LOS issues resolved in CMN?

[Battlefront.com]

You say it all the time, that i don't understand it.

Again you are confusing two things. I say that you don't understand that you are expressing an opinion and not a fact. This does not make you stupid, so saying that I called you "unintelligent" is not true. You just confuse personal opinion with fact.

I don't like your kind of discussing with strawmans.

Well, I can't see what else to say about this. I pointed out that you are not expressing fact but opinion, then you say that I'm calling you stupid when I did no such thing. If you are not stupid, then there must be some other reason for you confusing fact and opinion.

Your post was about putting into readers mind, that i'm not capable to understand the superiority of CMx2 and that i were just writing about liking or not liking.

You didn't even touch the aspect of fantasy as part of the sucess of CMx1.

Sure, there are lots of "parts" the success of CMx1, including the timing of our release. But the major reasons for success were the 3D environment, real world physics (as best we could), and careful attention to detail. Fantasy was not the major reason the game did so well.

There are those who thinking playing with a wooden train when a child is superior to playing with an electric scale model train set because the wooden train requires more imagination than the exact detail electric train set. I think it's great to have both, but we're interested in making the electric train set type games. Even CMBO was made with this philosophy. The degree of "fantasy" that was in there was due to technological limitations of computers of the day, not because we wanted it to be there.

But how is that possible, if CM:SF sells much better than CMx1?

What do sales have to do with wargame clubs? Wargame clubs are an extremely small number of people with a strong bias towards WW2 gaming. If CM:SF sold 1000 times better than CMBO I'd not be surprised to see less wargame ladder games for CM:SF than CMBO.

No, it is just one aspect that fills the picture.

As another argument you can take the forum itself and compare how the discussions have changed from CMx1.

You mean the arguments over how many mm of armor were on the chin of a Tiger 1E produced on the 3rd day of May 1943 when Hans Grübermeyer was on shift that day? Or what species of wood was used for the German "potato masher" hand grenades? Or did US made C-Rations made available to the British forces make them fight harder or spend more time unsuccessfully trying to evacuate their bowels? Yes indeed, we do not have the sorts of long and extended discussions about historical trivia that we once had. I don't see how that has any relevance to the games we make. It simply confirms that WW2 fans are a very special bread of wargamer.

The forum, even years after release, was full with excitement, of what was happening in the game. Gamers were reporting, as if they were there themselves.

All this has vanished - in quantity and in quality. But i fully admit, that your sales raised.

There's plenty of "I am there" discussions in this Forum. In fact, many have said they feel "more there" than they ever did in a CMx1 game. But if someone doesn't like Modern warfare they won't be playing CM:SF. And if they aren't playing it then they can't possibly be posting about it.

Again a strawman. I fully accept, that you don't share my opinion. My initial post was about nothing else, that you have a complete different vision of where you want to develop the game and i tried to show the differences and where the psychological reasons are. And i'm also not bitter at all, since i'm glad i don't spend time for gaming anymore.

Your position was stated very clearly at the top:

"IMO that is not true and is one of the core mistakes of their development decision for 1:1 representation. I will try to support this with facts:"

You claim to be speaking facts, yet you are speaking opinions. Opinions which others do share, but others do not. Therefore, you can not pronounce the 1:1 change as an error and a fact. It is just your opinion, and others do not share it. Obviously we don't share it.

But i do like to come back from time to time and check, if CM:N is out already, since i will give it definately a try.

Honestly, I do hope you like CM: Normandy. But of course we know we can't please everybody so we are comfortable with you being disappointed by it.

Steve

[Battlefront.com]

Hoolaman,

But as others have said, if your unit has spotted the enemy already, it should be shooting. I've never known trees to prevent LOF on a spotted enemy. At worst, the shell will explode on a branch or something.

In general this is correct. If you have LOF to a target you will shoot at it. If you can see the target from that unit's perspective you should be able to shoot at it. However, there are two conditions which may result in the unit not shooting:

1. If the target is too difficult to hit and the unit is conserving ammo. It tries not to waste ammo on "long odds" shots.

2. If the target is infantry it is very possible that you can see all of the enemy soldiers but can not actually shoot at all of them due to intervening terrain. This is because Spotting is done by Action Spot and LOF is done by polygons.

Where there are no exceptions is with two way spotting, LOS, and LOF between any two Action Spots. If Unit A in Action Spot 10 can spot, draw LOS, or draw LOF to Unit B in Action Spot 20, then theoretically Unit B can spot, draw LOS, or draw LOF to Unit A in Action Spot 10. There are no exceptions to this.

However, differences in the units and their states can influence the chances of spotting or returning fire. For example, if Unit B is a Conscript cowering after taking some fire, it may very well be incapable of spotting or returning fire at Unit A. But in theory it could.

The game system isn't perfect, but when you understand its limitations and the tools you need to allow you to reliably do what you want to do, it turns the frustration level down a lot.

Correct. No game system will ever be perfect. Forgiving problems/limitations is something all gamers must do in order to enjoy the games they play. The question is always if the problems/limitations are worth forgiving.

Steve

[stikkypixie]

The only real quirk I've seen is that when three tanks are lined up (in one line), it's possible to see the last tank before the tank in the middle. Very rare though.

[c3k]

To me, the only two remaining issues with LOS/LOF are regarding friendly vehicles and action spots in buildings. Everything else seems to work quite well. Heck, the two examples I'm about to elucidate work well the VAST majority of times.

Friendly units can use friendly vehicles to block incoming fire, but are able to fire through the friendly vehicle. This violates the two way LOS/LOF rule. It doesn't happen much, but I will be the first to admit I love using friendly wrecks to help my guys out.

The other issue has been brought up before, and addressed in detail by BF.C. An enemy unit may fire from a window on the edge of a building. As long as that particular enemy soldier is in LOS, friendlies may fire at him. Two LOS/LOF is maintained. However, as soon as the enemy ducks/cowers below the level of the window, he disappears from LOS. At that moment, the friendly will be forevermore incapable of area firing on that window. The enemy unit MUST be in LOS to allow LOF to the window. This is a result of the limited number of action spots in each building. Again, this is known and has been discussed at length. (Note: this only applies to SOME windows, based on action spots.)

Otherwise, am I sometimes frustrated by my BMP or T-72 not spotting something? Yes! But that adds to the WeGo drama. I can only imagine the shouted curses going on inside the armor vehicle. So I do imagine them - and speak them aloud!

Ken

[Sgt Joch]

The answer to that is, that it is. The blitz allows CMSF games to be reported on their ladder and over at The Few Good Men i believe Bootie is going to create a CMSF ladder once the auto ladder is up and running for CMx1. Also at FGM i'm taking part in a CMSF campaign.

I checked out the "few good men" CMSF campaign "operation patriot swords", it looks very nice.

It's nice to see more sites are now catering to the CMSF multiplayer community.

At "World at War", we also have a CMSF section with a dedicated forum, ladder and tournament.

[Jonny(FGM)]

Yes patriots sword has been fun to take part in, it's like no campaign i've taken part in for CMx1.

I'll go check out World At War after thursday (my last ever exam ) to see if i can pick up some games : )

[Battlefront.com]

The only real quirk I've seen is that when three tanks are lined up (in one line), it's possible to see the last tank before the tank in the middle. Very rare though.

Yes, that is because Vehicles don't block LOS. So it's a simple result of someone spotting them out of order. The thought of asking the CPU to check LOS based on variable blockages scares me My guess is the CPU can't handle it.

Plus, if we suddenly had more horsepower to fool around with I'd want the frequency of spotting checks to be increased rather than become more refined. That's because the more common issues revolve around spotting check frequency, not fidelity.

C3K,

Yup, the building thing is definitely the most annoying bit in the game. It's a limitation for now but, eventually, will be fixed by having each Action Spot up into "mini" action spots for targeting purposes. It's something we hope the computer can handle in a year or two.

The problems with vehicles blocking LOF one way and not the other is a big, nasty TacAI issue. We're not likely going to try and fix that for some time to come, unfortunately.

Steve

[MeatEtr]

Yes, that is because Vehicles don't block LOS. So it's a simple result of someone spotting them out of order. The thought of asking the CPU to check LOS based on variable blockages scares me My guess is the CPU can't handle it.

Plus, if we suddenly had more horsepower to fool around with I'd want the frequency of spotting checks to be increased rather than become more refined. That's because the more common issues revolve around spotting check frequency, not fidelity.

C3K,

Yup, the building thing is definitely the most annoying bit in the game. It's a limitation for now but, eventually, will be fixed by having each Action Spot up into "mini" action spots for targeting purposes. It's something we hope the computer can handle in a year or two.

The problems with vehicles blocking LOF one way and not the other is a big, nasty TacAI issue. We're not likely going to try and fix that for some time to come, unfortunately.

Steve

So.... then what's the word on Duo and Quad core support? No the excuse that wargamers are typically behind the normal gaming tech trends is not valid anymore, was fine three yrs ago. At least in relation to multi-core support, these CPUs have been on the market for quite awhile now.

Besides, we all know CMx2 is CPU extensive, so it certainly needs it. I'd think it's time to take advantage of already existing CPU horsepower.

[c3k]

Agreed: hepta-cores are the wave of the future...

The two issues I noted above are, to me, minor and well understood. Yes, there are circumstances in which they have a pivotal role in a firefight, but those circumstances are few and far between. Knowing how they occur lets you plan around/for them. The same for the rooftop hyper-elevation issue. Minor annoyances at worst. It would be nice to sort them out, but they are not gamebreakers by any stretch.

Ken

[Battlefront.com]

IIRC close to 20% of Steam's customer base (i.e. much more mass market than our customer base) still has a single core processor. That was quite surprising, actually.

The primary blockage for doing multi-core support is the several months of dedicated recoding that is necessary just to get the same game we have now using the additional processor speed. Add to that whatever time it takes to expand feature sets to use that additional power.

Even after Normandy is out it will still be a bad use of our time because we'll still have to be very conservative with the new power else we alienate a big chunk of our customer base. I'm not just talking about those who have single processors, but also those who have less beefy dual processors. Add those two groups together and they probably amount to a big chunk of our customer base.

Steve

[Redwolf]

People who like game that aren't mass-market shouldn't get CPUs with many cores, but slow per-core speed. Has never been different, probably never will.

The same issue comes up with quite a few popular games, too, some MMORPGs in particular. People just like to skip research before spending money.

[Paper Tiger]

How many folks here are using Duo Core processors, or Quad? I have a Duo Core 64 bit processor and while CMSF actually runs fine for me as it stands, as the game engine develops, it's going to demand still more and more of our CPUs so surely, this would be development time very well spent, perhaps after Normandy is released? This really is a very CPU intensive game and it's only going to get more so.

[MeatEtr]

IIRC close to 20% of Steam's customer base (i.e. much more mass market than our customer base) still has a single core processor. That was quite surprising, actually.

The primary blockage for doing multi-core support is the several months of dedicated recoding that is necessary just to get the same game we have now using the additional processor speed. Add to that whatever time it takes to expand feature sets to use that additional power.

Even after Normandy is out it will still be a bad use of our time because we'll still have to be very conservative with the new power else we alienate a big chunk of our customer base. I'm not just talking about those who have single processors, but also those who have less beefy dual processors. Add those two groups together and they probably amount to a big chunk of our customer base.

Steve

Keep in mind though, older dormant Steam accounts are also factored into that as well. So I'd say that 20% single core users is certainly higher than it should be. This is driven down when more multi-core users submit their survey. But yeah, mainstream Steam users will always have higher end systems than BFs customer base. Still a good measuring stick though.

I bet if you guys did a survey of your customers you'll be surprised single core is not the norm at all. Times have changed since CMx1, if you bought a new PC in the past 3-4 years, odds are it's at least Duo core. Perhaps a newsletter survey poll sent out to all who bought CMSF and any modules would work wonders.

I honestly don't think customers would feel alienated, they'll just simply upgrade. I know a few that did when CMSF went gold. Your customers are much older with income rather than little Jimmy bugging mom and dad for a beefy PC. I think most would be willing to pay more for a CPU to play a superior product.

[moneymaxx]

I have to admit I do not understand too much about programming but wouldn´t it be possible to use something like CUDA (use the GPU as a paralell computer) in the distant future? Thousands of parallel threads seem to be ideally suited for e.g. LOS/LOF checks, bullet trajectories and so on.

[Thomm]

But if that were true, why isn't CMx2 used in the wargaming community for turnaments, ladder-gaming and huge campaigns like CMx1 was used?

Steiner14,

You certainly put a lot of thought into your initial posts, and you certainly have a point.

Please allow me to add the following:

a) All the talk about one-to-one modeling being inferior implies, in my opinion, that there is no room for improvement in the one-on-one modeling. For the sake of the argument, let's assume that in - say - five years, the CMxn engine will show no noticeable flaws with regard to the 1-1 modeling, that is, soldiers will be spread out nicely, take cover, peek around corners, etc. When this point is reached (perhaps it will take ten years, but it will eventually happen) don't you think it will have been a good idea to start going this way (I guess the answer is 'No!', but I tried). On the other hand, if they stuck with the lumped squad model, how would they have improved the game? By putting more polygons on the three soldiers?

You mention the huge visible following of CMx1. Let me describe my situation just to balance out one of those of the 'hardcore CMx1 followers'. After buying CM:BO and CM:BB, I refrained from buying CM:AK until I picked it up from the bargain bin. I played two battles, then decided that I did not have time to sit down for hours to finish even a single battle and uninstalled it. Along came CM:SF. So far I bought every module except CM:Marines, for which I was involved in beta-testing. I dare not calculate the monetary worth of all the testing hours that I put into this game. I played it almost daily (compare CM:AK above) with great pleasure. And I am still looking forward to CM:NATO! I doubt that I would have bought another CMx1 style game. So there you have it: Even though I never even finished the one lone CM:BO PBEM that I started a lot of Euros and hours went to BFC, even though I am not visible on any ladder or campaign or whatever.

Just one counter-example.

Best regards,

Thomm

[Elmar Bijlsma]

Judging from my experience from games that were patched to make use of multi-core CPUs the gain is pretty minimal. It would appear that most of the additional CPU power is spend juggling the data to and fro.

Until such time as we see a huge leap in performance I'm fine coping with a single core.

[Ryujin]

How many folks here are using Duo Core processors, or Quad? I have a Duo Core 64 bit processor and while CMSF actually runs fine for me as it stands, as the game engine develops, it's going to demand still more and more of our CPUs so surely, this would be development time very well spent, perhaps after Normandy is released? This really is a very CPU intensive game and it's only going to get more so.

I'm running quad @ 2.83ghz and a dual core 4870x2, it's kinda sad to see CMSF slow a bit a on big battle when your only using 25% of your processing power. At least the graphics card seems to use both cores.

The issue is at what point do they make a dual core required? While you can use multiple cores to speed things up, at the moment CMx2 would still be constrained by having to be playable on just one core. At some point they could go to requiring a dual core and ramp up the engine a bit (probably not till CMx3 from the looks of it).

A survey would be interesting though...

[Thomm]

Hi MD,

I would greatly appreciate it if you did not drag any of my posts over to a place where I cannot respond (technical issue that apparently cannot be resolved even by the mighty Geordie.)

Thank you very much!

Best regards,

Thomm

[Elmar Bijlsma]

Oh gawd, you made me look.

The lunatic fringe just as loony as ever.

[Ron]

Oh gawd, you made me look.

The lunatic fringe just as loony as ever.

Yep, too rich.

Question: How many disgruntled CMSF detractors from GS does it take to change a light bulb?

.

.

.

.

.

.

.

.

Answer: Change?

[Fat Dave]

HA! nice one.

Cheers for that Ron.

[Redwolf]

I have to admit I do not understand too much about programming but wouldn´t it be possible to use something like CUDA (use the GPU as a paralell computer) in the distant future? Thousands of parallel threads seem to be ideally suited for e.g. LOS/LOF checks, bullet trajectories and so on.

Computation on graphics cards and multithreading on the main CPUs are entirely different matters.

Computation on graphics cards is surprisingly unhelpful with anything but straight floating point number crunching. Most LOS mechanisms are not straight number crunching but go through extensive lists of previously excluded regions, shifting the problem more to a logic crunching exercise.

Also keep in mind you need to copy all the data to the card first, that's not free.

[Battlefront.com]

Elmar,

Judging from my experience from games that were patched to make use of multi-core CPUs the gain is pretty minimal. It would appear that most of the additional CPU power is spend juggling the data to and fro.

Until such time as we see a huge leap in performance I'm fine coping with a single core.

This is what Charles keeps telling me every time I ask him He said the amount of noticeable improvements people would see are far less dramatic than people think. I don't understand the technical reasons for this, but clearly 4x Processors ≠ 4x Game Speed Improvement. Plus, as I said, unless the game is specifically coded to use this extra power you won't likely notice any improvements at all.

As for the lunatic fringe... I can only imagine, which is more than I care to do. I haven't read any of the tripe for well over year now because they never have anything new to say. It is their opinion that CMx1 is better than CMx2, always was and always will be. And they are welcome to their opinion, as irrelevant and out of touch with reality as it may be.

Steve

[Sgt Joch]

a) All the talk about one-to-one modeling being inferior implies, in my opinion, that there is no room for improvement in the one-on-one modeling. For the sake of the argument, let's assume that in - say - five years, the CMxn engine will show no noticeable flaws with regard to the 1-1 modeling, that is, soldiers will be spread out nicely, take cover, peek around corners, etc. When this point is reached (perhaps it will take ten years, but it will eventually happen) don't you think it will have been a good idea to start going this way (I guess the answer is 'No!', but I tried). On the other hand, if they stuck with the lumped squad model, how would they have improved the game? By putting more polygons on the three soldiers?

good post.

a few more thoughts, 1:1 representation is not just a graphical change.

In CMx1, infantry does not, for all practical purposes, exist in the 3d world. Infantry is treated as in a 2d wargame, an entire infantry unit is seen by the program as just a single point. Your infantry unit interacts with "terrain" as a 2d unit would... so many movement points to enter a road, woods or buildings with a combat modifier for incoming fire depending on what terrain hex it occupies.

It also has the effect that terrain does not really exist for CMx1 infantry. You can have an infantry unit enter any hex, except impassable terrain, by paying a certain number of movement points. Infantry can walk through walls and enter buildings whether there is a door or not and can fire at enemy soldiers through a solid wall.

In CMSF, infantry exists as fully 3d. Each soldier exists in the 3d world as a separate entity, basically a mini-vehicle. This has many effect, for example, soldiers need to go through a "door" to enter a building. They need to look through a "window" to spot enemy soldiers or fire at them, etc. The major change brought about by 1:1 was much more to allow infantry units to interact in a more realistic way with the 3d world, at least as far as movement, pathfinding. This may not be as crucial in a rural map, but makes a big difference in an urban environment. It is, of course, not perfect since issues still remain (i.e. men bunching up, lack of formations, etc.) but IMHO, is a distinct improvement over the infantry modeling in CMx1.

You mention the huge visible following of CMx1. Let me describe my situation just to balance out one of those of the 'hardcore CMx1 followers'. After buying CM:BO and CM:BB, I refrained from buying CM:AK until I picked it up from the bargain bin. I played two battles, then decided that I did not have time to sit down for hours to finish even a single battle and uninstalled it. Along came CM:SF. So far I bought every module except CM:Marines, for which I was involved in beta-testing. I dare not calculate the monetary worth of all the testing hours that I put into this game. I played it almost daily (compare CM:AK above) with great pleasure. And I am still looking forward to CM:NATO! I doubt that I would have bought another CMx1 style game. So there you have it: Even though I never even finished the one lone CM:BO PBEM that I started a lot of Euros and hours went to BFC, even though I am not visible on any ladder or campaign or whatever.

The huge continuing following for CMx1 is not supported by the numbers. So far, CMBO was the biggest seller for BFC. Sales dropped off for CMBB and dropped off yet again for CMAK. Sales of CMSF and modules is close to the CMBB level.

CMAK had all the qualities which the CMx1 diehards say they want, (i.e. CMx1, WW2, western front, cherry-picking QB w. random maps), yet had the lowest sales of all. Yes, CMAK is still popular in various CM gaming clubs, but the fact that several hundred gamers are still playing CMAK is irrelevant if BFC does not earn a substantial profit.

[hoolaman]

In CMSF, infantry exists as fully 3d. Each soldier exists in the 3d world as a separate entity, basically a mini-vehicle. This has many effect, for example, soldiers need to go through a "door" to enter a building. They need to look through a "window" to spot enemy soldiers or fire at them, etc. The major change brought about by 1:1 was much more to allow infantry units to interact in a more realistic way with the 3d world, at least as far as movement, pathfinding. This may not be as crucial in a rural map, but makes a big difference in an urban environment. It is, of course, not perfect since issues still remain (i.e. men bunching up, lack of formations, etc.) but IMHO, is a distinct improvement over the infantry modeling in CMx1.

The problem and the reason the two "camps" can't find any common ground is that the difference is a philosophical one, not an objective one-better-than-the-other situation.

The CMx2 knockers believe that infantry bunching up and lack of realistic infantry manouevres and formations means a graphical representation that is unrealistic, and therefore combat that is unrealistic. This is true in some situations unfortunately.

I believe CMSF infantry modelling is much better, but I find it is much easier to be happy with the results if you see that there is much more of the CMx1 infantry model with all its neccesary abstractions in CMSF than the CMx1 devotees really understand.

My view is that CMSF infantry is still not really explicitly simulated, instead of a single point, a squad spreads out into an area more or less representing where a real squad would go. The combat calculations are far more "design for effect" than anything, with some abstracted cover from terrain still in there, HE effects modelled for effect rather than realism. When a HE shell landed right between your three-man squad tile in CMBO and not all of them died, you understand that it is just a fudge. CMSF is the same.

Of course individual bullets still kill individual men, and individual men fire individual weapons, and sometimes individuals end up in places that get them killed unreasonably. Some people have a big problem with this and believe there is some kind of fundamental scale mish-mosh of part 1:1 and part abstraction. I think it is all abstraction to an extent, I use my imagination just as much playing CMSF as I ever did in CMx1, and I think of CMSF infantry as being pretty much what the community asked for CMBB infantry to become: for every soldier to be represented, every weapon to be shown, individual morale states instead of whole squads routing off the map (this is a big one!), animations of guys mounting up on half-tracks etc.

Really all this does is make for a much more interesting tactical environment. CM games are all just chess after all IMO. CMSF just has more squares and more pieces to play with. I don't think we will ever see BFC prduce that perfect 1:1 modelling that Thomm mentions above, but I don't think they really need to.