PDA

View Full Version : 7.7mm vs 12.7mm



mike_espo
04-13-2004, 10:37 AM
I was flying online yesterday flying the CR.42. Flying vs a spitV and was hitting him real good with deflection with the 12.7mms Got him smoking but no kill. http://ubbxforums.ubi.com/infopop/emoticons/icon_frown.gif

Later, decided to fly the Zero 21 in another server. Shot down 2 yak 9s and a La 5. the last one with just the 7.7mms. All in the same plane. My point being, did'nt think the 7.7mms were that effective. I hammered the spit with the 12.7mms and no clean kill. With the 7.7mms it was a relatively short burst. I know the 7.7 type 97 machine gun had a high rate of fire vs the Breda, but I would think the weight of fire would be much more with the Bredas.

Not complaining, just wondering.

thanks

"Fatte vede che ridemo!"http://www.flying-tigers.net/caccia%20WW%20II/g50.jpg

mike_espo
04-13-2004, 10:37 AM
I was flying online yesterday flying the CR.42. Flying vs a spitV and was hitting him real good with deflection with the 12.7mms Got him smoking but no kill. http://ubbxforums.ubi.com/infopop/emoticons/icon_frown.gif

Later, decided to fly the Zero 21 in another server. Shot down 2 yak 9s and a La 5. the last one with just the 7.7mms. All in the same plane. My point being, did'nt think the 7.7mms were that effective. I hammered the spit with the 12.7mms and no clean kill. With the 7.7mms it was a relatively short burst. I know the 7.7 type 97 machine gun had a high rate of fire vs the Breda, but I would think the weight of fire would be much more with the Bredas.

Not complaining, just wondering.

thanks

"Fatte vede che ridemo!"http://www.flying-tigers.net/caccia%20WW%20II/g50.jpg

Jippo01
04-13-2004, 10:45 AM
I find LMG equipped planes were hard to get a kill in. For example Ratas are exceedingly difficult to down flying a Gladiator.

Maybe you just had a very good hit with the Zero?


-jippo

LeLv28 - Fighting for independency since 2002
http://www.lelv28.com

Falkster's Ju-88 fan site:
www.ju88.de.tf (http://www.ju88.de.tf)

mike_espo
04-13-2004, 10:52 AM
Probably right. But I just hope the heavy MGs are modelled right.

"Fatte vede che ridemo!"http://www.flying-tigers.net/caccia%20WW%20II/g50.jpg

Dmitri9mm
04-13-2004, 10:53 AM
Sounds like some sort of one-hit-wonder, that thing with the Zero http://ubbxforums.ubi.com/images/smiley/34.gif
I think the problem is more the .50 cal and the likes being too weak than .303 being too strong.
I find it nearly impossible to hit anything when flying CR 42 because of the massively overdone muzzle flashes which blind me completely in that plane. Are you quite sure you actually hit that Spit with more than half a dozen rounds, I wouldn't be able to tell if it was me. http://ubbxforums.ubi.com/images/smiley/59.gif

http://ww2photo.mimerswell.com/air/italy/fiat/00305.jpg
>"Flatspin", what is that supposed to mean? This aircraft is entirely spin-proof!<

mike_espo
04-13-2004, 11:02 AM
No Definately hit him. You can see the pieces flying off him. On another server, got two kills flying the Falco. Actually had a real good day yesterday. But it takes up most of the ammo. CR.42 has alot.

Actually this could be historic. The Breda had a reputation of having lousy hitting power. Could be modelled accurately. But Ive seen other posts that the 50 cal is underdone.

BTW, shame on you for displaying the Falco as Crashed...lol.

"Fatte vede che ridemo!"http://www.flying-tigers.net/caccia%20WW%20II/g50.jpg

clint-ruin
04-13-2004, 11:52 AM
Some brief notes on network gaming:

Modern internet games since the time of Quakeworld typically use a method called Client Side Prediction to display any active entities in the game world.

Client Side Prediction is superior to non-predicted activity in that each client does not need to receive confirmation from the server that any event has taken place. The activity in the game world can be inferred from network frame to network frame based on what is considered near-certain or likely to happen as a result of a clients actions, without needing to get every single action transmitted back to it from the server

Non-predicted game: client action > client xmits frame > server > server xmits frame > client receives results of action.

Predicted game: client action > client displays action > client displays most likely result of action > client xmits frame > server > server xmits real results frame > client displays confirmation.

This is a big step forward in terms of network gaming. Without prediction, even something as simple as pulling the trigger and seeing the animation for that weapon firing may require confirmation from the server in order to display any result of pulling the trigger. For modem based games this could result in a delay of easily up to 1/2 to 1 full second between action and received reaction. Not very fun. Quake 1 and Quake 3 allow clients to manually disable prediction code and the results are not very entertaining except on the very very high speed [10mbps+] connections.

The drawback to network prediction is that the actions you see on your client are not necessarily the actions that the server sees, or that any other client sees. Because you are getting the results of your actions displayed as-they-happen, you may see yourself shooting an enemy plane and scoring multiple hits, when in fact the enemy plane was absolutely not in the 3d space where your bullets were fired. Actions such as a plane being de-winged or exploding require confirmation from the server in order to display that state change. Actions such as seeing your guns fire or seeing sparks on an enemy plane from your fire do not need to be confirmed by the server.

Just because you have seen hits on your end, does not mean that those hits actually occured. This means that occasionally you are going to pump ammo into a target beyond the point at which it has actually died [but the results not yet transmitted to you], too.

The way that a server determines what is valid data from what is invalid data can vary from game to game a great deal, but typically, the last several frames from both clients are stored and compared with each other in order to work out the results. Very few games since the original Peer to Peer model games [Descent/Descent 2 etc] accept every piece of data every client sends since the potential for cheating is far too great. A player could simply throttle their connection, leaving the target hanging in space and unable to move, shoot at it, then continue normal transmissions. Not good.

In the event of severe lag or packet loss, the client/server transmission has to start from an arbitrary 'zero' point and re-transmit the current client-visible state of the game world from scratch [ie send the last several frames as well as current updates], which can lead to even more problems if the connection continues to lag or lose packets.

If you are going to post to ORR about weapon damage effects in non-local online games, please get an ntrk both from your own client and the targets client of the event in question. They will not always show the same things happening in the same order.

http://users.bigpond.net.au/gwen/fb/leninkoba.jpg

|CoB|_Spectre
04-13-2004, 02:50 PM
Hmmm...it's hard for me to tell if IL-2/FB/AEP uses the prediction method nor not. One interesting aspect of being on voice comms is being able to audibly confirm that you and at least one other player witnessed the event simultaneously. On rare occasion I have seen players with low pings experience strange anomolies like having their aircraft collide with a vehicle crossing the runway when, to me, it was obviously right in front of them while they reported having no traffic on the runway at all. Even though I was the host and saw no visible reason for their aircraft's sudden destruction, it was what their computer experienced that mattered. This would suggest network prediction is being used. Sudden and unrequested disconnects produce some strange phenomina even when it is the host that loses connection. It is quite common for the clients to continue gameplay for several minutes. I have always marveled at the efficiency of the code in the IL-2 series and how well it works for a wide range of pings. As for gathering track files from multiple online players to use as persuasive evidence for changing online weapon damage, it is unlikely anyone will go to the trouble if there is no movement in the developer's position regarding demonstrated offline weapon damage. I am hopeful there will be some improvement to that.

WWMaxGunz
04-14-2004, 03:36 AM
The "Breda" in the CR.42 is, IIRC according to Gibbage, really an M2 50 cal and subject to extra wide dispersal. This has yet to be confirmed.


Neal

MatuDa
04-14-2004, 04:09 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by clint-ruin:
Some brief notes on network gaming:
*lots of stuff*
<HR></BLOCKQUOTE>

Basically correct and thanks of reminding me of qw, it was a huge thrill at the time http://ubbxforums.ubi.com/infopop/emoticons/icon_biggrin.gif

I don't think the prediction works exactly like clint explained. I think anyone can be killed in anyone elses client and the server is informed after that so in my view it goes the other way (clint said server determines who gets hit)

I have 10Mb unshared conn so when playing on a nearby server with a ping of 5ms or less I shouldn't be affected by lag if server decides if I am hit. Several times I have been on turning climb and watched my opponent, seeing that he cannot bring his guns to point me. Then he opens fire and shoots, his bullets missing me by 100meters behind me and then I suddenly miss a wing. http://ubbxforums.ubi.com/images/smiley/16x16_smiley-surprised.gif
In his view, because of his lag he sees me behind of my actual *server* position and can get a solution to shoot.

BTW a problem caused by the same thing can happen in headons: Many times I have dodged at the last second to avoid collision and 1.5 seconds after the merge my plane suddenly blows up for no visible reason. Because of my ping server has "known" I dodged and no collision happened but on the other playsers screen we collided and once his confirmation comes to server it gets sent to my client too and thus boom. http://ubbxforums.ubi.com/images/smiley/16x16_smiley-sad.gif

Clint was correct on the pushlatency-issue in FPS-games. FB just works a bit different.

WUAF_Co_Hero
04-14-2004, 04:17 AM
I had been under the impression (fueled by that of many others) that in IL-2, all hits shown clientside, are actually scored. Thus, when you see a plane warping but still manage to him him: when the enemy a/c comes back (if he does), he will have the damage, or be completely shot down. I've seen this happen many times, and it would further reinforce that theory.

Build a man a fire, keep him warm for a day...

Set a man on fire, keep him warm for the rest of his life.

clint-ruin
04-14-2004, 04:23 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by MatuDa:
I don't think the prediction works exactly like clint explained. I think anyone can be killed in anyone elses client and the server is informed after that so in my view it goes the other way (clint said server determines who gets hit)
<HR></BLOCKQUOTE>

As I said lower down, there's a few different ways the server can use to distinguish what's valid from invalid all with different appropriateness for different sorts of games [FPS is going to differ from flight sims and RTS games]. One of the easiest ways to do this is to do a quick check on angle / velocity / position of fire and allow any data that is not totally and obviously incorrect. There's a lot clients can do on their own and be far more efficient at than a server can, but for arbitration the server has a much better perspective on what has really happened at any point in time.

A lot of people have spent a lot of time reverse engineering [or just looking at source] for Quake/Quakeworld/Quake2/Quake 3s network layer so that's by far the most common source of info on how this stuff works. Other flightsim games that have been around a while - airwarrior/warbirds etc use a slightly different, almost MMORPG style system with massive server involvement and very low bandwidth requirements per object/client. But there's much less public info on how they work, or at least, that I have been able to find. Different approaches are needed when you can have objects with a combined approach speeds into thousands of kmh [two Go-229as fire Mk103s at each other], rather than players walking/running around firing subsonic rockets and hit/scan weapons at each other :&gt;

Since there's no Forgotten Battles:C or any ability for user modification, the release of a lot of this kind of info is probably just seen as a way to encourage cheaters/exploits. But it would be nice to get some concrete information on how the basics work in FB all the same.

http://users.bigpond.net.au/gwen/fb/leninkoba.jpg