Multiplayer Ideas

[red knight]

several of things lost in my mind for your new UFO

what about napalm bombing crash sites? :uzzi: only the first player at the place can do it

:aliencool:  what about human fractions against the aliens? it could be story for 2players humans vs 1 (or 2) computer players at aliens side

:alienmad:  what about Diablo/Diablo2 lvl up/promotion system? you can choose at "start" of soldier in what is he good - but you have limited "points"/"percent" (10% = 1 point)|then if lvl up - you can upgrade skills/move/etc.

:alientalk:   what about "civilian mobilization"? - you have several secret agents in towns/or something - they're mobilized in time of alien attack/base raid/ufo crash......... more people to kill them!

:alienlol:         what about on alien fractions (1st can be good in psi - but weak.etc.)? they will fight against each other for earth assault contract in last phase - like ROBO RUMBLE

Most of this suggestions had been made before.... Some are interesting like lvl promotion, but i will give you a little excercise... Think what makes a game FUN?... I will give you a hint. Most of us wants to kick alliens b.u.t.t.s but everything that helps you is making the game easier and not fun on the long run... you maybe want to include bombarding, but killing aliens in that way is not rewarding at all (not to say FUN)...

 

:uzzi:        what about fighting with each other of X-COM

organisations..........?

So, Are you asking for another Diablo (put your prefered frag feast game here) Frag Feast game? Today thread in game development (except on certain genres) are looking for way to avoid it... take a look at US Army, Most MMRPG games, and check on www.gamasutra.com on the game design articles...

 

Greetings

Red Knight

[0rganism]

For what it's worth, here's my take on RT (Geoscape) vs. TB (Battlescape).

 

The one thing that seems to be a consensus is that the Geoscape should be running nearly all the time for everyone. No definite decision on whether it should be massive multiplayer (i.e., a hundred clients hooked to a central server) or mini multiplayer (i.e., a handful of peer connections, perhaps over a LAN), except that resources for the former and interest in the latter may be lacking.

 

I think the most reasonable way to handle it is put a hard cap of 8 players on the game, make it localized, and save state between sessions. Persistent MMOGs, while phenomenally cool, are inherently unfair to those who have less free time -- or "real lives" of any sort. Wasn't there someone who literally died of malnutrition while playing Evercrack a couple years ago? Now THAT has to be a textbook example of misplaced dedication! There are ways to structure short, non-persistent MMOGs, that might be just as fun tho.

 

If I were going to make a persistent MMOG, I'd accept that some stuff has to go. First, as has been noted, the Geoscape and Battlescape are inherently different creatures. There's a good possibility to have varieties of either one without the other. For instance, suppose we have a decent auto-resolution algorithm for combats where battles take place non-interactively? Then you've got your real-time Geoscape running with minimal interruption. Turn based combat can be extra-fun, too, so maybe we want to put that in separately as an "arena option" that doesn't affect the Geoscape world. Or maybe we do away with the Geoscape time altogether, and just have open tactical ops sessions and base management.

 

If I were going to try to have both the Geoscape and the Battlescape, then there's a huge problem with scaling which I see taking shape in this very thread. I can think of tactical battles that have taken well over two hours of Real Life, so simply pausing the Geoscape for every battle is clearly not an option. Likewise, considering that a Geoscape fast-forward month can easily pass in two minutes, if one does away with interrupts for management, UFO shootdowns and other missions, just letting the two run concurrently is going to seem ridiculous. "It took me a year of gametime for that tiny raid? Wow!"

 

We all speed up the Geoscape while waiting for stuff to happen. What kinds of stuff? Research, recruitment, UFO interaction, facilities building, etcetera. Strangely enough, the temporal disconnect is not so much in the turn based portion of the game, where a tactical operation that would realistically take under ten minutes is stretched out into an hour's worth of distributed decision making, but in the Geoscape where time distorts by orders of magnitude. Let's take a typical UFO shootdown and salvage event, then which progresses something like this:

Event                    Gametime (Geoscape)      Realtime 
--------------           -------------------      --------
UFO detected             5:00 AM                  9:00 PM
Interception ordered     5:05 AM                  9:01 PM
Inter/UFO chase          5:05-10:00 AM            9:01-9:02 PM
"dogfight"               10:00-10:01 AM           9:02-9:03 PM
UFO salvage ordered      10:01 AM                 9:03 PM
Salvage team transit     10:01 AM - 3:00 PM       9:03 PM - 9:04 PM
Begin Mission?           3:00 PM                  9:04 PM
Battlescape              3:00 - 3:30 PM           9:05 PM - 9:55 PM
Score screen             3:30 PM                  9:55 PM - 9:56 PM
Salvage team returns     3:31 - 8:30 PM           9:56 PM - 9:57 PM

Off the cuff, the only reasonable way I see to synch the two is have a player or players involved in a given interception become incommunicado with the rest of the players until tactical resolution, using the implicit transit time as padding. The geoscape would have to progress at a constant 12:1 ratio to make this work. That also means it takes 3 days real time to build your hyperwave decoder.

 

So all this really says is the original storyline and game features are ill-suited to direct conversion to a multiplayer game.

 

I like some of the alternate scenarios, though. Take the post-alien victory storyline, where your job is to compete and cooperate with other fragmented X-Com bases to win back governments to the human side while funding your team instead of the other players. This way, there are always alien ships flying around, there are always aliens terrorizing the population, there are multiple alien bases on each continent to eradicate, and there's no need to do research. Your base is already built. Manufacturing, funding, and recruiting/recovering troops are the only geoscape-level events. Communication with other players is necessarily covert, and the time lag between a UFO shootdown/salvage the outcome becoming common knowledge would be easily explainable this way. If your gang liberates Chicago from the grips of alien opressors, maybe you get some additional funding from the residents of Illinois and game points. On the other hand, maybe the Mutons offer another player an opportunity to take Chicago back for them, in return for a gift of technology or cash and support of some alien troops for that mission. Could get veeeeery interesting.

 

And the Xenocide storyline could involve internal conflicts between the different species of aliens; perhaps the Sectoids wish to throw off the yoke of their overlords as well? You might just find yourself with some alien troops on your roster of mercenary resistance fighters! Or maybe you frame the Floaters or Ethereals in an attack on a Snakeman base by disguising your flying-suit troops, to instigate a race war between the two. Lots of possibilities for a planetary resistance storyline, with or without multiplayer.

[red knight]

Organism: I had been arguing about the time frame incompatibility long ago, for me there is not reconciliation for time issues like flying half the globe to get an alien... in a 5 seconds or maybe 10 seconds time frame in geoscape it would take a lot of time to get to destination... time you will be sitting around doing nothing... but as everybody wants to do multiplayer (i was the only one against it) i had to accept. This is a democracy after all.

 

Note: A battlescape only multiplayer its posible, and im not against it..

 

Greetings

Red Knight

[gangsta]

Organism: I had been arguing about the time frame incompatibility long ago, for me there is not reconciliation for time issues like flying half the globe to get an alien... in a 5 seconds or maybe 10 seconds time frame in geoscape it would take a lot of time to get to destination... time you will be sitting around doing nothing... but as everybody wants to do multiplayer (i was the only one against it) i had to accept. This is a democracy after all.

 

Note: A battlescape only multiplayer its posible, and im not against it..

 

Greetings

Red Knight

This is quite simply solved. Realism can be sacraficed in the multiplayer game. the fun factor is more important here. So the solution when you remove time controls in multiplayer is to make the skyranger move at unrealistic speeds to get to it's target. in fact we can have it speed up and slow down like someone in x-com would do with the time controls to get to it's target. It is not important if we are breaking any laws of physics here. People probably wouldn't notice anyway and just would be happy how cool the multiplayer game is. BTW good point and I'm gonna bring it up tonight.

[0rganism]

> A battlescape only multiplayer its posible, and im not against it..

 

Since you already have the Geoscape functional, why not go with a Geoscape-only multiplayer for proof-of-concept? As one of the enhancement requests has been for a tactical autoresolver to speed up repetitive missions, you can nail two floaters with one grenade that way. It'll also be a good platform to test out modified multiplayer storylines, if the mechanics work smoothly. It doesn't solve the underlying timescale problem of combining tactical battles and the Geoscape time warp, but it could make for some interesting strategy in its own right.

[gangsta]

BTW RK, what other things make you think that the multiplayer geoscope is not doable? I never thought about that skyranger thing but we might be able to find a solution for each problem. I remeber having same kinds of objects once against Warlords way of doing the battlescape but now there are solutions to make it work.

[red knight]

BTW RK, what other things make you think that the multiplayer geoscope is not doable?  I never thought about that skyranger thing but we might be able to find a solution for each problem.  I remeber having same kinds of objects once against Warlords way of doing the battlescape but now there are solutions to make it work.

You can alter skyranger velocity without problems, the issue is with the interceptor... i think i had posted it before... You can accelerate the interceptor, but you have to accelerate the UFOs too, causing that the same velocity difference. With the option of accelerating the interceptor you get a faster machine, but for gameplay issues you have to accelerate the UFO causing a similar DeltaT. So interceptor can cross the world in less time, but in the same time UFO is on the other side... If you dont accelerate UFO you always can catch the UFO, what its not true cause your technology is far below theirs...

The issue of the time taken by a battlescape mission and a geoscape is another... For example you end up using 5 minutes in battlescape to finish a mission... in Geoscape time that would be like 2 days... so your soldiers are moving like a 1 cm / minute in battlescape... i say we can bend a little time physics but that is too much.

 

Greetings

Red Knight

[gangsta]

We talked about this last night and I'm too tired right now to post a summary of everything we talked about. But let me get to that issue. First off we are willing to sacrafice any reality to make the multiplayer game playable and fun. So technology is not an issue. But with many of the things in this project I think here is a case where abstract thinking can help. Now in the single player version of x-com people associate how fast things move on the geoscape with time. The thing is as programmers we should think of those things as speed objects move on the geoscope and not time. In singleplayer mode the time control just changes the speeds of how things travel on the geoscope. Now in multiplayer mode we discussed that when an interceptor is chasing a ufo while everything else might be moving slow on the geoscape both the interceptor and the ufo can slowly speed up to the proper speeds while the chase is going on and then slow back down once the interceptor is about to intercept. Further for when we want more than one interceptor chasing the ufo we can let the user have a hyper button and the chase won't speed up until he presses it. Now forget about what technology aliens and humans have, we throw that out the window to make the multiplayer game enjoyable in this case.

[red knight]

Good for you... do you consider what is happening to other players as well, they have to wait cause someone (maybe they dont even know is there) is chasing a UFO nearby??? how about time passing faster from time to time without some player agreement?... Let alone the problems in transactions (if it is fair to call them like that) because of asynchronous events caused by different times per player. Just a glimpse of the problems involved... i had been thinking a lot of the issue, cause after all i was forced to think in that options... and for every solution i find there is a lot of added problems. EDIT: In that case you can only talk about a "quasi multiplayer" because nobody can coordinate actions with others... cause of timestamp transactions involded... you end up with a 4 single player instances game stored and managed by a unique server, no more than that.

 

Make time passing faster for alien and intersection crafts in there... well maybe some other player had tried to send a craft too to the same alien... but guess what... you acelerate both crafts, and the poor guy that missed 10 seconds because was doing base management couldnt send an interceptor to a UFO that has targetted his base (and accelerated by other player). That just plain unfair...

 

As you can see there are a lot of design challenges ( we can call them "design imposibles") that prevent you from being fair to players. Just take a minute to think about it a minute... from a multiplayer game mechanics point of view.

 

Greetings

Red Knight

[0rganism]

Yeah, I have to agree with red knight on this one. I hadn't even thought of it from the angle of Geoscape variable speed unsynching players or unbalancing the game before; you'd have to simplify base management to the level of RTS to have it make sense. And speaking of RTS, who hasn't had the pleasure of playing Starcraft on Battlenet against a speedjacking Zerg rusher? Cos that's basically what is proposed. The only way to make it semi-fair is have each player set an optimum speed, where the geoscape progresses at the slowest setting. Mix that with combat autoresolution, and maybe there's something workable hiding underneath. Base building would be done by simple pulldown menu, no fancy building configurations or equipment planning, scrap all tactical combat, assume all research is either very fast or already done, and just focus on the highest strategic layer where players race to intercept and salvage UFOs. Maybe the game pauses once per month for diplomacy and bathroom breaks?

 

It could be sort of a middle ground between Frisbee-500 and Risk. On the other hand...

 

Think back to the stupidest D&D session you can remember, where there were half a dozen pimply hyper pre-teens gassed up on Pepsi and Cheetos, trying to assert their immature fantasy plumage over one-another to impress a half-competent power-tripping DM hiding behind "the screen" rolling dice. Half the time is spent arguing over some fine point of rules minutae or waiting for someone else to finish up their saving throws. That dark vision is the lowest common denominator for multiplayer gaming, as practiced in the 9th circle of heck and junior high schools everywhere. I think it can be avoided, with a little care.

[gangsta]

Good for you... do you consider what is happening to other players as well, they have to wait cause someone (maybe they dont even know is there) is chasing a UFO nearby??? how about time passing faster from time to time without some player agreement?... Let alone the problems in transactions (if it is fair to call them like that) because of asynchronous events caused by different times per player. Just a glimpse of the problems involved... i had been thinking a lot of the issue, cause after all i was forced to think in that options... and for every solution i find there is a lot of added problems. EDIT: In that case you can only talk about a "quasi multiplayer" because nobody can coordinate actions with others... cause of timestamp transactions involded... you end up with a 4 single player instances game stored and managed by a unique server, no more than that.

 

Make time passing faster for alien and intersection crafts in there... well maybe some other player had tried to send a craft too to the same alien... but guess what... you acelerate both crafts, and the poor guy that missed 10 seconds because was doing base management couldnt send an interceptor to a UFO that has targetted his base (and accelerated by other player). That just plain unfair...

 

As you can see there are a lot of design challenges ( we can call them "design imposibles") that prevent you from being fair to players. Just take a minute to think about it a minute... from a multiplayer game mechanics point of view.

 

Greetings

Red Knight

 

Well I think it already has been posted before that in the multiplayer method peter, warlord and me came up with that there are no time controls and time runs at the same speed for everybody. In fact, as far as the multiplayer game is concerned time can be the local time of the server.. In the previous message about the ufo and interceptor in the chase mode I never talked about time. What I said before was that we the programmers should think abstractly here and disassoiate speed from time. How fast things travel in both single player and multiplayer is really speed and not time programming wise. In single player we also add a value to how fast 'time' increases. In Multiplayer mode we don't do that. Anyway to keep things simple I'd just keep everybody's time in multiplayer mode sync to the real local time on the server. Time isn't as important in multiplayer.

[gangsta]

Yeah, I have to agree with red knight on this one. I hadn't even thought of it from the angle of Geoscape variable speed unsynching players or unbalancing the game before; you'd have to simplify base management to the level of RTS to have it make sense.  And speaking of RTS, who hasn't had the pleasure of playing Starcraft on Battlenet against a speedjacking Zerg rusher?  Cos that's basically what is proposed.  The only way to make it semi-fair is have each player set an optimum speed, where the geoscape progresses at the slowest setting.  Mix that with combat autoresolution, and maybe there's something workable hiding underneath.  Base building would be done by simple pulldown menu, no fancy building configurations or equipment planning, scrap all tactical combat, assume all research is either very fast or already done, and just focus on the highest strategic layer where players race to intercept and salvage UFOs.  Maybe the game pauses once per month for diplomacy and bathroom breaks?

 

It could be sort of a middle ground between Frisbee-500 and Risk.  On the other hand...

 

Think back to the stupidest D&D session you can remember, where there were half a dozen pimply hyper pre-teens gassed up on Pepsi and Cheetos, trying to assert their immature fantasy plumage over one-another to impress a half-competent power-tripping DM hiding behind "the screen" rolling dice. Half the time is spent arguing over some fine point of rules minutae or waiting for someone else to finish up their saving throws.  That dark vision is the lowest common denominator for multiplayer gaming, as practiced in the 9th circle of heck and junior high schools everywhere.  I think it can be avoided, with a little care.

well I'm just talking about ufo chases speed wise not time wise. The time moves kind of slow in the multiplayer game because those games can last months running nonstop on the dedicated server. The idea is to form teams in the multiplayer game so people could be playing the game 24/7. perhaps the time should be accelerated slightly from realtime but not that much. Multiplayer is a totally diffrent beast that has diffrent geoscope rules. Also we had previously discussed doing a LORD (Legend of Red Dragon) type winning system where someone can win the game (fight the big brain) and then the game starts over for that person and the rest of the people continue playing and another team has a chance to do the same. Anyway, people play muds for months and years and I think this should be the same kind of thing.

[gangsta]

ok don't know if I made it clear in my above posts but time in multiplayer should be constant and not increase or decrease in speed for anybody. Everybody has the exact same time.

[red knight]

Yeah, I have to agree with red knight on this one. I hadn't even thought of it from the angle of Geoscape variable speed unsynching players or unbalancing the game before; you'd have to simplify base management to the level of RTS to have it make sense.  And speaking of RTS, who hasn't had the pleasure of playing Starcraft on Battlenet against a speedjacking Zerg rusher?  Cos that's basically what is proposed.  The only way to make it semi-fair is have each player set an optimum speed, where the geoscape progresses at the slowest setting.  Mix that with combat autoresolution, and maybe there's something workable hiding underneath.  Base building would be done by simple pulldown menu, no fancy building configurations or equipment planning, scrap all tactical combat, assume all research is either very fast or already done, and just focus on the highest strategic layer where players race to intercept and salvage UFOs.  Maybe the game pauses once per month for diplomacy and bathroom breaks?

 

It could be sort of a middle ground between Frisbee-500 and Risk.  On the other hand...

 

Think back to the stupidest D&D session you can remember, where there were half a dozen pimply hyper pre-teens gassed up on Pepsi and Cheetos, trying to assert their immature fantasy plumage over one-another to impress a half-competent power-tripping DM hiding behind "the screen" rolling dice. Half the time is spent arguing over some fine point of rules minutae or waiting for someone else to finish up their saving throws.  That dark vision is the lowest common denominator for multiplayer gaming, as practiced in the 9th circle of heck and junior high schools everywhere.  I think it can be avoided, with a little care.

well I'm just talking about ufo chases speed wise not time wise. The time moves kind of slow in the multiplayer game because those games can last months running nonstop on the dedicated server. The idea is to form teams in the multiplayer game so people could be playing the game 24/7. perhaps the time should be accelerated slightly from realtime but not that much. Multiplayer is a totally diffrent beast that has diffrent geoscope rules. Also we had previously discussed doing a LORD (Legend of Red Dragon) type winning system where someone can win the game (fight the big brain) and then the game starts over for that person and the rest of the people continue playing and another team has a chance to do the same. Anyway, people play muds for months and years and I think this should be the same kind of thing.

Changing the UFO Chase speed has the same problems stated in a post before... think for example...

 

A UFO controlled by a person or not... is moving at a constant speed of 1000 Km/h... Your interceptor depart from the other side of the globe at 1000 Km/h. Normally it takes 12 geoscape hours to get him (im aprox)... well using your solution... You can either:

 

State that the Interceptor is moving 4x faster meaning its absolute speed is 4000 Km/h taking 12 / 4 hours = 3 geoscape hours to get there... now you have two alternatives acelerate the UFO 4x too or not... We will discusse it one by one...

 

Not accelerate the UFO: You move at 4000 Km/h vs 1000 Km/h outcome you always catch the UFO, even theres not even a slight posibility to do so in a normal chasing...

 

Accelerate the UFO: You move at 4000 Km/h vs 4000Km/h. Your DeltaT is equal as in 1x movement... so supose both aircrafts are in the equator, then both are moving along the equator in the same direction... the time needed to get the UFO is the same in 1x, 2x, ... , nx cause the deltaT is not changing...

 

So its the same.

 

Greetings

Red Knight

[gangsta]

I'm for accelerating both at the same time. But I see what you mean about increasing the speeds for both. I think this is something we still can solve on the server side. We just will put the things in hypermode in a temporary diffrent time frame on the server. The people on the userside still will get the time stamp from the real time frame of the game. If that is not possible then having the chaser speed up on the prey will work too. It is not important if this is unrealistic in reality what is important is that it will make the game work. also in this case we can have a button to go into that hypermode to give the hunter more time to get more people involved in the chase.

[gangsta]

as for the method of only speeding up the chaser to get to the ufo. we can disable hypermode when the ufo is too fast for the interceptor like the battleship

[Maverick]

this is a thought out of the blue that may have been considered before, but I wanted to bring it up just in case. What if the player didn't have control of the clock in multiplayer, but the computer did -- and it was turn based (with a time limit of course) when events happen. Say it gives you a minute (or 2) to buy/manage the base (there is only so much you can do -- but that time limit could be decided on by the players when the game is created; newbies could set it for 5 minutes to build a base while pro's will create with 60 seconds of base management) and an additional 30 seconds for orders (ie interceptor attack ufo-102) This way, everyone gets the same amount of time (or as much time as they need) to deal with the stuff in the geoscape and then the computer accelerates time until the orders that were just assigned are carried out (ie interceptor downs ufo-102), time speeds up again and the interceptor returns to base and you go through the same timed turns again. (thinking about this, it would probably be pretty difficult to program -- but I leave it up to you guys to tell me what can/can't be done) This would also provide a COMPLETELY different game experience than in single player. You could be an expert at single player and not know the first thing about playing multiplayer (which would add more life to the game -- something new to conquer) And finally, I was wondering how the timing is run in the game. (the problem most common with the original game is that it runs too fast, i don't know how they programmed the time settings, but on faster systems they make the game unplayable) Knowing that this was an issue for the older game, are we taking steps to prevent the same thing from happening to our version? (I mean if we used a similar clocking mechanism, we might be prone to the same problem -- effectively reducing the life of our own creation to the days when the technology to use it is still common)

[gangsta]

The multiplayer game is more of a team game where you can have diffrent people on your team play at diffrent times considering the internet provides international users. The multiplayer game can take a long time to beat. To make up for the lack of time controls there are more ufos zooming around for you to shoot down and you are not penalized for not shooting them down. We have to discuss more of the details of how to implement multiplayer and it is mostly geoscope rules that are effected. Anyway I do believe multiplayer is doable and we can come up for a solution for any problem we come across.

 

As for the last thing you mention about X-Com 1 and timing well that was just a case of bad programming. All games should run at a set speed no matter how fast the computer they are running on. I really don't know why X-Com didn't because at that time that was already a well known problem and delt with. Perhaps the people who made X-Com were better designers than programmers.

[Maverick]

Then along the lines you've suggested we should try to have servers in different areas to facilitate such a 24 hour game -- as many servers as possible, which wouldn't be much (similar to the way Diablo II runs); also they would have to be capable of loading previously played games into the server to start with -- nothing would turn me off from multiplayer more than having to restart a game I'd been playing for 24 hours because the server crashed. Multiple servers would be required because #1 it would be just plain annoying to have to try and play the same time someone else on my "team" is playing, and #2 it would str8 piss me the F$*% off if I was playing and the host quit. Is this our current model? (I know that I could run one such server on a 1.3 ghz machine on a 1.5 mb/sec line, and given the time frame for the game at least one other on a similar bandwidth line). Should this type of server-client model be adopted, I suggest that the server only be in charge of data (#'s and so forth). HLTV (for half life) gives each player a variable and records the actions taken by each player at a specific time. In that way it can re-create the game from any perspective and takes much less space than video information. Similarly, this server-client model could be designed to store data changes and leave graphics work to the client computers. Also, a potential difficulty with this is that it would limit the effective range that players could interact over (FE might have a hard time joining a server I hosted, as I would have difficulty reaching a server in Estonia) not to mention server variations (if I have the greatest server in the land [hahaha - doubtful] and everyone wants to join my server it lags the heck out of it having people all over the planet trying to play on it. And finally, with this model the difficulty becomes one of abnormal usage, regardless of how many servers we would have available, if we produced anywhere near the result of HL or D2 we could reasonably expect more than 20,000 people to be playing at any given time. The server software would have to be included with the game and relatively easy to set up so that people could create and manage their own private server boxes. *sighs* The application of multiplayer presents difficulties that may be greater than the design structure and coding themselves.

 

I think that D2 should almost be our model for this type of multiplayer game. We don't have the same vested interest in ensuring people don't hack (so we don't need to go quite as extreme as storing every gamers file details in servers we control) What we could do, however, is use a waypoint system (based on some form of authentication system stored on their computer...if we used encryption on a log file of the progress a player has achieved with a character, that string could be called up at the start of the game, allowing for them to start a multiplayer game that many others are in at a much later date -- two computers don't necessarily have to have the same date displayed.) It would be possible then to have someone with a base that has avengers right next door to someone who has a base still using interceptors. Technology within the missions could vary as well -- however i contend from this point on that mind control be useless against other players agents; we need to develop a model now (assuming multiple players can have agents in the same missions) for how to deal with damage players can deliver to each other's troops (It would be very bad if I joined a mission and had my skyranger deck nuked with a blaster bomb by a player who thought it would be funny to screw my score)

[gangsta]

Dunno about writing a massive multiplayer engine. I was thinking of writing a multiplayer game that lets people from 1 to 20 (maybe 50) play at a time. I was thinking about someone running the game on a dedicated server with high bandwidth. I don't even want to think about multiple servers as a programmer right now. I'm sure RK doesn't either since the single player game is more important to him. I think at 20 could be done on one server without problem perhaps more. no way to tell until we have a dedicated high bandwidth server and the number is changeable via code anyway.

[Maverick]

When testing for that aspect of coding begins remember that I will be able to provide one such server that I can set up for complete remote use by the programmers -- allowing you to update, install, modify existing installed programs (and install new ones if need be) to run a dedicated server on a 1.5 mb/sec line (which is sufficient for most other games, i don't see why ours would be different).

[gangsta]

That's cool. btw any idea what the max number of players for most games is on a 1.5 mbit line?

[Maverick]

For HL that will support about 18-20 effectively if there is not high traffic (regardless of the dedication of the line itself), and there are no modems playing on it. I have not tested it in an environment like BF1942 (support up to 64 players). One more contributing factor is distance from player to server.

[gangsta]

well maybe we can have a main server connected to several high speed relays. I don't know how that would effect things. but doing a system where there is shared processing is hard. Anyway 18-20 players sounds fine

[Xcen]

Hmm I've been reading this topic on and off so it's hard for me to know what has already been discussed (we should summarize the most active topics at least once a month or something...I don't know). Anywayz my point is, the more I read this, the more it sounds like you guys want to make a MMORPG out of the multiplayer game? Am I completly lost? Seems like a project we would need to leave out until version 5 or maybe even more hehee...

 

 

Is there some1 that can sum up what is the vision currently discussed? Sorry I should've read the thread but a lot of stuff is going on and it can be hard to follow

[red knight]

Given the synchronization problems found in the geoscape, i really wont like jump into the issue until battlescape multiplayer, and general game is done... not to mention find a consistent way to do geoscape multiplayer...

 

Greetings

Red Knight

[gangsta]

Hmm I've been reading this topic on and off so it's hard for me to know what has already been discussed (we should summarize the most active topics at least once a month or something...I don't know). Anywayz my point is, the more I read this, the more it sounds like you guys want to make a MMORPG out of the multiplayer game? Am I completly lost?  Seems like a project we would need to leave out until version 5 or maybe even more hehee...

 

 

Is there some1 that can sum up what is the vision currently discussed? Sorry I should've read the thread but a lot of stuff is going on and it can be hard to follow

20 people is hardly massive . That probably can all be done on one server.

[gangsta]

Given the synchronization problems found in the geoscape, i really wont like jump into the issue until battlescape multiplayer, and general game is done... not to mention find a consistent way to go geoscape multiplayer...

 

Greetings

Red Knight

Well like I said the single player game would be given priority. As for the geosape I think that synchronization problems can be avoid and you would do that in steps anyway. At first multiplayer would have no time controls and have ufo's and skyrangers move at regular speeds. (the geoscope part doesn't have to be fully playable till later) later on you would let the hunter have unrealistic speeds to catch up to the prey. Finally you can try putting hunter and prey into a diffrent temporary time frame and then place them back into the proper time frame.

[Maverick]

RK/Gangsta -- can either of you give me the run-down on what the time problem is with interceptions etc? I understand the concept of needing faster speed when you launch an interceptor etc -- but is there a coding difficulty to making that happen reliably syncronized? I understand that the player does not have control of time and that each computer cannot run at its own accelerated speed in a multiplayer game -- perhaps there is the possibilty of staggered time (ie it switches between speeds at a constant interval). Final suggestion -- design it for fewer players and make time controls a group choice. If 8 people were playing and all of them clicked "go to next event" button, they could all fast-forward until the next action event for the nearest event on ONE computer and then slow down. (for example if I launch an interceptor at a UFO 800 miles away and someone launches one at the same time at a UFO 600 miles away and we click "go to next event" it will slow us both down when his interceptor gets there before mine).

 

(and I do think it would be a good idea to use an encrypted database file stored on the users computer for their "saved" multiplayer games so that they can't hack it -- I can expand on this later if necessary. One of the things I know a lot about for programming is how encryption works, and it would be relatively simple to implement -- the client need only an IP and decryption code, though because it is available internationally we will only be able to implement "weak" encryption. [still worth it!!!])

[gangsta]

Well I don't see any problems syncronizing time in the multiplayer game. 1 method is to make the hunter have unrealistic speed to catch the prey. This method time stays the same for everbody and speed of the hunter increases. the other method is thinking of virtual time fields that the hunter and the prey are in. That way both speed up and down as if there was someone using time controls on only them. This can also be done by keeping the time the same for everybody. all you would do in that case is multiply time by another variable like 2 (that gives you 2t) to make it seem like time is going by twice as fast for those two objects while all other objects movements are only effected by t. That way you have the effect of having hunter and prey seem like they are in a diffrent time frame when they really are not. anyways this is assuming that movement is a function of time. If you multiply time by a factor there is no need to resyncronize time when something jumps out of the hyper time frame back into the normal time frame.

[Maverick]

ok -- well, i was just wondering because it seems to have been a large concern until now. The other thing is that the UFO should not appear in double speed; it should be there in the slowest speed, but when someone launches an interceptor time speeds up. (consider that this method will add a coordination requirement between people playing at the same time)

[gangsta]

time in multiplayer never speeds up. time always goes at a constant pace. instead, we'll make it seem like an interceptor and ufo in a chase scene are in a diffrent time frame. However, that is just a special effect and in reality in the movement equation for each we are just multiplying the time variable with another variable to make it seem like time is going faster for each those objects. That is the best way I can explain it in plain english. in an equation with 2t time seems to be going twice as fast than in an equation that just has t. That is the math behind it. In both those equation the time is the same value. So there are no problems of syncronization here that I see.

[gangsta]

ok here is a old brainstorming IRC log about multiplayer Peter Dragon, Warlord and me had once. I already gave copies to some people so I'm posting it online. Also there is another multiplayer related log that that only warlord has ATM. gotta get it from him.

multiplayer_log1.txt

[red knight]

Good log, some nice ideas... But the only one that is worrying me from geoscape had not been addressed.

 

As i had told i think Deimos some days ago, i have an idea for multiplayer that address the problems with the geoscape, but i have to sort some thing out yet.

 

The idea is: If we have problems with the geoscape time issue, then get rid of the geoscape in multiplayer... easy...

 

I will explain it a little in detail... Instead of a geoscape you have a XNet command terminal, you have availables missions like landed UFO, crash sites, etc... With inmediate access to the Battlescape mission. You share the base with others and spend money as a group when you are improving the base... (you have your own to pay for use tanks in a missions, construct researched armament like plasma tech at very high costs, trading with your base mates, and the like) suppose you are squad from a base (8 soldiers, maybe with support after base upgrading for 4 soldiers more), etc. Better communication and surveilance equipment, etc... When you decide to go on a mission you get money and the artifact you recover like clips and the like, you cannot sold them for money (except to other players, maybe you have to pay to convert them to a human usable form)... You can add as much constraints as needed to make the game balanced... You can go alone or with 1 or 2 partners depends on the aircraft (upgradeable too) you own, each with 4 soldiers. And if the mission is getting complex ask your bases for reinforcements. With mostly instant access to the battlescape...

In this way you havent got funding, the countries pays for the missions you do... But if those countries get infiltrated they start to pay less and less... Not to mention that in this way the game can scalate to a Masive Multiplayer Game. With multiple bases... The GOAL: To survive in A VERY HOSTILE game, get enough money to pay for manteinance costs, advance your soldiers, and act as a group. Your base get honor points for their battles and lose them for not only bad performance, but killing or wounding own soldiers (or soldiers from other bases)... Then after a predefined time like 3 months in Utopia the game ends and the bases with highest points will appear in a special ranking...

Another plus is that the goverments make deliverate (after infiltration) or casual mistakes, telling you for example that a UFO is an scout, when it is a battleship...

 

About trading: you can trade (for money) in the white market (your base) or in the black market (the universe - the other players -) but if you sell things in the black market you get only 75% or less of the price you are selling, so all that money go to secrecy issues, then it is not a good idea to sell to other player. But you can place a request order if you really need it, and have the money to afford it.

 

More or less this is the multiplayer idea... Take it or leave it, it is just an idea.

 

Greetings

Red Knight

[gangsta]

I still think that the multiplayer geoscope game is possible. What problem about the geoscope still worries you? Constructive critism is a good thing because people can find solutions to problems they didn't think about before. I actually want you to point out all the reason why you think the geoscape in multiplayer is non doable. War_lord once came up with a system on how to do multiplayer in the battlescape that I didn't like because of problems that *I* saw. So I grilled him about those too, the thing is I've came up for the solution to those problems and now favor his system to the one I initially suggested. If there is a problem let it be known.

 

On that note, the way that I see we should do game logic is to seperate it into two parts. I call these parts the Macro World and the Micro World(anyone got a better name). Anyway, The Geoscope is in the macro world and the battlescape is in the micro world. The microworld checks out data from the macroworld which the macroworld can lable as onMission, inUse or checkedOut. So far to me the microworld is made up of several threads each running a battlescape. These battlescapes are all indepdent and once the thread has run its course the macroworld is informed. Now the macroworld, is a bit more complexed. We might have diffrent layers in there to do the logic. however the top layer is interchangable. It can be a Geoscape, A RAndom Generator Screen, PlanetScape, galaxyscape, etc. Each of those we would have to implement code wise before they can be used obviously. What you described I am for finishing before finishing the multiplayer geoscope. In fact before that there should be an even more simpler interface to launch a single battlescape game with all the options for random generation. Before implementing the battlescape for the single player game it is easier to implement the battlescape for the multiplayer game. You know this but the reason is that multiplayer battlescape doesn't need AI.

 

Anyway, like I said single player is getting top priority with me. I want to work on the backbone of the game to make sure that multiplayer is eventually possible. I'm also skilled enough in software engineering to know that the way to the ultimate goal is not to just jump and start working on the multiplayer game. Like I said there are several steps to take to getting to that goal. Just want to make the right ones so there won't have to be much rewritting to make multiplayer work once the time comes. I still want to have the multiplayer geoscope as an ultimate goal and I haven't yet heard any reasons to make me less optimisitic on why that can't be done. I like to hear what you think the problems are. I think the time thing I came up with a reasonable solution paper wise.

[Maverick]

I think both ideas would be fun and I think both are worth looking into. Anyone ever played Strike Commander? Very cool game -- flight sim. It was a borderline RPG flight sim with you taking control of the "wildcats" and piloting them to victory...if the price was right. RK's idea reminds me of that (like maybe everyone works for their own merc squad that the countries can hire [after xcom's collapse?]) There are definate options here that can be discussed. The alternative is to make the geoscape work (which would definately be cool) but might not offer the massive multiplayer that the other one offers. Also, with the x-net concept you could make it so that joining a mission could be on a different server (there is a lot to go with this, the simple version is that x-net runs on my machine -- battlescape is joining a server, all the servers do is restart battlescapes--almost like CS only the money you earn from the missions transfers back to base, you only have to leave battlescape to return to base and re-equip, etc) Just thoughts, don't take them too hard. I'd like to see what we can do with the geoscape, but I like this idea as a backup plan. best yet, see if you can design both aspects in the same program -- offers mod capability, and if they used the same networking programming it would be easy to do (relatively).

[gangsta]

well the mass multiplayer thing isn't an option for RKs either unless we design diffrently than originally planned.. so far what has been talked about is writing a single server game. The battlescape *could* be put on other servers since the macroworld and microworld logic are mostly indepedent. However that would add more context switching if run on a single server. The system I was aiming for wasn't a massive multiplayer system but more of 1 to 20 (maybe 50) users on at the same time. masive multiplayer is alot more users than that and the bottleneck is the internet. A geoscope probably could be done too for a massive multiplayer game too depending on how many users. But quite honestly I don't think we have the servers to do a massive game. 20 people on at once would be cool enough for me.

[Maverick]

I don't know how to do multiplayer, and I've talked to the guys here (who are learning Java) and they all say that Java could do such a thing -- but would have a very difficult time interfacing with another program. So that option is all but out. So, I'll pretty much leave it up to you guys to decide which monster you want to tackle. I'm not going to say that one option is better than the other, they'll both be great fun and each one has its own difficulties. I'll continue checking this post for the decisions that are made and as soon as we have some concrete information the design doc will be updated with it.

[gangsta]

Well for a project like this you really want to use C++ Java main disadvatages when it comes to game programming is garbage management and how it does strings (expensive)

[mikker]

I just god an idea to my mind...What about a server running non-stop? And when you are not playing, your funds, and such is saved (but you lose any craft+crew in missions, and interceptors in air will return to base), but that is just me. Maybe a AI can continue the game, while you got to bed, go to school, have a cop of coffee (or te, if an english server), or any other non-computeristic activitys.

[Guest stewart]

RK can we take multiplayer Battlescape as a given?

 

As for multiplayer Geoscape are the problems really technical programming ones or are they actually just game mechanics problems. For example could we implement multiplayer geoscape but at the moment it might not be fair or logical but still implementable? Don't get me wrong though unless multiplayer geoscape can be shown clearly to be an exceptable level of work we shouldn't consider it for v1.0.

[red knight]

RK can we take multiplayer Battlescape as a given?

 

As for multiplayer Geoscape are the problems really technical programming ones or are they actually just game mechanics problems.  For example could we implement multiplayer geoscape but at the moment it might not be fair or logical but still implementable?  Don't get me wrong though unless multiplayer geoscape can be shown clearly to be an exceptable level of work we shouldn't consider it for v1.0.

Multiplayer battlescape is a given because of some things like:

 

It is useful for debugging...

We dont have an AI programmer YET (See point one)...

It will be used to test and debug the multiplayer backbone (see again point one)...

 

Multiplayer Geoscape had game mechanics problems right now... i think the issue should be defered until battlescape completion...

 

Greetings

Red Knight

[gangsta]

Well what problems do you still think that the multiplayer geoscape has? It is good to list them because once they are out in the open people can try to find solutions for them.

[gangsta]

BTW for the multiplayer battlescape I suggest we do it how Warlord suggested in the past. That is during the team turn allow everybody on the team to move at the same time. I didn't like this idea at first cause it made the game less like chess but warlord main reason to do this is so you don't have to wait forever on slow players. He stated that what made UFO2000 suck is waiting on the other player which against him was Peterdragon. Peter would take a long time to move each of his soldier carfully calculating everything. The last time Warlord posted this idea I was critical of it like RK is critical of the multiplayer battlescape. The main problem I saw was two players trying to walk into the same position. That could really piss off a player if his strategy got messed up cause someone walked into that spot and now he wasted all his TUs. The solution for this problem is that since the server gets those kind of movement request sequetially anyway is to let the server send a "Request Denied" packet back to the client when another player walked into the spot thus allowing the player that got denied to come up with another strategy for his guy. The aliens would still move the same way as X-Com 1 there is no need to make em move all at once like the players. Also single player and the turned based multiplayer battlescape can all be implemented this way too since programming it that way makes it more flexible.

[red knight]

Well what problems do you still think that the multiplayer geoscape has?  It is good to list them because once they are out in the open people can try to find solutions for them.

They had been listed in this very same thread, you can retrieve them and post them in a sticky thread to others to try to solve them. I had even post an alternative to the geoscape that can allow Massive Multiplayer Approachs without problems.

 

BTW for the multiplayer battlescape I suggest we do it how Warlord suggested in the past.  That is during the team turn allow everybody on the team to move at the same time.  I didn't like this idea at first cause it made the game less like chess but warlord main reason to do this is so you don't have to wait forever on slow players.  He stated that what made UFO2000 suck is waiting on the other player which against him was Peterdragon.  Peter would take a long time to move each of his soldier carfully calculating everything.

Thats easily solved, before starting you both agree on a time frame and then both of you stick to it (bahh the machine stick to it)... You use a timer.... And it adds more presure to your actions too, say you have to complete every move in 30 seconds or better 15 ... Just like "Blitz Chess game". In that way we dont introduce untested game mechanics issue before we have something that works... Im worrying in scalating complexity, remember that we are spare time programmers, not full time programmers in this project... not to mention that we are not a lot of programmers either. For instance i couldnt do anything in this month (except work on some design issues in the 3D Engine -untested changes yet-, and a little research on a bones character animation file format -ms3d- ).

 

The last time Warlord posted this idea I was critical of it like RK is critical of the multiplayer battlescape.

Im not critical about battlescape, i am about geoscape multiplayer game mechanics, not battlescape. In fact battlescape game mechanics excellently fit on multiplayer because of its fixed time - turn based nature...

 

Greetings

Red Knight

[gangsta]

Well what problems do you still think that the multiplayer geoscape has?  It is good to list them because once they are out in the open people can try to find solutions for them.

They had been listed in this very same thread, you can retrieve them and post them in a sticky thread to others to try to solve them. I had even post an alternative to the geoscape that can allow Massive Multiplayer Approachs without problems.

 

Well the concerns you posted here that I remeber I did provide a solution for them. From what I remeber was your main concern being about syncronizing time. Since I'm sure that the movement equation of UFOs involves a variable t for time in the multiplayer game you can create the special effect of having the hunter (interceptor) and prey (UFO) being in diffrent timeframes when in fact you just manipulated the equation by multiplying t by another variable to make time seem like it is going faster or slower for those crafts. There is no time syncronization problem there. Also the other solution was to sacrafice reality for fun which is a basic tradeoff. I don't know of any other concerns you posted that I haven't responded to.

 

Thats easily solved, before starting you both agree on a time frame and then both of you stick to it (bahh the machine stick to it)... You use a timer.... And it adds more presure to your actions too, say you have to complete every move in 30 seconds or better 15 ... Just like "Blitz Chess game". In that way we dont introduce untested game mechanics issue before we have something that works... Im worrying in scalating complexity, remember that we are spare time programmers, not full time programmers in this project... not to mention that we are not a lot of programmers either. For instance i couldnt do anything in this month (except work on some design issues in the 3D Engine -untested changes yet-, and a little research on a bones character animation file format -ms3d- ).

 

That was my original argument but then he brought up how long you still have to wait. Like say you control 2 people in a team of 14 people then 30 x 12 = 360 secs = 6 mins. if you control 4 people that would be 5 mins of wait and from what warlord said waiting that long is very annoying. Now 15 seconds might be too short and rushing people is another untested game mechanic issue since we haven't tested it to see if it is worth it. Yes there is chat in the game and you can participate in multiple battlescapes but some people don't want to chat when playing and may just want to play one battlescape. As for if this kind of system works that is a yeah since other multiplayer games like this do it this way.

.

The last time Warlord posted this idea I was critical of it like RK is critical of the multiplayer battlescape.

Im not critical about battlescape, i am about geoscape multiplayer game mechanics, not battlescape. In fact battlescape game mechanics excellently fit on multiplayer because of its fixed time - turn based nature...

 

Well you still have to say more of what you think doesn't work. From what I can tell you had two problems with it. 1) Time Syncronization to which I posted a reply offering a solution to which you haven't further commented on. 2) You didn't like some things that didn't seem realistic. My comment on that was that in multiplayer you sacrafice realism anywhere to make it work or more fun. It's a valid tradeoff and people are playing multiplayer more so for fun than realism. The realism thing is more important in the single player game.

[Guest stewart]

Very well RK we can defer Multiplayer Geoscape discussion for now, let's not forget about it though. If the API for interfacing with the networking Protocol is designed correctly then we can defer it, I think.