How to solve warp-problem in Multiplayer: solution

[razark]

3 hours ago, EpicSpaceTroll139 said:

Player 1 is landing his first rocket on the Mun. Players 2 and 3 are designing turboshaft helicopters. Player 4 is learning how to build a plane from Players 5, Players 7 and 8 are building a base on Duna.

The problem I see there is that you should have four different KSP instances, instead of all those players in one.  Player 1 can do his thing, 2 and 3 can be on their own, 4 and 5 off on another server, and none of them bothering 7 and 8.

And then implement a cross-server chat system that allows all the players to communicate with each other as they wish.

 

3 hours ago, EpicSpaceTroll139 said:

However, it allows players to do whatever they want, whenever they want.

Which they can still do, in single-player mode, with no one to stop them.

 

1 hour ago, EpicSpaceTroll139 said:

Many multiplayer games allow people in the gameworld to do different things.

The nature of KSP makes it somewhat different from other multi-player games.  Using them as an example is not entirely feasible.

The way I envision any multiplayer version of KSP working would be with small groups (maybe four players at a time), working on a similar things at the same time on the same server.  That way, one can be launching a supply ship to a station while another is moving modules around on that station, or some folks can go race airplanes together, or build a Munbase, and there's no reason to fight over timewarp and worry about who is when.

 

 

18 minutes ago, Veeltch said:

I'd much rather have multiplayer in a form of Telemachus/Houston mod where one is the pilot and one is the mission control.

I'd love to see this as a possible multi-player mode.  However, I'd still like to be able to fly with people in a multiplayer mode, as well.

[EpicSpaceTroll139]

6 minutes ago, razark said:

The problem I see there is that you should have four different KSP instances, instead of all those players in one.  Player 1 can do his thing, 2 and 3 can be on their own, 4 and 5 off on another server, and none of them bothering 7 and 8.

And then implement a cross-server chat system that allows all the players to communicate with each other as they wish.

<snip>

The way I envision any multiplayer version of KSP working would be with small groups (maybe four players at a time), working on a similar things at the same time on the same server.  That way, one can be launching a supply ship to a station while another is moving modules around on that station, or some folks can go race airplanes together, or build a Munbase, and there's no reason to fight over timewarp and worry about who is when.

Ummm... that sounds similar what DMP does... Are you suggesting something like multiple persistence files, so that, say, if Player 2 wants to go to Duna with 7 and 8, he loads a onto a new save/game instead of launching there and syncing? I'm a little confused

[razark]

Just now, EpicSpaceTroll139 said:

Ummm... that sounds similar what DMP does... Are you suggesting something like multiple persistence files, so that, say, if Player 2 wants to go to Duna with 7 and 8, he loads a onto a new save/game instead of launching there and syncing? I'm a little confused

No, I'm saying if Player 2 wishes to interact in the same "shared universe", he logs on to the same server as 7 and 8.

If 2 then wishes to join 3 and 4, he would log off of the server that 7 and 8 are on and log on to the server 3 and 4 are using.

Once 8 is tired of playing on Duna, he messages 1 and says "Let's go play on Minmus", logs off, starts up his own server, and logs onto it with player 1.

The key is to keep small groups doing similar things partitioned, and each server would maintain it's own version of the Kerbal "universe" (persistence file). 

 

I'm thinking a system like this would need some sort of import/export functionality to allow players to transfer creations between "universes".  Say a player builds a station with some friends on Server A.  Once completed, he could "export" that station's craft file and orbital information, log on to Server B, and "import" it into that universe so that players there could then interact with it.  Whoever is running the server could set whether imports are allowed, and "object" owners could set whether to allow other players to export their creations.

[Arugela]

7 hours ago, Red Iron Crown said:

Sync Issues [ ] Waiting [x]

"None of you can warp for the next 20 minutes while I do this reentry and landing. Have fun."

One thought on the waiting is if you can freely switch between kerbals/ships you can be a copilot etc. Just add to the depth of the game so there is more to do. Then waiting and single que warping won't mater. You will need help or use it. Just need to add depth to the flights. The game is already heavey on being easy/simplified. Add some of teh stuff so a NPC does an OK job but is predictable. Make is so a player can do more. Either way stuff to help each other should/could make up for multiplayer waiting issues. Plus you can organize mission so the boring stuff is done while the others are offline or something then sitting in orbit or on the way.

But the depth idea can help eliminate this also. Just need to add more realism to what it takes to get into orbit etc. what else would they do on a real mission potentially. Either way you can have enough permissions to do certain steps while others aren't playing. A 20 minute launch Doesn't need to be 20 minutes. It can be warped if no other live missions are happening. which means it will not be 20 minutes real time unless others are busy.

Plus what are the sync issues. can you have it center on one player or server centered in multiplayer then have people moving around with the server having the universe around them minimally. I sort of thought that sort of issue would be the same regardless. and if you have one warp with the single player current setup wouldn't you have less sync issues. wouldn't each machine process independently and be like any other mmo. Isn't there a minimal amount of info needing to be passed as long as they are out of each others ranges to some extent? That or it can go into a more synchranized mode when they are near each other helping. Although I kind of thought that would be no different as the switched around units near each other. Wouldn't that be simpler than the other mode.

They can always add in a single mission que setup(which is single player with a few extra slots) and then add asynchronous warp later. The waiting sort of fits the nature of teh game anyway. And you can always help each other or do it offline. And if not they could focus on adding things to the game for people who are working on the game at the same time on the same mission. Those are things needed in the game anyway. Basically non player generated things to do on destination and possibly more technical details/potential difficulties on flights etc. As long as they are not too boring and superfluous. Having a basic mutliplayer would give a mindset to start figuring those things out as people would deal with the realities and naturally start seeing more of the issues.

So to simplify give everyone more to do so doing the same mission together is not boring. Hypothetically that would deal with the whole issue. And it's a badly needed part of missing depth in the game currently. I think most of the issues with multiplayer is it's really just missing things it needed anyway to make multiplayer and single player more fun no matter which way you do it.

Edit:

Just noticed the X is no in synch in the above thing. Derr.... Either way.. 8) I definitely think the waiting can be dealt with. and it would make the game much better in the process. It would force the need for more gameplay additions over time. All of which we've been screaming for anyway! As I said before. Why not add asynchronous warp abilities over time if they are needed. Start with synchronous work your way up.

If they did dedicated host vs player hosted game could they do hosted with only a small file noting the games needed data. Like location of objects etc. Isn't the required data fairly small in this game? Or is it more complicated than I'm thinking. I don't know why I brought this up... Couldn't object location be kept track of with just a small file change constantly. then whoever is hosting just need minimal data and it updates and references. If the game already does this when out of range in single player you just let the computer side deal with the life info until you have to do things in range. Is it expensive to track other moving objects out of full loading range if you simply stuck a new person in a current single player match. I was assuming most of the live stuff was separated and could be loaded onto different players machines already. can you ahve players not in the middle of the universe. AKA having them rotate around the other player. Or the base player as the reference? Possibly simply adding multiple people. Or does that not work?

If you did synchronous first could you simply do it as a single guild then add asynchronous if you find a way and make it guild(s). It could get multiplayer in the game faster hypothetically. Assuming a bunch of stuff. Then you would have solar systems of players against other solar systems of players all rotating around(referenced to) each other... Maybe swapping between the base players as needed. Even if not on the same team. Assuming non coorperative/ competitive stuff is put in by default at some point.

Edited January 23, 2017 by Arugela

[tater]

What's the presumed point of multiplayer? 

Serious question, with a few possible answers as I see it. Note that I will leave out the obvious one, combat, as we're talking about stock MP, and combat in stock KSP is not a goal.:

1. You are within visual/interaction range with other spacecraft, and those craft are controlled by other people (since the game has no AI (which it should)).

2. Players are cooperating somehow, which is basically single player that sometimes leads to point #1.

3. Players are competing against each other (a space race of some variety). In this case players might never be in situation #1, but only know of the other's existence via "milestones" the other players achieve.

 

As I see it, #1 happens very rarely unless players are intentionally cooperating. If players are "doing their own thing" then MP is pointless. In the case of #2, we're fundamentally talking about sharing spacecraft designs, and docking where one player needs to effectively be passive since if 2 ships are docking without the players agreeing beforehand on the exact choreography, it would be a mess. In the case of #3, with the exception of Apollo-Soyuz type cooperation (already mentioned as #2), this might as well be 2 single player games sharing data. This is the most interesting version of MP as I see it, but it would require a complete redo of career to make time meaningful, and perhaps even include failures, so it's pretty much off the table.

So what's the bottom line? #1 which will happen about as often as you see spacecraft dangerously swing past your station within physics range that were not ever sent to it (very, very unlikely), or #2, where at the point of actual "multiplayer" (2 people doing something that can actually have the craft interact) it pretty much requires on one craft to be entirely passive. Fun.

I'm just not feeling it.

 

Edited January 23, 2017 by tater

[dimmy]

Guuuys!

Please, try to think!

WARP as is can't be solved it multiplayer KSP - its true!

But we don't have to had it in multiplayer KSP in that version as we have now in single player or DMP (with its issues).

 

There are many developer's troubles and difficulties which had to be solved with WARPing.

But how to travel in deep space, you'll ask me.

Answer: hyper-drive. None of current physics and game mechanics ARE NOT  NEED TO BE CHANGED for it implementation!

All players play in WARP 1x speed as usual, can interact with each other, but, if you want to travel further and faster, you'll fly away from "warp-denied" zone, and, you can do hyper jump using traveling points on the map as usual, it'ss be in cost of the electricity fuel. Come out from "hyper-space" (simply dissaperaing and appearing in new place)  can be done only in safety zones (not in "Warp-deny" zone, not near than 10km near the closest ship or asteroid). But what about planets travel prediction, you'll ask. Answer is simple to implement too: the planet's and world time doesn't change dependable of your jumping. It's time and position always calculates as 1x speed of all-users-world-1x-time. So, when you jump from kerbin to moon, the moon's position in space will not be changed! Yep, it somehow other, than in single player, but it can be MORE SIMPLE TO BE DONE in KSP.

Next: where our ship will be after the hyper-jump? how to predict it using current mechanics?

As usual KSP's map points, you set, where you want to speed up (do hyper jump) on your course, and the second point is - where you want to come-out from hyper-space, KSP for it just calculates estimated electricity needed for jump and ensure you are not in denied-zones.

 

So, you all discuss many variants of it and think, that Space Engineers or something like that should be played instead of KSP for multiplayer, but it is not the last variant. Just try to find mistakes in my logic.

For multiplayer in KSP it need to be:

1) added hyper-drive engine (yep, like Space Engineers maybe), but available only in multiplayer. Or just an option in the game's map. Hyper-drive works on electricity or liqued fuel, your ship is not controllable while you are jumping. It taked near 1-2 seconds to jump anywhere (mostly need for KSP's mechanic to remove your ship from the world, and calculates new position and syncing with the world and then appear in that place).

2) added deny and failsafe zones for entering and outing hyperspace

3) added hyper-drive traveling points on the map (start jump, end jump), and turned off planet's speed prediction and incrementalling for the map.

4) added ship's and part's access organization for players and their factions.

5) added build-in chat system.

6) fully removed warp-ing for multiplayer. For more interesting gaming, planets rotation speed and traveling on their orbits can be increased as they almost always will be on that position (We turned off warping! as you remember). So planet's orbit speed can be set so, that it can do full rotation over the sun on its orbit by 4 hours, for example.

Any troubles and problems with such solution? Lets analyze it. What is difficult or wrong in this idea? Almost all what need's to be added and done - already exist in the game!

So, KSP can still be alive and attract more players to play with!

Edited January 23, 2017 by dimmy
removed spaces

[dimmy]

5 hours ago, tater said:

What's the presumed point of multiplayer? 

Serious question, with a few possible answers as I see it. Note that I will leave out the obvious one, combat, as we're talking about stock MP, and combat in stock KSP is not a goal.:

2. Players are cooperating somehow,

 

Of course cooperating. One builds orbital refueling base. The second - mothership, based on 5 rockets.

Goal? Its easy:

1) teams competition who can faster travel to the moon like google's X-PRIZE and solve tasks in it (photo, traveling, etc.). Also, if google or NASA knew about such possibility to make this advertisement for MOONAR X-PRIZE, i think, they even can give money to devs to help implement it.

2) teams competition (or sandbox - like gaming) to build duna's base with farming, etc. like "The Martian" movie.

3) do i should to offer more examples of super-interesting command multiplayer gameplay in KSP? there are huge possibilities! From studying physics and astronomy to battles like capture-the-flag!

You should only turn on your imagination and you'll see.....

Official CAPTURE-THE-FLAG competition: 50 players, 50 starting platforms on KERBIN, 1 target on the EVE, 50$ prize... =)

 

It's not unreal to be developed in KSP. Just read my suggestion one post before.

[Red Iron Crown]

23 minutes ago, dimmy said:

Answer: hyper-drive. None of current physics and game mechanics ARE NOT  NEED TO BE CHANGED for it implementation!

That would make it an entirely different game, IMO. The core challenge of KSP is space navigation bound by (mostly) realistic physics and rocket performance, some sort of instant teleport drive completely sidesteps that. There are quite enough space games that ignore how spaceflight works, I would hate for KSP to become just another one of them.

 

[dimmy]

Its Realistic (in future):

https://www.nasa.gov/centers/glenn/technology/warp/warp.html

If you would like to know more about the theories of interstellar flight, you should visit the Tau Zero Foundation.  Marc Millis, a former NASA Glenn physicist, founded the organization to consider revolutionary advancements in propulsion.

https://tauzero.aero/

also, there are Nuclear-detonation engines, etc.

 

We, game designers, science fictions writers are making the future to become real. All starts from such suggestions, games and ideas.

Why to refuse from start of them?

Edited January 23, 2017 by dimmy

[tater]

2 hours ago, dimmy said:

Of course cooperating. One builds orbital refueling base. The second - mothership, based on 5 rockets.

Goal? Its easy:

1) teams competition who can faster travel to the moon like google's X-PRIZE and solve tasks in it (photo, traveling, etc.). Also, if google or NASA knew about such possibility to make this advertisement for MOONAR X-PRIZE, i think, they even can give money to devs to help implement it.

I addressed this one. The only time you are actually doing "multiplayer" one player is active, and the others are just watching.If I "help" by trying to rotate to point the docking port at him, but he intends to dock as it is now, I just screwed him up. Better to do nothing.

2 hours ago, dimmy said:

2) teams competition (or sandbox - like gaming) to build duna's base with farming, etc. like "The Martian" movie.

Also addressed. Aside from being within a couple km of each other, there is no "multiplayer" as much as it is sharing milestones. Competition without time meaning something seems odd, too.

2 hours ago, dimmy said:

3) do i should to offer more examples of super-interesting command multiplayer gameplay in KSP? there are huge possibilities! From studying physics and astronomy to battles like capture-the-flag!

You should only turn on your imagination and you'll see.....

Official CAPTURE-THE-FLAG competition: 50 players, 50 starting platforms on KERBIN, 1 target on the EVE, 50$ prize... =)

It's not unreal to be developed in KSP. Just read my suggestion one post before.

I'm not seeing anything all that novel that requires "multiplayer" in the sense of players on the same server.

Hyperdrive is a fine solution, but it's not even slightly "realistic."

EDIT: It's important to add that the very first thing anyone at Squad working on MP should think about is, "What's the point?"

The first thing. Career mode was built to make a career mode, without really thinking it through, and as a result doesn't actually do any sort of career very well. A "sandbox" MP would be the worst of all worlds. I would argue that time needs to be meaningful for almost any interesting use of MP, which adds all sorts of complications required (including multiple launch sites, as pad turnaround should not be instantaneous).

[dimmy]

29 minutes ago, tater said:

"What's the point?"

Dev's point is: Attraction more players to KSP, and not invite them to play MP Space Engineers, let them stay on MP KSP.

Devs can create MP KSP on Steam to have money support for it.

 

To us this is:

Maybe realtime teaching and helping each other during building ships, navigation and travel?

In the first iteration this is to add "Spectator Mode" for guests with mics and chat.

In the second: function "take control", on which host can pass the control of the ship to the "spectator" and get back if wants.

In the third: "competitive" (or name it as you want) full MP mode, as described above with hyperdrives.

Edited January 23, 2017 by dimmy

[KSK]

7 hours ago, tater said:

What's the presumed point of multiplayer? 

*snip*

In the case of #3, with the exception of Apollo-Soyuz type cooperation (already mentioned as #2), this might as well be 2 single player games sharing data. This is the most interesting version of MP as I see it, but it would require a complete redo of career to make time meaningful, and perhaps even include failures, so it's pretty much off the table.

Agreed. I'm also thinking that Space-Race type multiplayer will be extremely hard to balance and/or have reasonable matchmaking for. An awful lot of things in KSP are really difficult until you figure them out at which point they quickly become routine. I think it'll be hard to find a sweet spot where players can meaningfully compete without one completely curb stomping (or should that be kerb stomping?) the other. This isn't helped by the fact that:

a) A lot of KSP still isn't very discoverable. The game is still heavily based on a 'learning by failure' model, which is (arguably) fine for single player but dreadful for multiplayer since it makes it hard for players to compete on a vaguely level playing field.
b) There's not enough variety or trade-offs in possible mission modes for fun competition. Put another way, I can unlock a bare handful of technology nodes and get to the Mun. From there, I can use the same spacecraft to get to Minmus and, with some tweaking, Duna. I wouldn't claim my designs are absolutely the most efficient but there's not a massive amount of optimization left either. I think competitive KSP will very quickly be min-maxed to death and then have nowhere left to go.
c) There's nothing else to do in KSP apart from building and flying rockets. Competitive multiplayer games generally rely on giving the player a lot of things to do at once and seeing how well they juggle them all. KSP is currently extremely limited in that regard.

[dimmy]

Agree with KSK, time, spent in KSP make sense for players and competitive mode. If this is so huge problem,

what about a complete different look into KSP MP:  MP career. Where you with your team can simultaneously build the same ship and control it?
There is nothing to change. Old good warp. You decide which cockpit can control thrusts, which cockpit (player) can control legs, etc.

After some time you can divide ship onto 2 or more. But the warp button works only if everyone agrees with it to do it now.

 

So, what about career MP?

 

* It seems like you all say: "we dont want MP, let stay KSP without it, as is".

Edited January 23, 2017 by dimmy

[CaptainTrebor]

If I were to implement multiplayer, I'd make it so that everyone plays as a specific kerbal on a single mission, and make it so that only the mission commander (person hosting the server) can warp.

[KSK]

1 hour ago, dimmy said:

Agree with KSK, time, spent in KSP make sense for players and competitive mode. If this is so huge problem,

what about a complete different look into KSP MP:  MP career. Where you with your team can simultaneously build the same ship and control it?
There is nothing to change. Old good warp. You decide which cockpit can control thrusts, which cockpit (player) can control legs, etc.

After some time you can divide ship onto 2 or more. But the warp button works only if everyone agrees with it to do it now.

So, what about career MP?

* It seems like you all say: "we dont want MP, let stay KSP without it, as is".

Could be quite fun if done as @Veeltch suggested where one player has the job of Mission Control - with access to Map view and the ability to place maneuver nodes - and the other player(s) are restricted to IVA view and have to follow the instructions from Mission Control to complete the mission.

I don't think it's something I'd play a lot of but I can see the appeal. I'm not as sure about the way you describe it, at least not with stock KSP. The player deploying the landing legs doesn't seem to get much to do compared to the player doing the piloting for example.  

[tater]

The "mission control" model is OK, and eliminates the warp problems as everyone in on the same ship/mission, so always warps in lockstep. Ditto with a spectator mode... It;s not something I'm terribly interested in, frankly.

I'd much prefer some well-designed AI, frankly.

I'm not utterly discounting "sandbox" MP, but I think to me, it would be a lot like IVAs in the stock game. Cool to look at once or twice, then never looked at again. I play flight sims, OTOH, entirely from "IVA"---but this is because such IVAs are actually useful, unlike in KSP, since at the very least I can look around, even lean, with my hat switch, etc. KSP IVAs are too limiting, and in RL spacecraft are in fact more automated than KSP allows for (when the player literally micromanages everything, then sometimes pure IVA is less, not more realistic).

[Arugela]

That was why I suggested a MMORPG guild permission systems. The base system is mission commander. He has his game and says lets open up for four players in single my single player game and putit up in a game finding program. People see 1/4 slots open and join. Or they get theirknown buddies to join.

But we coudl also allow free moving of people between objects. They can freely go to the KSC and go to the menu for craft and switch craft. Their PC woudl start processing from the point of the new craft visually. The game might use the logic of having them rotate around the commander to address the world rotating around one player. Use a ready check system for warp with the commander to have override. and have other permissions for customization to allow anything. It could be almost the same as mutliwarp in flexibility.

This could allow people to switch between multiple kerbals and ships to help play the game freely. Or go get the other ship and get it ready to be on a docking path. When needed warp, etc. Lost of ways to give control or make it work out. It's single player with multiple people. Add in view mode and more than one person can be on the same ship/object. One controls the others view. That covers all logical grounds.

But then the game needs or will be oriented on more things to do in mission. Which I would think would be good. Maybe we will come up with and see more ways to play the game from the games standpoint. More goals. As the game would be dealing with them more. Maybe analyzing rocks or doing whatever.

It would be fun to help each other unload kerbals and run around loading up vessels for travel or whatnot. I would think you would need to have people choose the vessel they are in and switch freely though that way you can efficiently help each other. Or else it would be hard to get stuff done.

If there are technical issues with the games seeing each others movements out of loading they could have an option and the ability to change when it loads the changes.  Kind of like the physics option per second. You don't need to see the changes that fast in order for the other guy to do stuff. Except maybe for comms satelites or something that affects the game. Assuming it needs to be visually represented and not just a check for use even if desynced, which is more realistic. It could update every few seconds. Not to mention in real life delay of info is a real thing and could be used to make some new cool game features. Information update timing. It could be applied to the whole game to add more depth. In essence Fog of War! And multiple other layers to require exploration in multiple new ways. Adding depth to the game slowly. Heck make the upate of visuals based on the technology in the comms etc. Then it matters if you upgrade the comms(note I haven't used comms yet much. But if anything can be added add it tehre for more realism/mechanics). It should be predictive by nature anyway. Then as you get close it can check for accuracy and delays and off information are normal just like potentially in real life.

And, like I said in a previous post. I bet asynchronous warp can be added later on top of synchronous warp. So starting with synchronous wouldn't be bad to get it in game and start seeing what needs to be added to make multiplayer interesting and functional. Asynchronous could be a second KSC. The russian analog version with different base tech/parts and mechanism to trade/share tech and do group research or other things. Basically going from 1 to 2 guilds! 8) Then you could have competitive with different tech and the game progresses to cooperative between two groups etc. Eventually info/satelite sharing etc. A space race as you will! 8d Use real life for ideas to put in the game just like their RL analogs. Would make it easier to come up with mechanics. LIke giving each their U.S/Russian tech base seperately. It could give some ideas for new parts. Especially if each sides analog part is not exactly the same.

If asnych if finally added on top of synch! "KSP 2.0 Let the space race begin!!" >< I bet that could add a lot to non sandbox games. Both in things to do and new ideas to add to the game. End game could be cooperative play opening up between guilds.

If I knew how to make mods I'd be working on this right now! >< Has anyone done this sort of multiplayer in a mod yet? Synch mod based on MMORPG guild systems. I wonder if the idea of players rotating reference around the commander works practically. Or his assumed position. I assume the game can still rotate around him even if he leaves game. If you get two guilds going can you send players from one reference to the other if needed for things. Even if staying on the same team for cooperative play or other random needs. Assuming it's helpful to anything.

Edited January 23, 2017 by Arugela

[StupidAndy]

my multiplayer warp solution: WORD INSERT HERE!!

instead of the warp deny zone thingy, everyone has a month, plays that month, sims that month, does whatever they want to in that month, and then they stop at the end of the month, simple! you can have as many warps as possible and it pretty much solves the warp problem in multiplayer!

and here's the other good part of it: it means unlike Stellaris and other games like this, you can play for about an hour, and then stop, then play again tomorrow, you don't have to set out an entire day for nothing but the multiplayer game, then get disconnected for about an hour and lose everything, you can keep the data in one save file and you can keep coming back to it day after day, week after week, month after month? whatever you want, you can just do it for all of eternity! because you don't have to be completely dedicated! and another thing that might work is that you can have more then one game available, and you don't have to play on one game at the same time.

the problemsolving part: so you have a station in orbit of Kerbin (nice job!) and someone is coming to resupply it, but you already have the docking port they need taken, so what do they do? you give permission to separate people to be able to control ships around the station! it allows the people you want in your station in and the people you don't out!

IT WORKS!

[Arugela]

If that involves playing at the same time with separate warps you still have the problem of if people do something in the same area during that period while not synced. You have the issues of time travel in fiction. the only change is you lessen the time where you can warp at once. That works best for TBS(turn based strategy) games where the actions are simple. If one guy moves faster the piece is moved faster and beats him to it or the action does not matter as it is calculated at the end of the turn. There is no way to calculate everything at the end of the turn and still do anything without it becoming boring. This is a live interactive game.

If anyone in this game moves out of sync you have more wait times. You just moved the wait times to later or deal with them in a funny manner. You must always deal with wait times. Either deal with them live or deal with them later when you run into their already moved item. It's picking between options. There is no way to have the ideal. That is why I say get a synched warp instead of asynchronous until you can figure out what type of asynchronous does not interfere with the game design if any.

Unless I'm misunderstanding the previous idea. Which I might be.

I believe there are worse problems with waiting for asynchronous warps because of unrealistic (Unrealized, Farscape reference! ><) realities it introduces. Not to mention it is more complex to potentially deal with/keep track of. If you do synchronous, in this game at least, you can still override and do live warps if all items are in stable orbit. There is little to worry about past a few situations(takeoffs, midflights, landings etc). Unless they add life support. Since you just loose time waiting for a next rendezvous potentially in the worst case scenario. Even with life support you just need a pause feature and wait for each other, or since it's multiplayer do the mission yourself while the other guy is away(cooperative play design benefits). Assuming you make that a fundamental of the game. Freely moving to do any mission no matter who started it. AKA single player with multiple people.

This leads to an obvious statement. Some fundamentals need to be added to multiplayer no matter what though. One of which is, unless there are separate teams, people need to potentially be able to freely take control of structures(or at least partially). Even if they are not theirs.(Unless a controlling player decides not to allow it because of specific rules for his game.)

One thing is obvious though, whichever way you make it, adding asynchronous elements, even if added to synchronous, is making the game literally into a TBS(Turn based strategy game) That is literally what a TBS is. It is literally defined by adding the elements of an asynchronous nature to a game format on a technical level. And you are then dealing with all elements whether that TBS is a hard turn based version or allowed to play out more freely. You are literally adding every literal element of a TBS to KSP. This is because the point of TBS games is to add turns to allow asynchronous actions between players for short periods. Even if on the surface it appears like it is synchronous. It is really only synchronizing players at periodic time intervals with what are technically asynchronous actions between(AKA you are synchronizing asynchronous play periodically.). Which is fundamentally an asynchronous game at it's core.

So, KSP has to do synchronous to stay a pure live game. Or it has to start adding TB elements to become asynchronous on some level. Asynchronous is hard because this game is not design for periodic asynchronous actions that can be tallied in a predictable manner(unless I'm missing something). It is a potential live interactive game by it's core at all times. And they literally must start adding the infrastructure of a TB game to add any asynchronous elements. From what I can see anything not tallying the game in the end(AKA set turns) is just another form of synchronized warp. And anything tallying the game(Or any asynchronous elements) is adding endless complexity(technical and non technical) and is no longer potentially live play. While, ironically to many, wait periods are actually live play. Which can actually be completely avoided/bipassed with player control/permission structures! 8)

To simplify, asynchronous gameplay(in all forms) is literally turn based play. That is it's true technical definition. You are adding periodic checks to a live game. Recognizably or not. It is technically backsteping the design from live play to something more primitive on some level to achieve it. And there is literally no way to avoid it! Technically you can only try to control it.

Edited January 26, 2017 by Arugela

[Earthslug]

Omg!

I was brainstorming today and thought I found a 100% working solution to KSP MP.

Then I decided to read the current existing suggestions only to find out my master plan is a total fail and filled with paradoxes...

[Tyko]

 I was thinking last night about this convo while I was playing. I noted all the times I time warped - I probably time warped a dozen times an hour often-times only for 15 minutes or maybe a few hours. There were a few time warps of a couple of days.

If i was in an MP game in which all players had to agree on time warps, most of those small time warps probably wouldn't have happened and I would have spent many hours of my valuable game time reading a book or going for a quick walk while I waited an hour for my ship to reach it's next maneuver node.

Anyone who's an advocate of MP cooperative time warps should really spend a couple of days playing the game without ever using time warp except for major events because in a coop environment only major events would work, otherwise there'd be a overwhelming demand for mini-warps.

Edited January 26, 2017 by Tyko

[tater]

21 minutes ago, Tyko said:

 I was thinking last night about this convo while I was playing. I noted all the times I time warped - I probably time warped a dozen times an hour often-times only for 15 minutes or maybe a few hours. There were a few time warps of a couple of days.

If i was in an MP game in which all players had to agree on time warps, most of those small time warps probably wouldn't have happened and I would have spent many hours of my valuable game time reading a book or going for a quick walk while I waited an hour for my ship to reach it's next maneuver node.

Anyone who's an advocate of MP cooperative time warps should really spend a couple of days playing the game without ever using time warp except for major events because in a coop environment only major events would work, otherwise there'd be a overwhelming demand for mini-warps.

I think the fact that I agree with you underlines the fact that MP and a game that requires time warp are two mutually exclusive things.

[dimmy]

Yep, So as i said before, the better solution is to refuse using warps at all.
Just use hyper-drive instead of warps.
So, we will not have to wait somebody, and all be ok.

But, as someone wrote before, with hyper-drive it will be very different KSP from we have now.
KSP with MP based on hyper-drives is ok for me.

How many of you agree to use MP KSP, based on hyper-drives, as i described before?

[53miner53]

We could have the server simulate the orbital craft not controlled by a player, and the players simulate the craft they are controlling, and the server relays the controlled craft to the other players. Time warping would only affect the player that is warping, and any craft that the time shifted player interacts with would first ask the player who launched the craft if the craft should be able to be interacted with at that point. Then, if the person agrees, that craft is locked and will not be interacted with until that point, other than executing planned manuver nodes. If the interaction is refused, than the crafts will ignore collision entirely. Interaction is defined by the game as a collision or a docking with the craft. Any maneuver nodes would be assumed to have been executed as well for craft positioning. If a maneuver node is missed or passed with 25% or more of delta v remaining and no player controlling it, then the craft that missed the node would be repositioned and the other craft would be assumed to have gone to that spot and not interacted with the other craft. 

 

Edit: I attempted to strike a balance between realism, ease of use, and avoiding people griefing other people's crafts.

Edited January 26, 2017 by 53miner53

[Earthslug]

Not against the idea but might as well be seperate game.

Hohmann transfer gone. Just point nose to duna and fire... halfway, retro burn.

Even if we get used to that play style (like many games use it already), we have more issues.

If we keep te hohmann transfer system and just speed up the craft along the projected orbit, it defies the laws of gravity.

Even if we go for that technology...

Planet alignment may be trivial at those crazy speeds so we don't need 100k fast forward to get our inter planetary transfer windows right before launching.

But that fat class E roid enters soi in 327 days... no fast forward or hyper drive will speed that up.

And if we decide to ignore roids and planet positions, it still is a disaster.

I cheated a infinite fuel 50+g acceleration vessel and flew straight line to eve. Still took minutes and only to crash at 40km sec speed. Oops retro'd a bit too late.