Jump to content

The Tanklord

Members
  • Posts

    42
  • Joined

  • Last visited

Posts posted by The Tanklord

  1. Steve,

    The Tanklord,

    A thoughtful, strong, and constructive counter post. Just what I like and what I so rarely get when talking about something like this.

    I hope we can continue with it!

    I hope so too, discussing stuff like this with a gamedev is a treat. (No matter how stubborn said gamedev might be ;))

    That's exactly what I meant when I said "We predicted that years ago, as we've discussed many times before. The people most bothered by it

    are the ones who left, which makes sense." I made that in response to your comment that "some (lots?) have left." So I guess I don't

    understand your point.

    I was just trying to make the point that a couple of players that changes their minds proves as much as couple of angry players leaving,

    not very much. Other factors will judge your success.

    No argument there with the theory. The difference is that in reality I don't believe it's very relevant to CMx2.

    ...

    What I'm doing here is not disagreeing with your core argument, I'm just saying that in practical terms it doesn't mean much.

    Sure, in theory there are nearly unlimited resources at our disposal if we go with WeGo, but in reality there are other considerations

    that do indeed impose limits. The other one is...

    Great, then I think we agree on the theoretic/abstract limitations, these are the ones I wanted to claim as "fact".

    The practical impact, if any, is obviously more about opinions, especially from the outside where cause-effect relationships

    are even harder to deduct when it comes to such a complex piece of software as CM.

    (Agreeing with the point you made about pathfinding and working in complex environments and the increased possibility of "non logic" bugs)

    Since I don't own the game, so I'm saying this very weak on hard facts (and the rest of my post should be seen in this light when it comes

    to game specifics), but I do believe that you could design situations where the engine chokes.

    If so, then it becomes a discussion about how rare/common these situations are in common scenarios and

    their importance in the bigger picture.

    I do agree that WeGo/Turnbased isn't magically gonna solve the problems that are hard to compute, BUT it does put the decision

    about the size of the map/forces involved more in the players hands. If you felt like playing one of Runes more evil creations you could

    do that, even if your computer wasn't top of the line and still get basically the same results as a faster computer, in theory atleast.

    The limitation was how long you were willing to wait or when your graphics card hit the wall, both looser limits than in RT for almost

    everyone.

    If you have W amount of work, T amount of time and S amount of work done/time unit, if W/S > T something has to give, in a turnbased

    environment it's the time constraint you loosen (up to a point, of course). Since you only have control over W and T (cut work or delay),

    how do you handle this in the new engine? (I'll sign an NDA to hear I'm right :P)

    I doubt that you have baseline W so low so that peak(W) never invalidates peak(W)/S < T (when T is in realtime).

    In respect to LOS/LOF and Pathfinding, of course these are things which require every increasing amounts of computing resources to

    achieve ever diminishing returns. So one can take these things so far that the computer would basically sit around for a half hour

    computing nothing but LOS/LOF for a few seconds of gameplay.

    Well, when it comes to pathfinding in a complex enough environment you might have to compute the whole path to find the/a solution, in

    a labyrinthesque situation for example. Obviously this will be a problem in WeGo too, but the problem could be somewhat larger before

    that happens.

    It might also make a difference is in those games where in CMx1 the basic turns took maybe 50 seconds and

    the high intensity combat/lots of complex movement turns took 1½ - 2 minutes. In RT complaints about lagging comes up fairly

    quick and lagging has a (very?) negative impact on the players ability to control his forces in RT whereas 2-3 minute wait time for a turn

    in the best PBEM you ever played might not disturb you as much...

    [optimization]

    No argument there.

    Sure, to "some degree". I agree with that. Just like going with WeGo imposes tradeoffs that impact the gameplay to "some degree" as well.

    yes, and if you want the intensity and other features of RT it's a rather large tradeoff to go Turnbased!

    I'm only discussing this from the very narrow point of view regarding some of the computational problems that are core to a game engine and

    the theoretical limitations. And, since we seem to agree about the theoretical, now more about what I believe the practical impacts are.

    I do agree that the positive impact of a RT engine to gameplay for players that like RT is much larger than the negative impact of the

    harsher technical constraints impacts WeGo/Turnbased.

    The best example of this is in CMx1. Sometimes a shot would pass through a building and impact a unit behind it. From a physics and gameplay

    standpoint, this did not put smiles on people's faces. The reason for this was that we didn't track things in RealTime within the game engine

    and therefore there was no way to account for the movement of the unit after the shot was launched. Or put another way, real trajectories were

    not possible because it required a degree of tracking that simply did not exist because the system was inherently based on discrete time increments

    instead of continuous time.

    I don't really see your point here, whats to technically stop you from doing a pure WeGo-engine have the same time-granularity as a RT-engine

    "under the hood" and just have very loose deadlines for those problems that benefits from it? You still need some sort of internal game clock to

    synchronize against. Something like a RT engine where you allow it to lag as needed (within before mentioned limits on WeGo) and then playback the result in 1:1 time. It should be very close to what you have now.

    I do see that once you have all that stuff in place, the step to RT is rather small and probably very tempting and possibly even the right

    thing to do ;)

    What I'm saying is that it isn't a tradeoff in the way you say. You're arguing that a turn based system has no cons, only pros.

    Technical pros yes, I have a hard time seeing what would be better/easier to do in a RT engine or what would lead to a better

    practical results with regards to the algorithmic challenges posed by a game engine.

    Gameplay wise the differences are huge with many pros and cons on both sides and since you seem to be doing good, despite the (imho!)

    lackluster syrian scenario it might even the right one. Although you might have inertia-sales from your earlier games both in SF and

    the first WW2 game, so after that I think the real tests of the new engine/path will come. I'm crossing my fingers though, needed or not ;).

    I'm not sure that we are completely on the same page with regards to the differences between engines, and since I'm only speculating this is

    most certainly the case.

    The difference from my point of view is that in RT you need to calculate and results in realtime and output it to the screen/player also in real time,

    whereas in a turnbased/pre calculated/blue bar engine you can pause the simulation when you get a peak in the workload and get arguably better results.

    And obviously you can only get better results when you have more work than time.

    That's simply not the case. Both have cons, both have pros. We feel that, on balance, the pros of the current system (inherently

    continuous time based) outweigh the pros of the old system (inherently turn based). Therefore, we didn't have to tradeoff anything

    significant when the whole is examined.

    No argument there, I'm just arguing for some pros for a WeGo type engine in some (small?) areas when compared to RT. I felt that a Turn Based

    engine (and the CMx1 engine?) had some features/pros worth defending and maybe, just maybe, get some of these pros back in for a (far) future

    release just for the WeGo/PBEM crowd.

    Absolutely. I think the track record of what is in the patches since CM:SF came out really underscores our ability to deal with both of these

    issues. We didn't throw out 3 years worth of coding, as some suggested we should, yet we spent many months not only fixing what was broken but

    also adding in new features that people requested we put in. And for some of those things which we agree with, but are not practical right now,

    we've said we'll get to them as soon as we can (committing to a timeframe for some, not for others). Since the early days we've been accused by

    people of not listening to our customers... it's never been true, and never will be. What is true is that we don't listen to the customers that

    would wind up putting us in the unemployment line

    Yes, your support of your games is excellent and combined with the personal contact on these forums it's truly outstanding.

    It goes without saying that you can't throw away 3 years of work, I'm not arguing for anything like that either.

    And none of them have anything to do with the RT nature, just to point that out QBs are getting a massive overhaul and, I think, will wind up

    being far superior to what was in CMx1.

    Yeah, I saw that and I'm still worried ;). One of the very best games I have played was in a CMBO-tourney over at band of brothers that was played on maps

    that posed "unrealistic" tactical challenges, buy anything you want, but no mixing branches and you got to see the map before you buy. (Perhaps some

    limits on arty too) This one was a small rectangular map with a circular hill with almost no cover, a church and a large (and only) flag all in the middle.

    Some mixed terrain around the hill and you had both diagonal corners to setup in, lower right and upper left for me.

    It was almost a game in it self just planning what to buy and choosing a strategy.

    I went with crack Fallschirmsjägers and an extremely aggressive plan, lost the game in a Draw 48-52 or something similar but the game and

    everything around it was a blast.

    I've had loads of fun in more historical scenarios too,for example in longer meta campaigns where losses carry over and gets replaced with green troops,

    those pose another type of challenge. Anyway, I feel that the flexibility of a point system and the powerful QB generator is hard to beat and I fear

    that we have lost the ability to try out the more exotic setups.

    Hopefully

    the bulk of this response will show you that it is a fallacy to think that you're not expressing an opinion since I've been able to disagree with

    your facts with other facts.

    I think we are pretty much in agreement about the major stuff and maybe I was unclear earlier, the only parts I wanted to claim as fact are

    the theoretical limits/differences. And I'm certain enough that you could create an _artificial_ situation in the game where these theoretical

    limitations becomes practical to some extent, I would call that one a borderline fact. By practical I mean that they show them self in the game,

    not that they necessarily impacts gameplay.

    Everything else, including the possible impact of claimed facts are opinions.

    Hopefully not your "last words" because this has been enjoyable! I hate it when people post an opinion and then run away from the debate that

    follows. It seems the more strong the opinion, the more likely that is to happen. In our case we really don't disagree much in a fundamental

    sense. We simply disagree in what it actually means in reality.

    Steve

    No no, I'm not running away, not even if the discussion should turn "serious".

    I was thinking more along the lines that the forums seems very calm and I haven't seen Matt around lately. In combination with my

    "assumption" of correcting you... *Que hitman-theme*

  2. Steiner14, that is a very valid point and obviously BFC should go the way they feel is right, imo one of the biggest mistakes for a gaming company is listening too much to the vocal part of their forum. On the other hand it's good to keep them from getting hubris too, memento mori and all that.

    As long as BFC is in business there is still the possibility that "the next game" might be just right, although beating their first games might be impossible.

    Regarding the design decisions I'm just saying that a couple of them are wrong from the-best-possible-game-for-me-perspective. They might be very right in more general terms.

    I'm not so sure about the smaller "slices" of ww2, losing the good old QB or losing the point system either, but I feel the points against those decisions has been made well enough.

    The technical aspects are less about opinions and more about facts. Steve has been saying the same things regarding RT vs WeGo and the calculation aspect since the game was launched (and before?) and I find the technical reasoning fundamentally flawed (the NO difference at all part).

    Now that the forum seems rather calm (ie no major fires that might explode from extra gasoline) and with a perfect thread to post it in to boot I felt that it was time to correct Steve ;).

    (These could be my famous "last words")

    edit: The time limited edit function is really annoying too!!!

  3. We predicted that years ago, as we've discussed many times before. The people most bothered by it are the ones who left, which makes sense.

    Yes, as it should be. What I find disturbing is the whole "those guys don't get "it"" and "the idiots even started complaining before CMSF was out" and then showing some recent re-converts as if they prove anything.

    I think a couple of people that are bothered by it might still hang around to see what you will make of the WW2 release, and I think that we (or I if its only me ;)) have heard enough about how little we "get" or how little of the vision we saw. The hotheads that you might be targeting with this have already left.

    That is only true if the requirements outstrip the ability for the computer to keep up. Tic-Tac-Toe works just as well on a 1960s computer as it does the latest and greatest, does it not?

    That's true. But comparing any part of CMSF with Tic-Tac-Toe is deeply flawed, the pathfinding for the smallest possible unit over a small distance is much more expensive than solving the whole tic-tac-toe game.

    Take chess, 8x8 grid, 32 units, and very limited movement possibilities for each piece. Compared to CMSF it's still stupidly simple, but here you need very powerful computers and precalculated endgames and whatnot to give the best players in the world a run for their money. Impose some hard ("RT") time constraints and I doubt that computers would perform nearly as well. And chess could be seen as a much simplified problem of StratAI for CMSF.

    Add layers of imperfect information, C3 and so on...

    As it is CMx2 is doing far more than CMx1 could have ever dreamed of doing in WeGo, but it is doing it in RealTime.

    Computers keeps improving at a rapid pace yes, but my argument is about how much more could have been done if it wasn't RT not about how it compares with an older game from many hardware generations ago.

    We don't have time to make a Deep Blue AI for CM anyway, therefore the endless CPU cycles we would need for it aren't relevant.

    Looking at other forum posts, there are threads about pathfinding, LOS and TacAI behaving strangely or not at all. Pathfinding and LOS are theoretically simple problems, but expensive to compute. The mention of "tweaking" suggests changing timelimits, searchdepth, pruning and so on and if you are not computing the complete solution more time (or power) will in the right/wrong circumstance improve the results. Or in another way, if you calculate the optimal answer you can't tweak the algorithm for a better one.

    Not to mention that the human player surely could use as much AI-assistance as possible when trying to organize his/her forces in RT.

    Plus, when a programmer knows he has unlimited CPU time he tends to code less efficiently than when he knows he doesn't.

    That might have some truth to it, but then there is the state the game was released in and the many patches since then, I doubt there has been enough time for the really hardcore optimizations. And to quote Donald Knuth "premature optimization is the root of all evil".

    Not to mention that it takes more of the programmers time to do things for RT.

    But to get back on track. I'm not saying these things are bad now (since I don't own the game I have to wait for the demos and the right mood), I'm saying that RT brings with it a tradeoff and that some areas could have been better without the RT constraints.

    If the tradeoff was worth it, history and your balance sheet will show that but saying it's not an issue that doesn't impact gameplay to some degree is simply not true.

    The tradeoff can only be argued for if it exists. It doesn't, so in reality there is no problem. Plus, trying to tell the world that CMx2's RealTime is "broken" or a "click fest" isn't that line of argument. It's an opinion tarted up to appear like fact, despite the evidence to show that it is wrong.

    Well, it does exist considering all the complex stuff CMSF must do and the limitations of modern computers. The impact can be discussed and is subjective, but the tradeoff is there. I'm not saying the game or its design is broken, it's a choice with pros and cons, but some of these pros and cons are technical by nature. (although I can't think of a technical pro(?) right now).

    And regarding RT and the "click fest", back in the day I played Red Alert competitively so that part is not an issue and I don't agree that CMSF is a click fest anyway. I just feel that WeGo was the way to go for CM with all its imo added benefits. But this, as I'm well aware, is just one mans opinion. And it's even possible that CM2:ww2 might be able to change my opinion on that.

  4. Thankfully, the WeGoers who screamed at us to get rid of RealTime (this call started long before CM:SF was out, which says something :)) have largely quieted down.

    Yes, that seems to be the logical thing to do, before SF was out there was still a possibility that things might change and after the release they let you know how wrong they think you are. Obviously nothing lasts forever and they have at the very least made their point heard.

    Some have even admitted that RT is their primary method of play!

    And some (lots?) have left...

    So the occasional person who still tries to claim that there is something inherently wrong with RT can be ignored since it is obviously an emotional position rather than a rational one.

    From a computer science perspective RT inherently comes with limitations , there are hard deadlines that needs to be met which will limit you in other ways. If you want the best possible pathfinding, TacAI, stratAI and so on, WeGo (or any type of offline resolution engine) is the way to go. RT can at the very best only come up to par with WeGo in these areas and this is (imho) only under unrealistically optimal conditions.

    I'm not saying that RT isn't good for CM, although I don't care for it (I've only played the demos though), but the tradeoff is there and taking the other side would still be rational.

  5. Originally posted by Steiner14:

    One exciting aspect seems to be, that it's not just programming spaghetti-code with dry algorithms and calling subroutines, but it's about modelling the real world into classes and objects (you create indeed the object "tank", or "soldier"). Then you give the objects attributes and capabilities, they send messages, do something and receive messages from other objects.

    And if you need a more specified object, they can inherit their capabilities.

    ...

    I really dont see anything that will persuade someone who rather diggs trenches ;) .

    The "dry" algorithms are mostly the same and still needs to be implemented. And if we are talking c++ OOP with some problems of its own. Memory leaks, null pointer dereferencing, deleting objects more than once etc. I'm not saying its bad or anything, just don't really see how it is a "solution" for the "rather dig trenches" crowd. (assuming they would want to program if it were more gratifying than digging trenches)

  6. Originally posted by Steiner14:

    No, you dont understand. There is no such thing, all known (algorithmic) random-generators are predicatable given their seed, no matter how "complex". So you save your random seed (time or whatever) and share that and you are good to go.

    I would love to see such a function that doesnt call external hardware and is fast enough to use for a game. Feel free to point me in the right direction.

    What are you gonna use for this? Time difference regarding to timeslices? Small variations in mouse positining? Not very usefull for a game...

    Yeah, but it has nothing to do with the random generators...
  7. Steiner, you are wrong.

    Your earlier examples (a couple of posts up) are easily saved, you seed the generator and then you can reproduce it, random isn't truly random if generated on a computer. Depending on how they have implemented it and with threads and other stuff it can be a pita to do, but you would never need to save the whole datastream.

    You need to ask yourself where that non-determenism would come from.

  8. Originally posted by Dandelion:

    Well, the basic situation after a sawmping is that he is in cover defending and you are out in the open, attacking him with a force that in firepower is in no way superior to his. Pretty bleak odds.

    Well, I don't really agree (suprise!). First of all he has given up the initiative, he is on the flags with everything hes got and intend to stay there. Secondly he has given away valuabel intel. This should maximize the effectiveness of your arty amongst other things.

    Then there is the point that usually there is more than one flag, which means he has to divide his forces while you can pick him of one flag at the time. The flags might not have optimum cover, then he have to defend from a poor position.

    To shift a determined such force one will normally have to reduce it by at least 40%. That's AI - a human opponent will in fact be able to hold on though suffering up to 85% casualties,

    and he will fight to the last man by that flag, not regrouping to anywhere. That's about 160-350 men that need to be killed or otherwise taken out of action. Can also be expressed as 6-11 dead men every minute, assuming fighting starts at turn 1 and ends in turn 30.

    Thats good since the your forces are worth a lot of points even more than the flags. It sounds like a lot of men but then again many players puts around 2/3 of the points into infantry which means you need to kill 4-8 soldiers every turn (assuming your math is correct smile.gif ) and this most people do without a problem.

    But of course, if you say it's not normally a problem, it probably isn't smile.gif

    Glad you see it that way smile.gif

    I'm unsure if we might be looking at this from different perspectives, with a historically "correct" force and doctrine you might be out of luck against this fenomenon, but then again your opponent isn't playing very historical either. If you and you opponent are playing to just compete I doubt only infantry will gain you much if any advantage esp not after the first game when the element of suprise is gone.

    I have a couple of exams during the next 9 days so I'm in all out study mode, but after that I'm willing to play you a game. You know what they say, seeing is beliveing ;)

    Regards

    Dandelion

    Cheers

    Tank

  9. Not offended at all, despite my handle I play all types of games/forces. smile.gif

    Apologies if I came across as so.

    I was wondering about the date because I thought that this might have been an issue before I started playing CMBO.

    I'm not intrested in the gamey vs non-gamey discussion either, in fact, I'm far from sure

    that this gives any type of advantage at all so the only reason I mentioned the gamey threads was because you called it gamey.

    Now regarding the intresting stuff.

    The default numbers of turns in a QB is 30 and imho thats plenty of time to clear 1500 pts of infantry with a 1500 pts combined arms force against an equally skilled opponent. At 20 turns

    it might be a problem, but then again, in 20

    turns you don't have time for much fancy stuff at all and then the obvious choice is to go for the flags.

    Of course there are maps were a infantry only force will win, what I'm doubting is that it gives some sort of advantage across the board.

    I've played with pure infantry vs CA a couple of games and I reached the flags before my opponent and I would say it's about 50/50 whether I get kicked of them or not. Admitedly I haven't tried it with a green or conscript force to maximize the quantity of infantry.

    The way I see it is that either you go for the objectives or you go for the enemy, if you go for the latter it doesn't matter if he's on the objectives or not. And yes, players that do the second usually don't rush the flags.

    And before you ask, yes I've played some good ladder players, under diffrent sets of restrictions, some silly, some not. But never have I heard anything about not going pure infantry or not using fast to get to the flags, thats why I'm curious about this.

  10. Originally posted by Dandelion:

    This used to be called "swamping".

    Suddenly, players stopped buying the formerly so popular "übertanks" - in fact they stopped buying tanks at all. Instead, they bought masses of infantry - the more the better, but quality was also a factor. At the start of a game they'd run like hell, disregard casualties completely and secure objectives. After that, there is simply no way of removing them within the timeframe of a ME. Over a period one encountered this a lot. But then word was out and it was considered very gamey. I think the ME setup dropped drastically in popularity.

    [...]

    Regards

    Dandelion

    Could you explain this further? When was this common practice? A month ago, a year, two years?

    I searched for "swamping" in the CMBO forums and archives and came up empty.

    The only time I can see this working is with parameters that allows you to dominate the map with ATGs, but then shouldn't it be the "explotion" of the parameters or the excesive use of ATGs that should be considered gamey and not the infantry horde?

    I could be wrong though, it has happened before, so I would love to hear where this is comming from and why it hasn't been in the gamey-threads on this board that I have read (far from all, but they tend to repeat themself so I should have seen it at some time.)

  11. No problem.

    Unless you already got them: (found them in another thread)

    http://gamesweb.com/pc/downloads/patches/detail.php?item_id=29252

    http://gamesweb.com/pc/downloads/patches/detail.php?item_id=31315

    If you have version 1.00 you need to get that one first, only one version availible, and then the "mutltilingual" 1.02 from the other url. 1.03 is a beta and it doesn't work for the CDV version

    and probably won't until it's "non-beta".

    For more info regarding the diffrent versions try a search on CDV, sucks, censorship and curse words of your own chosing smile.gif

  12. Mike,

    What does the computer know? It obviously didn't

    consider the excellent mansion I kicked out your troops from. The sweet rosegarden, that it's was my color on the flag on the big flagpole, no my friend, we aren't talking about what the computer thinks, we are talking about who is the moral winner, who walked away from that game with a big smile on his face.

    I know I did smile.gif

    You did mention that you had some scenarios from that tourney that we might play, or anything else for that matter.

  13. So this is your current excuse for not sending me a setup? You don´t need to do this charade in public, just let me know if it was to much for you when my Überhamsters threw your canadian lumberjacks and their "wheels" of the map.

    No need to ditch another mailservice trying to save face, running from the problem never helps.

    [ February 19, 2003, 01:01 AM: Message edited by: The Tanklord ]

  14. Voyeristic in the dark? Please explain how that's possible.... on second thought, please don't. The only thing thats certain is that seeing you in any sort of compromised position would leave me blind and it would be such a horrible way to lose ones eyesight.

    If your bedmate is protected accordingly you might want to turn on the light and go looking for the turn you owe me. smile.gif

  15. Originally posted by gautrek:

    Ok so where else am i supposed to go to complain about piss poor service then?.

    Try somewhere were someone gives a damn.

    If these people are going to set up forums and make quality games then they have to accept complaints and be big enough to take it on the chin.This is a major f*** up and i am not happy.

    Major f*** up? We are talking about a computer game that is/might be a few days late. You need to get a perspective on things.

    This forum at times seems like a mutal appreciation society where BFC can do no wrong.

    Seeing your post I understand how it might seem so to you...

    As to buying the game from another source WHY should i have to at more expense to me.This shouldn't have happened in the first place.

    Yes you are right, the BFC crew should come by your house to personally deliver your copy.
  16. Originally posted by C-J:

    Hello 2nd panzer in the LFHB tournament ended

    with a draw between Allied C(46p) and Axis Z(47p)

    It was a very exciting battle.

    Thanks Patrik for a good fight.

    C-J

    Thank you. Well played.

    Originally posted by mike the wino:

    Is that "The Tanklord" Patrik? IF so, please get back to losing to my Green Canadien troops. :D

    The one and only ;) . I would, if you could just attach a turn to one of your emails, or are you afraid, you feel the tide is turning once again?

    Not that I admit that my crack hamsters has any problems, ever, with green Canadians.

    smile.gif

    On a side note, all my games is complete, is there any of those results that hasn't reached the management? I'm a bit confused around here. smile.gif

  17. If you want speed go play a RTS or even better a FPS.

    I'm sure some counter-strike player on crack can plot his turns faster than you can say "timer".

    Would that make him the best CM player?

    Not to mention that when the timlimit/turn gets real low (1 sec or less) a monkey would be equal with anyone else. So Timelimit --> 0, Good Player

    --> Monkey. So bascially youre saying that TCP/IP players, under time constraint, are more monkey like... So, whos youre MONKEY?!?

    [ August 10, 2002, 01:51 PM: Message edited by: The Tanklord ]

  18. Welcome Sir!

    Having such a large and active reserve always pays off (two guys just notified that they are no longer available so I have none left. Don't drop out! :rolleyes: )

    LFHB Allied heads up. Set ups to the newcomer at once!

    -T[/QB]

    Thank you!

    No worríes short of major disaster I'll finish my games.

    So far I've gotten setups from A & B, I'll give the others until tomorrow before I start the email harasment smile.gif

×
×
  • Create New...