But really, it's a great game. Its awesome, actually. If you haven't played it, well, then you should probably just go buy the gold or platinum or other precious metal edition. While its downloading, I'll mention that the shear number of options you have is mind-boggling; this game was made for serious strategy fans by serious strategy fans. A friend of mine says it is "deep". There are two written play-throughs that are simply a joy to read ( here and here ), and give you a taste of what you can do and some of the brilliance of the AI players.
But what a more casual gamer might not realize is that GalCiv II: Twilight of the Arnor (TA), the final expansion, is broken. Okay, this game is pretty 'old' by now, and given some of the stuff I've found to help in poking around in data files, I doubt that it is really news to anyone who has cracked the hood of the game or paid attention to the patches.
I certainly did not realize how far the rabbit hole of broken it went. I got the game freshman year, back when it had no expansions. That game was compelling and very fun, and very, very, very hard (for me, anyway, but that is a different story). I always end up at least pondering modding any PC game I have; its in this programmer's blood. And though I wasn't a very skilled programmer yet, looking at the data files of GalCiv II, I realized that something was wrong. Nowadays, I call that feeling the 'stink' of bad code.
You can tell when code is bad by the data files that support the program and the type of bugs that are discovered. For instance, in Maya3D, you can crash the application by entering a negative number of subdivisions in the Subdivision Surface modifier's dialog box. If you know anything about how Maya was stitched together, you know that the front end is mostly a collection of scripts that get interpreted by the Maya kernel, where all the sexy stuff happens. Thus, this crash is completely avoidable and, in a word, stupid. No one thought to validate the field values before sending it off to the kernel, and since you're converting one type of geometry into another, the kernel eats that negative value and one can only assume it goes straight into the subdivision algorithm code and the whole program crashes somewhere between starting to generate the new geometry and displaying it.
That code smells. It smells like dead animals. You know its stupid and that only unclear thinking went into the system, idea and feature tacked on, one after another, with little control or overall vision. With that in mind, I will present something from GalCiv II completely out of context:
<Weapons ID="Beam Weapon theory">
<Category>Trade</Category>
<Cost>50</Cost>
<Group>None</Group>
<Alignment>None</Alignment>
<WillingnessToTrade>80</WillingnessToTrade>
</Weapons>
This is a snippet from one of the tech tree files in TA, specifically from the Iconian faction. Notice anything odd? Like, a Beam Weapon technology being categorized as Trade?
The categories of techs are used by the AI to aide in prioritizing research; they have nothing to do with the visual presentation of technologies (that's the Group). The tech trees are littered with these sorts of things; if you want all the information on the tech trees and many other tidbits of "WTF?!", you can read this discussion on the GalCivII forums. I had no part if uncovering this quagmire, mind you. Those guys are the detectives, not me.
In short: miss-categorization of techs in TA makes the AIs in the game act like morons. They don't research their unique technologies, they don't get very far, they hit roadblocks before getting to the techs that they need to even contend for dominance. Two of the AI personalities are broken, some tech-trees are basically engineered to thwart attempts by the computer to play well, and to top it all off, the AI is incapable of using spore ships or terror stars. It just won't. And those are big elements of the things TA provides.
It's getting late, so I'm going to turn in for bed. Tolmekian, MarvinTosh, and qrtxian did the work on the techs structure, and next time I hope to get into the badness that is the writing in GalCivII.
P.S. GalCivII, like so many other fantastic, buggy, glitchy, sometimes horridly designed games before it, is still one of my favorites. I love the Stardock guys for doing this game, and I speak so disparagingly of the guts because it warrants lamentation. It is still a great game even with so many problems by the time TA was released. I hope the Stardock guys don't hate me.
No comments:
Post a Comment