ZOOMIE1980
Posts: 1284
Joined: 4/9/2004 Status: offline
|
quote:
ORIGINAL: Stavka_lite This has been a very informative thread and I have learned a great deal from it so thanks to all of the posters. It is good ( to know that other people have the same issue and that there are ways around it. The Death Spiral or the Naval marshalling ' undocumented design features don't make this game unplayable for me because I use it as a "sparring partner" to learn how to do things and how not to do things. I originally posted because I was amazed at this bug. I do understand the complexities of AI design and have no desire to attempt it. I have no desire to pay hundreds of dollars for a blue sky game AI that Zoomie wants, I always have shifted to PBEM once I know the rules and how to use the interface. No AI short of the Blue series can handle a game like this with all of the variables and I am happy with this game and its attendant AI. First off, I don't want a $300 "Deep Blue" game, either. What I would love to see would be a good object oriented game engine API coupled with some powerful disk based data handling tools, and then use those tools to create some games with some more enhanced AI features, leaving the presentation layer to professionals, skilled in that area. Problem today, is there are NO decent API's or OpenSource projects out there at all. The few "game design kits" are totally lame, building block style, scripting tools for non-programmers to create simple minded arcade games. Just like before the C++ ISO standard came along complete with the STL, programmers had things like Rogue Wave, to help the low level stuff. We had MFC, wxWindows, IOCL, and Fox for GUI stuff. No need to code your own container classes or dialog boxes, frame apps, etc, because someone else with a lot more skill, money and time did it for you! The lack of robust toolkits leads far too many game developers to have to spend far too much time and resources on reinventing wheel (like UV and WiTP rolling their own UI stuff) instead of investing more heavily in game balance, AI and such.... In otherwords, far too much time is spent writing low-level, infrastructure code, and too little writing actual application level code. Thus a big part of the reason we see the general poor state of the AI in games, overall. The AI in this game is actually pretty good, as wargame AI's go. But it is obvious it has a few remaining, significant problems left that keep it from being a true "advancement of the art", so to speak. As Gnome states, it has the feel of being about 90% there. I can get around that last 10% through personal "house rules", but I wish I didn't have to. Especially since that last 10%, conceptually, should not be that difficult to overcome.
|