starting point. From what I saw on the website the map seems to be fairly well on it's way, though I don't know how old the text is.

I wouldn't trust anything on our website. I do believe that unless someone has done something during the last week or so (I've been out of the loop myself), the map is nowhere near "on it's way".

[Azail]

Ah, I mistakenly assumed that these were the kind of things that you thought would make the development cycle shorter. But if you were talking about simpler testing like seeing just how some particular features work out, isn't it equally simple to kick off a 2-player hotseat game than to start a single player game? Where else do you need the AI if not for game balance (after all, the game would be rather boring without opponents...)?

When I think of game balance I think of making sure the numbers are right, for example, making sure all the unit attack and defense values are right.

Sorry I must have missed it before... "hotseat"... I was under the impression you were shooting for internet play first. If hotseat is first on the list, I see no problem with it then.

[S. Kroeze]

Welcome Azail!

One question: Have you ever played any computer game with satisfactory AI?

Even a 'bad strategist' is a genius compared with AI in my opinion!