M&B and Multiplayer Discussion

[Eluem]

What I am trying to say is that I think the whole debate has wrong focus. The problem is not getting you hands on the sourcecode, but the technical limitations in the internet and it's protocols. In addition to insane amount of information that must be sent to make this internet playable. You must remember that all characters, the NPC ones too, must have instantaneous status updates on the battlefield. But nothing would make me happier that if someone proved me wrong.

I understand and thought about everything you've stated in your post. The only reason I said that I would need the source code was because of what the original post in this thread was saying about how I should code it and test it myself which would be impossible with out access to source code....

I honestly don't know how the combat is coded but depending on the way it is done could change how you could use it on the internet. You don't need to use so many variables for the combat system... depending on how you do it you can make it so it's either more CPU intensive by using more collision boxes and a couple of other variables and then just sending a couple of single point locations with some other basic data for attacks and such. Also I was never suggesting to make the game be the the whole huge scale battles. Maybe it could just be a couple of players in an arena like people have suggested OR you could do two armies against each other but you may have to scale them down. I know that the internet has many limitations currently for a lot of reasons (especially the messy lay out). Also the idea of sending velocities and such is fine, that's what I would have planned to do too. Then every couple of frames you could also send everything's location to be sure that there are no issues (when sending velocities, depending on how the game is coded, the FPS of each individual computer may actually effect the out come slightly which would eventually cause issues). This could be done, I'm sure of it. Not saying that I could code it for I've never coded much on the side of networking except for one small test project I did a long time ago with my friend. It would take a lot of research to figure it all out for me.

 

[janlulhannes]

To make this easier to transport as the internet is crammed with torrents, we must pack light, I would use an initial position in three dimensions. I don't know how many coordinates that are possible in a M&B battlefield, but i can imagine that this would make for three rather large integers. To avoid having to report my position every millisecond to avoid stuttering, I'd only send my velocity changes. I would also have to report actions like mounting/dismounting and attacking/blocking.

So far so well, now for the bad part: to make sure that you got all my velocity updates this would need to be a tcp connection. tcp connections are slow. One could always write your own network stack, I've done it, but i wouldn't recommend it. One could possibly use some existing framework like Directplay, but I am not sure if you can dictate what transport protocol to use.

What I am trying to say is that I think the whole debate has wrong focus. The problem is not getting you hands on the sourcecode, but the technical limitations in the internet and it's protocols. In addition to insane amount of information that must be sent to make this internet playable. You must remember that all characters, the NPC ones too, must have instantaneous status updates on the battlefield. But nothing would make me happier that if someone proved me wrong.

JEWTUBE KILLED TEH INTERNETZ

 

[Orion]

JEWTUBE KILLED TEH INTERNETZ

And you just tried to kill a very interesting conversation.

Get out of here, seriously.

 

[kallek]

I honestly don't know how the combat is coded but depending on the way it is done could change how you could use it on the internet. You don't need to use so many variables for the combat system... depending on how you do it you can make it so it's either more CPU intensive by using more collision boxes and a couple of other variables and then just sending a couple of single point locations with some other basic data for attacks and such. Also I was never suggesting to make the game be the the whole huge scale battles. Maybe it could just be a couple of players in an arena like people have suggested OR you could do two armies against each other but you may have to scale them down. I know that the internet has many limitations currently for a lot of reasons (especially the messy lay out). Also the idea of sending velocities and such is fine, that's what I would have planned to do too. Then every couple of frames you could also send everything's location to be sure that there are no issues (when sending velocities, depending on how the game is coded, the FPS of each individual computer may actually effect the out come slightly which would eventually cause issues). This could be done, I'm sure of it. Not saying that I could code it for I've never coded much on the side of networking except for one small test project I did a long time ago with my friend. It would take a lot of research to figure it all out for me.

OK, I read you. You have some valid points. What made me think combat is made the way I described it is that an attack always has the same trajectory and using animation scopes would require less data transfer (as the same information is stored on all clients) than using colliders. Since an attempt at doing MP would have to mem-patch the .exe, this would be the easiest part.
I agree with you in that doing this in an extremely small scale would make sense, at least until one had some sort of proof of concept. Is it possible to turn on some kind of debug output in M&b. Showing the positions of objects? Where the collisions appear? This would make it easier to watch for changes in values in memory. This would make it possible, but still insanely time-consuming, to spot variable candidates using visual studio and it's brilliant debugger.

If this succeeds I would then continue with dumping the data stream to a file just to get some idea of the amounts of data we are dealing with here.

As previously stated, I believe that the numbers will be to huge to make i feasible. One might, as you say be able to make some kind of "duel" proof of concept code, but this would also involve reverse engineering M&b (memory patch), something which might not be viewed upon too kindly. However if you think you can make it work, I am not here to stop you. I do not have the time to contribute with any code, but I might be able to contribute with ideas and pointers if you get stuck.

Good luck 

 

[Cergic]

First post 

I refuse to scroll through 108 pages of discussion when it obviousley contains 50% craptalk (or more)  , but i assume that a multiplayer mod can be done?
*note*
I am aware of the fact that it would require a ****load of work, & i honestly think that everyone should wait for M&B2 instead of messing around in M&B1.

However, i would love some multiplay in M&B1

 

[Wu-long]

First post 

I refuse to scroll through 108 pages of discussion when it obviousley contains 50% craptalk (or more)  , but i assume that a multiplayer mod can be done?
*note*
I am aware of the fact that it would require a ****load of work, & i honestly think that everyone should wait for M&B2 instead of messing around in M&B1.

However, i would love some multiplay in M&B1

Armagan said it may be possible for M&B 2 just not 1 yet

 

[italiano]

First post 

I refuse to scroll through 108 pages of discussion when it obviousley contains 50% craptalk (or more)  , but i assume that a multiplayer mod can be done?
*note*
I am aware of the fact that it would require a ****load of work, & i honestly think that everyone should wait for M&B2 instead of messing around in M&B1.

However, i would love some multiplay in M&B1

That is basically what most of the "crap talk" contains - either suggesting Armagan implement it, or requesting a modder to do it.

 

[Eisenhouwer]

First post 

I refuse to scroll through 108 pages of discussion when it obviousley contains 50% craptalk (or more)  , but i assume that a multiplayer mod can be done?
*note*
I am aware of the fact that it would require a ****load of work, & i honestly think that everyone should wait for M&B2 instead of messing around in M&B1.

However, i would love some multiplay in M&B1

That is basically what most of the "crap talk" contains - either suggesting Armagan implement it, or requesting a modder to do it.

Touché

 

[Flatline]

*returns from long M&B-slumber*

Haven't found anything about this on the forums:
Swedish publisher Paradox says M&B will have multiplayer this summer/autumn.
Article in swedish, but main points are:

*Focus on battles, not campaign
*Aiming for 32 players, maybe more, and bots.
*closed Beta begins in 3 months
*modes of play are standard deathmatch/team deathmatch, siege and more


Article here:
http://www.fz.se/nyhetsarkiv/?id=31031

 

[Dryvus]

Can anyone translate and verify this? Nothing against you personally, Flatline, I'd be suspicious of anybody posting this.

 

[arbites_augmented]

Well, if so, I CHALLENGE YOU TO A DUEL, DRYVUS!

*Slaps face with glove*

 

[Dryvus]

*Ignores, shoves out of way on return from Paradox Forums*

Holy ****, seems real.

http://forum.paradoxplaza.com/forum/showthread.php?t=391662

 

[Adorno]

Oh no, now I'll get my ass kicked 

 

[Shik]

Now that multiplayer is officially confirmed, it's time for serious multiplayer suggestions.

A command line that can be accessed with a certain key - to give players and server masters abilities to do assorted commands like calling a vote to change to the next map, or to kick a certain player from the forum would be VERY much appreciated.

 

[Guan_Chung]

kick a certain player from the forum

I support this!

 

[Archonsod]

Given the amount of crap in this thread, I think it's better to start a new one. Over in the court.