Multiplayer for EVO has so much potential. Can someone tell me the reasons why it's not possible to make an EVO that could be played multiplayer?
-
Quote
Originally posted by Sportzen:
Multiplayer for EVO has so much potential. Can someone tell me the reasons why it's not possible to make an EVO that could be played multiplayer?Man if you don't know why it's not possible then there is no point trying to explain it, it would be over your head. Lets just say that it would take very good programers YEARS to do, though they'd make loads off it. Any who is trying to make EVMP is deceiving themselves. People have tried and in the end gave up and admitted how impossible it is.
Dragon
-
Well, it is possible, but the technical difficulties are immense. There are lots of other minor issues, too... Ambrosia's position is that it will likely never make EVMP.
------------------
"His Strangelove All-Purpose Do-It-Yourself Defensive First Second or Third Strike Indestructible Fantastic State-of-the-Art B-Ware Offensive Attack Bomber can't compare with our M & M E & A Sub-Supersonic Invisible and Noiseless Defensive Second-Strike Offensive Attack Bomber."
- from Closing Time , by Joseph Heller. -
I think they said that it would be really difficult because of the difficulties involving dates, hyperspace jump times, etc. Since a big ship takes 3 days to complete a jump, and a little one takes 1 day, if the two people playing multiplayer were to warp out of the system, they'd quickly get out of sync with each other, where one would be further into the future than the other. I'm sure there are other problems too, that's just all i can think of right now.
-
Quote
Originally posted by Johnny Bubonic:
I think they said that it would be really difficult because of the difficulties involving dates, hyperspace jump times, etc. Since a big ship takes 3 days to complete a jump, and a little one takes 1 day, if the two people playing multiplayer were to warp out of the system, they'd quickly get out of sync with each other, where one would be further into the future than the other. I'm sure there are other problems too, that's just all i can think of right now.Things like that are minor problems in the face of a network game. If it would be anything like we all want, there would be a bunch of servers running full time for people to play the game, but they will also have to program latency code, and tons of other stuff dealing with the networking part of the game. The scenario design, and actual specifics of the game would be easy compared to this. You can't just stick in network code in a few places and see if it works. You have to make an entirely new game. With the networking code being created right next to the game play code, they have to support each other. And oubviously you don't have to make it exactly like EV is. Because the time thing would be annoying, so they'd just make it real time, faster ships get to the next system faster, slower ships get there slower, longer routes take longer to get to, and shorter routes are faster.
Chamrin
-
And EV/O-MP isn't impossible, it is just stupid. What you want is a MP LIKE EV. EV is a 1 player game. To make an MP and limit its features is stupid, you could do SO much more.
It doesn't take years to do, just alot of people with alot of time. Not one guy who likes to act like he can when all he has done so far is made an app that crashes after 2 seconds.
The people that will do it are those that work together and spend their time writting it, not trying to impress people by saying they are.
Sportzen, no one has said it is impossible, it is VERY possible to do. More than you might think..
------------------
tear it down / AIM: G2uidehatr -
There is also the LAG problem. Imagine 50 people all flying around, SHOOTING and having homing missiles. The lag would be HUGE. Face it people it's not going to happen. Things would have to be simplified to the point that it sucks and is no longer like EV.
Dragon
I know of at least 3 different EVMP groups that though well organized and thought out(much better than the idiocy that is goin on in the developer board) and it STILL failed misserably. -
By the way, mikee, kberg has made seven more builds after the crashy one. They no longer crash, and have many new features. I suggest you download it, look at it.
And really, people, do you have something against this endeavor? Let's stop the disparaging, pessimistic comments. If it fails, it fails. Let's just wait and see, ok?
------------------
God. Root. What is difference?
-PitrAIM: obormot345
-
Quote
Originally posted by evo_dragon:
**
Dragon
I know of at least 3 different EVMP groups that though well organized and thought out(much better than the idiocy that is goin on in the developer board) and it STILL failed misserably.**Interesting how you can say that without ever seeing either my design documents or my code :).
I am interested in knowing about those other groups though. If you can tell me who they were and hopefully how to contact them it could prove to be quite helpfull.
------------------
-
How's that MP Engine work? I would be very curious to hear something! (I am not very adept at programming yet...in fact, I have done almost nothing in programming, just a few simple BASIC progs, but I am still curious.)
If it's not too hard, could you email or post with a few simple concepts...(if "simple" is even a correct word)!
Thanks!
------------------
From Left Hand Phoenix of the AWL
(url="http://"http://www.awl.cjb.net/")http://www.awl.cjb.net/(/url)
Halo, Homeworld, and Diablo II. -
Quote
Originally posted by evo_dragon:
**
There is also the LAG problem. Imagine 50 people all flying around, SHOOTING and having homing missiles. The lag would be HUGE. Face it people it's not going to happen. Things would have to be simplified to the point that it sucks and is no longer like EV.Dragon
I know of at least 3 different EVMP groups that though well organized and thought out(much better than the idiocy that is goin on in the developer board) and it STILL failed misserably.**Ever play Starcraft over Battle.net? Often times me and another Protoss human will have 20 carriers of 8 intercepters. That is 320 interceptors flying around like Mosquitos - all over the internet.... and it works just fine... no lag.
Pretty effects don't cause lag, it only takes one piece of data to tell your program to do some huge explosion or other effect.
------------------
tear it down / AIM: G2uidehatr -
How I've currently got it set out? I WAS planning on using net-sprocket for most of my networking needs, but after this years WWDC... Instead I'm going to go with Apples open-sourced openplay project. At least I've got all the source for that...
In order for me to explain how it works, I need to give you some background on the current implementation. Every game-play object has "critical" info in the form of a vector. For ships and weapons that vector is represented by component x and y speeds and an angle for the current direction it's facing, and for spobs (irrelevant) it's a speed and relative angle to it's parent spob. Each object also has non-critical info such as absolute x and y positions. They are represented with the following structure:
struct PosType
{
float xSpd, ySpd;
float angle;
double xPos, yPos;
... and a bunch of stuff related to screen redraws
}The critical info pretty much has to be updated every other frame, while all the non-critical info such as the absolute x and y positions, shield and armor levels, ect can all be calculated independantly from the critical data.
The general structure will take something of the following form: Every ship has a master ship, this ship serves as a control and is never used in the game. Every ship that is added to a system is a clone of one of these master ships with a complete critical/non-critical information update during the hyper-in. This (might) be the biggest problem, since if more then one ship jumps in at the same time, it might cause some lag but it will be brief. From that point on, only critical data must be transmitted. With a custom random number generator and a common randseed, I can remove the need to transmit any positional data for weapons and AI ships (since they are all handled by computer anyways). The ships are a bit harder. First of all, floats are rather neccesary for x and y component speeds in the game... I need the flexibility of fractional movements. I'll probably do something of the following: since the component x and y speeds are limited by the ships top speed, I'll multiply the float components by some scalar to get an integer (losing some precision, but that shouldn't really matter) which I probably will be able to cast to a short. If I can do that, then I reduce both my x and y speeds to 32 bits in total! If I put some more strict limits on turn speeds I'll be able to do something similar to the angle as well. My target is 48 bits for critical info, and another 16 bits is left for me to compress stuff like weapons firing, messaging, and other generic info. If I can get every ship down to 64bits then this will work pretty nicely.
56000 bps / (60 fps * 64bits) gives me basically 14 ships per system to have NO lag for some poor guy with only his 56k modem. That seems pretty acceptable, considering that 14 human controlled ships in a system is actually quite a lot. I figure that double that number would still result in acceptable gameplay (every ship updates 30 times a second), and as internet performance improves and people start getting cable modems and DSL connections then we could start having those huge net-battles everyone dreams of without using a LAN.
Also, are Chris Burns or Captain Scurvy still out there? I'd be interested in talking to them, as I just learned they were part of a rather large group of developers working on the same thing...
------------------(This message has been edited by kberg (edited 05-27-2000).)
-
Quote
Originally posted by kberg:
**Also, are Chris Burns or Captain Scurvy still out there? I'd be interested in talking to them, as I just learned they were part of a rather large group of developers working on the same thing...**
Did you bother to EMAIL them before starting some rumor? Burns announced a while back (last...fall?) that he was working on an EVMP using some Apple game sprockets I believe.
Captain Scurvy has his MP-RPG and was part of a large group called AEON which was developing software and DISCUSSED an EVMP if that is what you are talking about. Aeon is basically dead now.
------------------
tear it down / AIM: G2uidehatr(This message has been edited by Soviet mikee (edited 05-27-2000).)
-
Kberg,
insted of critism, here's a little help.
I haven't a clue how to reach Chris Burns, but CS can me e-mailed at Markaz2001@aol.com, or contacted over AIM. His screen name is Markaz2001.
CS should be able to tell you how to find Chris Burns.
------------------
-
Kberg, as UEC gave you all my info, come find me.
I really would like to talk to you over IM. Really.
-CS
-
All who are intrested in this EVO topic should check out the "Multiplayer possibilities and solutions" topic in the developers forum. We have fought our way thought the majority of the problems we have seen. kberg, a regular poster in that topic, has aparantly started coding allready. If you have any comments on this topic should post there.
------------------
Zitchas