I'm experimenting with settings in GamePrefs.ini, in particular, MaxMWSmoothPercentage.
Setting this number higher will be
// slightly more jumpy, while lower numbers will tend to induce warp if
// remote players maneuver really quickly. - from the notes for MaxMWSmoothPercentage in GamePrefs.ini
What is the difference between a ship moving in a warpy way and one moving in a jumpy way? Is difference between the on screen world and the internal CPU world relevant?
thx
i think its the same
Yes, I thought they were the same, but in the notes for this setting, it suggests they are different and at opposite ends of the spectrum: a high setting causes one, a low setting causes the other?
My guess on that is that they are essentially the same, but different in degree. Jumpy means things will tend to jitter about some (tiny more frequent warps), but warp means things will look smooth for a bit but then suddenly make a big shift (less frequent but bigger jumps).
Ok then imo to make it more 1.2 like and able to predict where people will end up is to make the game more jumpy.
well warpying is more like.. u shoot a craft with laser and u watch it go striaght thru the visible enemy model.
jumppy is like of popping around from mwf resync to mwf resync, intervals in less than a second. but the player is always hittable @ these syncs
in 1.3 with mwf on i noticed a very different type of lag nathan didnt bother to look @. perhapps its my "scout speak".. any1 thats willing will understand what im typing even if the gammer isnt perfect.
Any how.. this new lag.. to put it ultra simple its like a mininute 1.2 blink effect that is in effect when a player/ai unit is moving.
ei the visible model is lagging behind so much so "landing the shot" in on a person/ai craft is actually harder...
i first noticed this when fighting a warrior one time in the beta Some build in a galaxy far away.... the warrior was facing me and i was facing him he was circle strafing to my right while shootin. But his shots where originating 1 whole ships width right of where i saw him. Now that is just plain bad. A unit moving as slow and as constant as a strafing ai warrior and u cant even shoot in the middle of the visible Model and hurt it..
if anything should be done at all.. the visible model should be *Ammo*ing glued to the hitbox so like any other standard fps u can actually HIT what u see when you SEE it hit the enemy.
a certain netsmooth will prolly reduce this..
I'm not sure if the battlezone ii internet world can ever really be made habitable for human pilots, at least this is my prejudice. This is why I rarely play strat or DM games and stick to MPI and single player BZ2. The shadowy relationship between what I see on my screen and what is going on from the other player's perspective is what breaks the more competitive aspect of multiplayer for me, and this doesn't necessarily have anything to do with whether scouts can hover indefinitely or not.
I don't think 1.3 has made any significant changes to this aspect of the game compared to 1.2, and it sounds like at best, with further tweaking, it might be brought back to what the dedicated team strat players have habituated to and learned to live with under 1.2.
I know this is an extreme example, but all I have to do is look at the behavior of a human piloted turret or guntower as it apparently is wildly spraying shots towards all points on the compass, when in fact the gunner is really only making mild adjustments to his aim, for me to give up on the idea of this game ever being a "realistic" multiplayer experience. This has nothing to do with high school physics and everything to do with immersive entertainment.
Tweaking the the "network.newskew", "network. maxsmooth", and "network.velsmooth" variables has virtually no discernable impact on the behavior of piloted turrets and guntowers, which suggests to me that these settings really don't get at what (at least for me) is the fundamental flaw in BZ2 internet play.
Ok, this is a gripe as old as the game, now; so, there's nothing new here. Probably best to just move along.
Quote from: Spacecomber
I don't think 1.3 has made any significant changes to this aspect of the game compared to 1.2, and it sounds like at best, with further tweaking, it might be brought back to what the dedicated team strat players have habituated to and learned to live with under 1.2.
I disagree; it does seem to be worse. I also think that team strat in BZ2 is the best multiplayer experience available, but I guess its all a matter of taste. I'd be happy to accept 1.2 levels if this were the best available, and have 1.3's other features as well. What I'd like to avoid is having to choose between 1.2's superior movement and 1.3's superior features. I'm not particularly concerned with infinite hovering either way. It only becomes a concern for me if removing the ability to hover also results in a degraded performance of mobility generally. I don't know if there is a close connection between the increased warp of 1.3 and the changed movement of 1.3.
@Scout (or anyone else): do you know if Shadowers lock on to the visual position or the internal position of an enemy ship? Because they seem to miss a lot more in 1.3. However, FAF missiles are at least as good compared to 1.2.
"The shadowy relationship between what I see on my screen and what is going on from the other player's perspective is what breaks the more competitive aspect of multiplayer for me"
Very Very well said.. if bz2 had netcode in the range of.. UT2k4 where they do have craft AIR/HOVER/Wheel that can move extremely fast, where i can still no scope with the lighting rifle, which i do...
Then Bz2 would have another dimention to it. The COmpeditiveness would be great fun for all and the community would just feed itself.. like Cs.. but with every RTS+fps hybrid ud have to minus the Jar head script kiddys...
besides, if bz2 was good enuf to make demos and had a spectator mode.. i wouldnt doubt it coming up into the CAL range.. but i guess ill have to settle for jetpacks and and lerks in CAL-NS =/
Spectator mode just finished for DM, without ANY bugs I can see. If you want the map to see, just PM me Scout. I tested it with stalker, and it works really well.
Battlefield 1942 is another MP game with server side hit detection. AFAIK it dont use any client prediction or smoothing at all. The effect needs to get used to but is perfectly predictiable for the player. You just have to know that fast moving objects are in fact (from the referee´s aka servers eyes) always a certain length in front of where the client sees them. If you want to hit a tank moving perpendicular to your shooting direction you have to lead your target so much that the shell passes the visible tank about 1 length in front of it, depending on the ping. On planes I have to lead about 3 plane length when ping is like usual 70-80. All movements are absolutely smooth except sometimes on 64 player servers. The only problem is that the client hit animations are all useless and irritating for newbies. To reduce this irritation and to provide a reiiable feedback they implemented a server generated hit indicator with one of the patches.
All the attempts to bring client animations in line with the server´s world only seem to increase warp and reduce predictability for the player.
I probably allowed myself to get too caught up in the spirit of frustration that is rampant on these boards, since 1.3 was released. Let me make clear, I think that 1.3 is a good thing. I just think that for almost all of us BZ2 has provided a bittersweet experience. For many, especially those coming to BZ2 straight out of BZ1, there was a lot to be disappointed in with the original release of BZ2. The two subsequent patch releases (up through 1.2) made some good improvements, but for me (and I want to speak only for myself on this) dogfighting has always been a disappointment for the reasons I mentioned above. With each new patch, there was always the hope that everything that every individual thought BZ2 could, or even should be, for him, would finally be realized. Of course, this was never going to be the case, so those old frustrations and disappointments just come back to life for another irritating round of venting. (Old vinegar in new bottles, I suppose.)
I know that, for those who enjoy 1.2 strat, what I find to be too irritating, in how the game handles multiplayer dogfights, has been counter-balanced by the game play, itself. I know that any of the faults that I complan about can be adapted to, just as we train ourselves into any new game. Whether we will or not is simply a matter of taste.
No game is perfect. As GJC pointed out in regard to BF42, what you see on the screen isn't exactly "right", either. It requires a player to "lag lead" his aim, which is pretty much the standard situation for internet "player on player" action (although BZ1 was peer to peer, this aspect of the expereince was the same). For my personal taste, this is much less disturbing to my ability to get lost in the expereince of playing the game. You need to lead your aim in BF42, anyway, since none of the weapons are instant hit (they all involve projectiles). You just exaggerate it a bit, depending on your ping to the server.
The advantage to BZ2's netcode is that is harder to run your typical hacks (increased armor, aim bots, inscreased hit damage, etc.), but it came with it's own bag of exploitable weaknesses, many of which had the purpose of making what I complained of as being shadowy into something completely illusory.
Speaking of BF42, BMR has just set up their own desert combat public server. It's a 30 person server running out of St Louis, so most of the USA should get a decent ping. We have no where near enough members to keep a server this big busy, so feel please drop by and help us fill up some of the slots. 66.165.90.43
Space
I wouldn't want to prevent anyone discussing bf42 or how other games behave, or aspects of BZ2 that people do or don't like.
I'd like to focus (in one or two threads) on how to improve one particular part of 1.3: warp. I understand that some people may not like dogfighting, hopefully others understand that some people particularly like this.
So...
I searched my BZ2 directory for "velsmooth", and all I could find was that this option had been removed from GamePrefs.ini. Can it still be changed via the console? Is this a path worth exploring?
I looked for my earlier post about the personal settings I was trying, couldn't find it (please let me know if discussion is inappropriate). I'm using:
MaxMWSmoothPercentage = 1.0
UseNewTimeskew = true
stock 1.3 version of BZ2 with default Recycler
Seems to work well. As server I was getting hit by blast, and was landing laser shots on target.
I have had several test games where there was almost no warp or lag what so ever and then I'd had several with lag and warp all over the place. The only thing I can point too that was different was the type of connection everone was playing on.
I would ask people to start keeping track of what type connect each player has during a game so we can start to see parterns with when it warps/lags.
I also have a the idea that movement date is being sent or receved out of order which might also cuase this funky warping stuff.
but the wrong data transfer also exsists in 1.2
and i find 1.3 more warpier with alot others
and the MWF is fixed - which was another cause of 1.2 to be real warpy
and still, 1.3 is warpier :s
Actually I find 1.3 less 'warpy'. I have a 3mbit connection, average of about 40 ms latency with the people I play. In 1.2 we would still warp around, jump occasionally. In all my time in 1.3 with the people I have 40ms latency with, I have seen a warp once, and that was a hickup in my connection from downloading files at the same time.
Quote from: aougliI wouldn't want to prevent anyone discussing bf42 or how other games behave, or aspects of BZ2 that people do or don't like.
I simply wanted to explain what the crux is with internet gaming. You either have a client prediction or have to predict yourself. I prefer the latter one.
Getting back to what can be adjusted in the game through gameprefs or console commands, as aougli suggested, the network.velsmooth [value] command still works. The default is currently 0. The numbers seem to work in the opposite direction of the network.maxsmooth [value] setting. (I hope I'm remembering what I saw right.) With a high network.velsmooth number (lets say 2 or 4), when you watch a human pilot take a step, after this stepping behavior has stopped, they seem to slowly slide into their actual new position. With lower numbers, less than one, like .4, at the end of the stepping action they don't continue to slide into their position. Network.maxsmooth seemed to work in the opposite direction. You would see human pilots doing their sliding with low values, less than one, but stepping and moving would stop simultaneously with higher numbers.
I know no one really cares about these pilot actions, but I give them as a way to try and get a sense of what these settings do. Hopefully, it will provide a better sense of whether the settings need to go up or need to go down in order to have the proper effect on what you are seeing in dogfights with human pilots.
@Silver: yes, warp symptoms in the game are not consistent. But as I am experimenting with my settings, and the same people are not in the server each time, perhaps that is to be expected.
@Over: I find warp in 1.3 is definately different; I'm not sure whether it is better or worse.
@Space: thanks for the info. If I've understood correctly, velsmooth is by default at its lowest possible setting, and that this is the one that provides the greatest positional accuracy for aiming an incoming shot. And as maxsmooth covers similar territory, is there any point to using both variables, especially as one may contradict the other?
The experimentation continues! 8)
I really don't know much about the two variables; so, I can't say how much they are basically doing the same thing by different routes versus how much they really are different variables. Commando probably has the most hands on experience with experimenting with them prior to the public release.
I've not gotten beyond trying to figure out an easy way for me to observe the effect of these settings using my two computers over a LAN connection.
Had some more games with stock 1.3 with maxsmooth = 1 and timeskew = true.
Basically it's OK. From my own experience and asking others after the game, 1.3 is not much worse, not much better, than 1.2 when it comes to warp. However, it is different. Seems to be necessary to acclimatize to the new environment, which is neither better nor worse, but different. The games I've had are fairly typical for me: me as server, most people on broadband, one or two 56k people in as well. Even with 56k across the Atlantic ocean the game is playable and fun. 56k people report the most variation in performance, with some saying there is a noticeable improvement in network performance, and other saying it is markedly worse.
I've lowered the server's MWF from my usual 5 to 3. Didn't make much difference to me, but others in the server wanted it and say it is better.
Do we have any data on what the possible/closest values are for these smoothing commands in 1.2? Maybe it's somewhere in the closed beta section?
Regardless if it's strat/dm/mpi/ctf/race, I do kinda see lots of warpage in 1.3. Apart from that, the best thing I've noticed (as I mentioned in another thread) is that my dialup ping settles about 5 times faster than in 1.2 right after joining the gameworld. That's pretty slick to me. :)
There are some notes in GamePrefs.ini itself by the various variables.
1.3 seems a lot more tolerant of joiners and leavers to a game than 1.2.
Host setting mwf low is CRUCIAL to 1.3's enjoyment in
strats.
I have played plenty of stock 1.3 games now and when mwf2 the
doggyfights are excellent, predictable movements and minimal
warp being the norm.
When Aug hosts at mwf5 the warp is considerable to me, having
a 56k modem connection.
I think ALL players having same mwf actually may reduce load on
the server in some occult way, never mind what the mwf is set to.
Try mwf2 Aug, it is good so far and your comp should do it fine.
Well good thing 1.3 sets everyone's MWF to what the host chose :)
Perhaps the default should be lower though...