Michael Dorosh
der Spieß des Forums
- Joined
- Feb 6, 2004
- Messages
- 15,733
- Reaction score
- 2,765
- Location
- Calgary, AB
- First name
- Michael
- Country
Starting a new thread on this because anyone searching for this topic later would have a hard time finding it in thread called "uh oh"
From Steve:
From Steve:
Thanks for your patience guys! We've had several discussions about this in the past, and two very recently, so I wanted the opportunity to try and keep a couple of issues straight so this thread doesn't go off on tangents and the basic points get lost.
Despite what we all want, in fact it's impractical (read impossible) for terrain to be "spottable" you would need a super computer to determine what to show and what not to show based on any given circumstance. On top of that, the computing resources required to keep three separate "maps" (original, Player 1, and Player 2), even with all the programming tricks in the universe, is also quite large (and let's not talk about when there are 24 players!). On top of all these problems, the code has to be so good and so clean that all of this information can be accessed, manipulated, and displayed with unerring consistency and speed while still leaving computing resources to do everything else. That's why CMx1 doesn't have LOS terrain, nor any other complex 3D game that I can think of. First person shooters don't even need it because you can only see what your character can see, so that is the ultimate form of FoW.
What happens when something, like a trench, is put into a CM map? The underlying terrain mesh is deformed in such a way that a trench is formed both visually and "physically" (i.e. defined geometry). Unlike an object, such as a tank, a trench is literally defined by the terrain mesh, not an externally produced 3D model which sits on top of the terrain. As noted above, since a trench is terrain that means it can't be spotted since terrain can't be shown/hidden depending on variable circumstances. There are no simple work arounds (like 2D "lids") for this.
With me so far?
When we designed the game engine's primary characteristics in 2003 we knew that anything which we made a part of terrain would become terrain and therefore would not be spottable. We discussed the impact of this on fortifications, specifically, since they were the only types of objects affected by this decision. For reasons discussed further on, it was a no brainer that we had to sacrifice SOME forms of spottable fortifications.
As much as we wished we could have our cake and eat it too, tradeoffs are a part of the reality we live in as game developers. For us it's normal to make decisions where we gain various things and lose something else. We have to see the big picture because we have to make it. Gamers, on the other hand, can easily focus on one thing and dismiss all of the others as being of lower value even when they are added together. Gamers live in a bubble of their own making, without any possible way to be proven wrong because... well... because they never produce anything that can be tested. We're not so fortunate
What was gained? A 3D environment with rules which are consistent and reliable. This is definitely the most under-appreciated thing about CMx2 because so much that it does is taken for granted, yet it's not present in CMx1 nor any other wargame. Much of the benefits of the system are invisible to you in direct ways, but you still benefit from it. Bugs, for example, are both easier to find and easier to fix for real instead of making bloody hacks. New features are easier to add because there is a stable environment to add to which mimics the real world and not an abstract one. In fact, the lack of this environment and years of bloody hacks is why the CMx1 code base was chucked out in favor of the new CMx2 engine.
Aside from the coding and gameplay stuff, there is the visual qualities of the 3D environment which must be considered. Why? Because it's what the vast majority of our customers want. Yes, even the CMx1 customers. The number of you guys that could pass a lie-detector test and say that you would rather have the CMx1 graphical environment would amount to such a small number that it's not even worth considering. The number of you that would say you'd rather have the CMx1 graphical environment, but would flunk a lie-detercotr test, is probably a little bit bigger
However, as it so happens, trenches (specifically) are so tied into the physics modeling that Charles would have to make some extremely bloody hacks to have 2D trenches work in a 3D environment. The visuals, therefore, are the easy thing to change... it's the underlying game modeling that is impractical and undesirable to change. It's also extremely damaging to the overall game environment to have such exceptions jammed into it, so it's out of the question.
I'll let this sink in before you read the next bit
OK, so what do we have for defensive objects and how are they handled? We basically have three types:
1. Those that rely upon the terrain to give them definition (both visually and game properties).
2. Those that do not rely upon terrain to give them definition, but modify the terrain if needed.
3. Those that do not rely upon terrain to give them definition and do not modify the terrain.
Starting with the last first, things like barbed wire obstacles, roadblocks, sandbags, and other things which sit on top of the terrain mesh, and don't dig into it ever, can be treated like units. That means they can, theoretically, be spotted/hidden.
Bunkers are similar in that they are separate from the terrain and therefore can be spotted/hidden. However, bunkers are often "cut into" terrain and therefore the terrain has to be modified. Trees can't be growing up through a bunker, for example. When a bunker is put on the side of the hill, having it conform to the contours of the hill would be visually ridiculous, so the hill's terrain mesh is conformed to allow the bunker to remain horizontal. Since terrain, once modified, must be shown that way all the time you get a situation where the bunker can be shown/hidden, but the bits of terrain removed to allow it to be where it is must be shown "as is". This means you can use the camera to hunt around and possibly find where the bunkers are without actually having spotted them. The possibility depends completely on how obvious a bunker modifies the terrain (it might not modify it at all, remember).
And last but not least... things which are defined by the terrain itself. Specifically, trenches and foxholes. As is, these things can't be spotted/hidden, but instead must remain visible at all times once the game starts. The work arounds for this are not practical to achieve, for one or more reasons.
In CM: Normandy we'll add functionality that allows people to place trenches and foxholes using 2D representations (similar to CMx1). The difference is these 2D representations have NO GAME ATTRIBUTES and are, instead, identical to the little 2D icons in the Editor in terms of their functionality. As soon as Setup is exited the 2D icons are translated into 3D modifications and the changes are permanently made to the map used for the game.
It may be possible, and I stress POSSIBLE, to have foxholes retain their 2D representations in the 3D environment and then add hacked in 3D properties for them. Meaning, the foxholes will look pretty much like CMx1 foxholes, but will behave correctly in the 3D environment. This is something Charles will look into at a later date prior to CM: Normandy being complete. No promises, other than we'll give it a shot. I'm hopeful that it will work, and IF it does then foxholes will have at least some degree of FoW.
Trenches, unfortunately, are out of the question. They are too deep and too complicated to even consider making bloody hacks to get them to work in 3D without a 3D representation. Plus, on top of that they would look like crap. Based on feedback over many years from dedicated CMx1 players, and a sense of what the wider audience wants, looks do matter. But as I said, the looks are coming for free with the 3D game engine so this isn't about sacrificing spottable trenches for visual reasons, it's about the impracticality of having significant 3D objects represented in 2D.
And finally...
The realism problems associated with not having spottable trenches is, I think, overblown. Sure, the attacker will always know where trenches are once the game starts, and act accordingly, but I will remind you in CMx1 we had the opposite problem. That was where the attacker was always denied benefiting from intel gathered by previous encounters/attacks. Even if the attacker shouldn't know exactly where the trenches, he most likely would know a) that the fortifications existed, b) roughly where they were, and c) roughly how tough they were. Sure, units bumbled into thick defenses all the time, but usually those engagements were short and the attacker withdrew so that it could come at them again with a plan. Extreme circumstances, such as the Hürtgen Forest battles, this was done over and over and over again in fact.
My point here is to remind you guys that this isn't a one sided thing. In CMBB/AK trenches were unrealistically favorable for the defender to some extent, in CMx2 they are unrealistically favorable for the attacker to some extent. If you're displeased with one because it is unrealistic and has an effect on gameplay, keep in mind that you've somehow managed to live this long even though you've already experienced something similar in reverse Therefore, perhaps the negative effects aren't nearly as game wrecking as some think they are.
On top of this, we have the fact that in WW2 the options for attacking trenches were somewhat limited. Or at least vastly more constrained than they are in CM:SF's modern setting. Artillery was slower, less accurate, and less effective. Look at WW1 for Pete's sake... every inch of the enemy's trenches were known and attacked for MONTHS by heavy artillery and raked with MG and sniper fire... yet the side getting hit more often than not was able to fight off a massed attack against it. Which just goes to show that seeing the enemy's positions is not the same as being able to eliminate them.
It's also true that the more involved the defenses are, the more likely they are to get spotted. Therefore, the degree of effort to mask the defenses was somewhat proportional to the effort involved in creating them. Massive, well hidden fortifications like the Maginot Line, the Atlantic Wall, and even the Siegfried Line are outside of CM's scope and aren't part of the equation. Neither are defenses built in other epochs under similar long term conditions (like Pacific Islands, for example). The majority of defenses in CM: Normandy, therefore, should be of the hasty type that aren't heavily camouflaged. The big exception to this would be the Hürtgen Forest battles which are not within the timeframe of CM: Normandy and are, for now anyway, not relevant. Trenches in Normandy shouldn't even be that common, from a realism standpoint.
A reminder... there was aerial recon in WW2, in case you guys forgot. Not only from specialized aircraft, but also each US division had a dedicated aircraft at its disposal to check things out for themselves. Which means a field with a lot of trenches in it could indeed be known to the ground troops long before they got there. Not as likely as modern days, for sure, but definitely not impossible.
Now, don't get me wrong... I'm not saying that having trenches shown all the time isn't a break with reality. It most definitely is, no argument about it. What I'm saying is that we need to be careful about the Chicken Little effect that is so common about stuff like this. The downsides of the system, as we have it, are often highlighted, taken out of context, and blown out of proportion, while the upsides are downgraded and often cast aside. Others, like the problems CMx1 had with fortifications (like no trenches at all in CMBO) are even forgotten about because they really complicate the arguments that the sky is falling.
Well, I think that about does it. Thanks for being patient