godarklight

@Dr. Death: Subspaces are a bit confusing at first, but they are just about the only way to do multiplayer and enable warping. I also support master controlled warp (although I haven't looked at it in months), it's generally a pain as you can't warp unless the majority agrees with you. I think what I'm trying to say is, "Multiple realms" per player is the entire idea of KMP / DMP. "It's not a bug, it's a feature". I've always generally been a co-op player, so I haven't put effort into any PvP stuff, nor will I . There's a big problem with the way DMP works and weapon mods though - DMP does not send full part updates for other players, so you can't actually shoot each other down. The reason for this is simple: "Spectator: I saw you crash into the ground while landing!" "Pilot: No I didn't!" <- Which one is correct. Kerbinside: Probably works. I wouldn't be all that surprised, most mods work fairly well in DMP. Load distance: DMP tip toes around that one, so any load range bumping mod should work fine - It's a 10 minute mod for anyone willing to try. @inigma: Wait, a load bumping mod doesn't exist? I might make one if people ask. I'm serious about the whole 10 minute mod thing. @dxdt: I didn't take offence, hence the wink - You're right though, I have been promising groups / vessel permissions / shared science for a while, and I thought the months I had were enough time. This wasn't the case - other things kept taking priority. @MK3424/Atheu/Duke23: That is almost certainly an issue with input locks (Alt+F12 -> Show Input Locks), and it's almost certainly an issue with KSP's faulty handling of "applicationLock*". I'll poke this one in a sec, It would be a 5 minute fix if I'm correct @Tery215: There's already quite a few http://d-mp.org/?p=slist @Preliator: I'm actually completely unfamiliar with RockyTV's plugin autoupdater - But I see what you mean, the link just relinks back to the download page. I'll get onto VITAS about that one - he's running d-mp.org. @Everyone: Actually - It was my faulty handling of input locks. Release 0.1.6.1 is available that fixes that issue .

@dxdt: It's mainly because my full time job - I haven't had a great deal of time to spend on DMP, and when I do it's usually hunting down the KSP weirdness that causes bugs . The idea is to ditch the 0.1.6.0 series as soon as possible so I can work on a few things that requires protocol breaks, like groups / permissions / shared scenario modules / making docking reliable and fixing up surface positioning (rotation/altitude can be funny - the ground doesn't spawn the same between different clients. Terrain detail makes this worse).

@Preliator: How about now? @mindstalker: It doesn't actually work - The position updates are broken if you try to use 0.1.5.6 in KSP 0.25. I've realised that this is hurting people more than simply not having shared science / vessel permssions, so I'll drop 0.1.6.0 now and move on to 0.1.7.0. @Everyone: DMP 0.1.6.0 is available! EDIT: It's now also on the CKAN! EDIT 2: Happy birthday KMP

@longbyte1: Rotation interpolation.. *shivers*. I know enough about quaternions and how I should be slerping them, but KSP doesn't provide you with a rotation it's slerping towards, it's UnityEngine that provides you with a angular velocity vector. It appears setting the angular velocity for the ship *kind of* works, but doing it on the root part isn't enough - the rest of the mass of the ship cancels it out, so I'd need to transform that single vector onto all of the other parts. I'm already attempting to do this, but it's not quite right... @hvacengi: Thanks for helping TheWattGaming @TheWattGaming: Only dev supports 0.25 - but I'm going to release 0.1.6 as is today - not having a release that supports the current KSP version hurts more than a lack of vessel permissions and shared science does. @Duke23: It's awesome when the krakens don't attack DMP isn't so sensitive to latency, but it's very sensitive to TCP connection drops. I might look into putting disconnected clients into "limbo" to see if I can auto-reconnect clients that have dropped out. @LoneRider: That shouldn't have happened in the first place though, DMP mainly works like singleplayer for scenario data - and I'm unsure of the approach that BTSM takes. I'm not interested enough in career though to download BTSM to find out though @Yru0: DMP kills the vessels to replace them on the full update - it looks like the only way to do things. You can imagine what happens to those poor little kerals when this happens. I'll have to figure out how to forcibly remove kerbals from the vessel before killing them... @Dacan: Might have to jump on IRC and get in touch, I'd like to work out freezes if you are getting them

I'd like to point out that this is *not* a 64bit only bug, it's definitely caused quite frequently by DMP's start. It's quite an annoying one as in my case, KSP does not throw any exception - It only throws after you leave the space center screen, and that particular throw isn't helpful. It's not a throw in GameEvents.OnGameSceneLoadRequested as that bug only prevents scene changes and entering of buildings, but does not lock the camera. The camera looks like this: http://steamcommunity.com/sharedfiles/filedetails/?id=330730377 I'll try to poke this one more - It's one of my biggest bugs. EDIT: I've obviously failed to read - Follow NathanKell's advice and upload your KSP_Folder/KSP_Data/output_log.txt (on windows) file. EDIT 2: Input locks during a game start were the source of DMP's problems.

@longbyte1: Very complex: https://github.com/godarklight/DarkMultiPlayer/blob/25997cb6b1f27fbbd9039ce7a0f334d4e4f612e9/Client/VesselWorker.cs#L2167-L2177 (Although this is technically a localhost flight, you did still see jumpiness with planes before due to the way DMPServer and the subspaces work).

@PikachuNL: The root part is kind of the way vessels are handled in KSP, Things that break off become new vessels. This should have been corrected on the next full update though. @Enthalpy: If you don't have the flag, DMP will reassign it to the default flag. Unless it's in DarkMultiPlayer/Flags/, at which point all flags are synced from and two the server - nobody uses this feature though . As for the disappearing vessels, PartKiller is obviously playing silly-buggers - It's supposed to remove vessels / debris that break off other players that haven't been synced through the server, they aren't legitimate vessels / debris, they are duplicates only. @Everyone, My internet has been shaped and I really need to reduce things using my connection, so I've stopped the server list scraper. As noted on the page itself, I'm handing the server list over to VITAS. The server list isn't completely done yet, but it's most of the way there . Go check it out. (It's also instant-updating, it uses a DMPServer plugin rather than making the server list scrape data off all of the servers - You'll likely need to visit me in IRC to set it up until we write some docs for it).

@PikachuNL: I need to code motion prediction. (this is my branch for KMP that was never merged). It's actually pretty low hanging fruit, so I'll nab that one today.@stardestroyer: KSP's terrain doesn't spawn the same for all players, which causes err... issues at the moment. But yeah - you can certainly try @longbyte1: Baby steps, I'll get there (although 31st october is sneaking up on my pretty quickly)

@Steven Mading: Dev supports 0.24 - I'm getting there, just want to finish off a few things before I release it. And because it's a good enough excuse, I'll release on 31st october as that's KMP's anniversary (I thinks). My little way to celebrate 1 year of KSP multiplayer . @inigma: Thanks for showing me the bug - It was a bug in the pid (the way KSP stores GUID's in ConfigNodes) to guid converter. I forgot to add a dash @RockyTV/BCSpazer: I'll code a global auth plugin for server admins to participate in, that way instead of being registered on one server, you register on all of the servers participating @nanowinner: Target actively refused to connect means they sent a TCP reject - which means the server either a) isn't running, or is firewalled. @Mulbin: I'll have to expose them via a server setting - But atm it's just default. @technologygeek: When you click upload on the craft sharing window it should come up with "Uploading / Uploaded!", then other people can download them. @Tery215: Yet another bug that was made worse by KSP 0.25 - This type of thing started in KSP 0.23 and that exact bug is the major reason KMP doesn't exist today . I'll poke it until I figure it out. @chiefymuc: *Plonk* *Explode* . I have to add the PQS altitude and use that instead if they are close, I believe the ground doesn't spawn at the same altitude in KSP all the time :-/.

@jakeb1993: After you leave the safety bubble (which is tiny) you should spawn in. Make sure you use the development version if you're playing on KSP 0.25. @Everyone, thanks to @Neeken helping me test the disappearing GUI bug, things should be fixed in dev. Turns out the "get a new unique id" call wasn't so unique . @The Jedi Master: Only the development version works with KSP 0.25 - Either download it from my build server, or if you're using DMPUpdater.exe, rename it to DMPUpdater-development.exe (This is the best way to keep DMP up-to-date on dev). @inigma: Try to catch me on IRC, the more reports / videos I have of bugs, the better the chances of me fixing them . Stock ships are disabled in career/science mode (because that's what KSP single player does), I haven't really tested the stock ships in a long time (because they aren't all that great...). @Mr Tiddles: You need to be on dev on KSP 0.25, and dev servers appear at the bottom of the server list. I'm hoping to release DMP 0.1.6 around october 31st (which would celebrate 1 year of KSP multiplayer via mods - It was KMP's initial release date). At the moment, the protocol 29 is the latest protocol version, you'll need to connect to those. @Rekrahttam: If you can reproduce this following some certain types of steps, drop by in IRC. It's much easier fixing things if we can actually speak in real time . There's not a lot that can go wrong in FlagSyncer, apart from the (flagURL or HighLogic.CurrentGame being null...) EDIT: I've taken a pre-emptive strike against that error, Test dev to see if that bug is still around

@Harrekin: Ah, the active vessel does indeed get time skipped if your client clock falls behind. I might get DMP to go into a free-running type of subspace while you're the only player in it, that would prevent time-skips in that situation. This is sometimes accompanied with spam in the debug log (Alt+F12, Debug). @Neeken: If it pops up again let me know, I'm not completely sure what happened there, Unless you tried updating DMP while KSP was running? @FlightNewbie: Caused by my sillyness (The token authentication system was replaced by a secure private/public key challange/response system). In the process I had to remove the player tokens, but forgot to check if the directory existed or not first . Fixed in dev.

@Neeken: Not without KSP_folder/KSP_Data/output_log.txt I don't @Steven Mading: You might want to record a video, but DMP doesn't really do stuff in this area - It only takes the flight parameters and gets the other clients to overwrite them each position update. DMP could be physwarping, but that still wouldn't account for that problem. The output_log.txt might come in handy, but a youtube video explaining it might be an even better bug report.

@inigma: If I had to guess, it's caused by using a kerbal to dock two vessels together that you don't control or have update locks for. Get two rovers, Player 1 EVA's, player 2 switches to the other rover that was eva'ed out of (therefore having both the update locks for the rovers, banning updates from the kerbal player 1), and then get the kerbal (player 1) to dock the two ships together. This may not be solveable though - There's many cases where allowing this type of thing would break gameplay, this is kind of the same category as allowing weapon mods to work correctly.

@Everyone: KSP 0.25 has dropped and I've done a very initial port to 0.25, but DMP isn't in a position to do a new release just yet - So I'll encourage people to use the development version for now. If you're using DMPUpdater.exe, you can rename it to DMPUpdater-development.exe to update to the latest development version. @inigma: That is almost certainly PartKiller playing up - it's supposed to remove illegitimate debris or bits that break off vessels. dsonbill: I'll try to reproduce it but I might need a few eyeballs just incase I miss something.

@tetryds: For weapons mods, I suspect the only sane thing to do would be to let *them* hook into DMP and tell the other player which parts were killed, likely with plugin-glue between a weapon mod and DMP (I haven't yet written the public domain plugin API reflector for inclusion into other projects). This does imply client side trust though so it's exploitable, but it's probably a risk people will have to accept for now. As SRB missiles are actual vessels, they should work assuming the lag is low enough (I've still got more to do in this area) @longbyte1: Those aren't stupid questions at all - they are very valid ones The first: I had this exact same question for gimp when he was active in the KMP project, the answer is you don't. Disabling the music helps preserve some of your sanity . Monodevelop gives quite sane warnings and compile errors which takes most of the problems away, and the debug window video above was quite a simple GUI, I copy+pasted the methods I needed from another window . I have a fairly good idea of what I need to do before I start coding, Generally I'll get ideas away from the PC and then try to implement them in either a simple test program or just directly into DMP. If I don't like the approach I've taken but the code does work, I'll put it in a branch, push it to github, and then start again using the first branch as a reference . Compiling takes around 5 seconds, so that's not to big a problem, and I made a command that compiles and copies the mod to gamedata which saves a few seconds each time. Symlinking DMP doesn't work due to the incorrect install detection, perhaps I should disable that on the debug builds Unit testing is basic automatic "does this work" type of testing. I've thought about adding unit tests to speed up debugging for docking and positioning problems, and I actually did write a unit test for KillVessel, but it totally wasn't worth it as it always passed . (Second post) I don't *really* have a workflow, I usually iterate a few times and I like testing things early to see if I'm on the right track, hence the "Ehhhyupp" and "Noooope" buttons. It took me a while to figure out how to say "DMP contacts the server you're connecting to!" without freaking people out over it. DMP does not phone home . Some of that video was waiting for someone else to test it to confirm I haven't broken all the things, someone did actually point out that DMP didn't start from a fresh install in that video I think... But that wasn't from the debug window Also I was never taught project managment / workflow skills, programming has *always* been a hobby for me, I've never ever worked in the IT industry. Everything you see I kind of learned by myself (or via some googling) . As far as my management of DMP, there's not all that much to manage, if a PR works and doesn't make me tear out my eyeballs, it will get merged. If I have a problem with the way it does something I'll mention the change needed, but it's good to catch me in IRC, I'll always answer questions there quickly if I'm online @snashkille: Leave DMPServer bound to 0.0.0.0 in the server settings file, connect to it with 127.0.0.1 locally, and give other people your public address. I should probably add www.canyouseeme.org to my sig, It's handy for testing. @dsonbill: The hardest part is waiting for KSP to restart... @inigma: I'm a stock player when I actually do play KSP, so I have the stock parts there still. The wait is bad, but it's not too bad if you can get a lot done in one go. The hardest bit is figuring out KSP weirdness where I constantly need to restart KSP to try and make some variable value match what I expect. @SeventhArchitect: DMPServer doesn't really do "kspy" things, nor does it run mods. The plugins folder is specifically for DMPServer plugins, of which there aren't many, and almost certainly no mods use this system yet. DMPModControl.txt just tells the clients what mods they are allowed to have and what parts are allowed to get synced to the server. A mod control file can be automatically generated from your own GameData folder (It hunts for all the parts and part adding mods), you'll just have to move it from your KSP folder to your DMPServer folder. DMPServer doesn't work like a minecraft server - It's mainly a message relay that allows syncing to the latest subspace (It only holds the latest state), not a server side physics simulator - which is why it should use nearly no CPU and nearly no ram, only bandwidth, and is why you don't put GameData mods in the DMPServer plugins folder . @kasperhangard: Shared* are all the same feature, and it will be built on top of groups, so you can have different groups with shared science/funds/contracts/kethane/etc, basically anything in the Scenarios/ folder of DMPServer. I'm a co-op player (as opposed to a pvp one), and this feature has been long overdue . I might be able to help with debugging a freeze though, but you'll need to catch me on IRC. @Kerbrud: You're correct, DMPServer just sends a list of mods that you need to have or not have, along with the parts allowed on the server. And yes, messing around with gamedata is a pain, which is why I will one day port KMPModClient (which basically connected to a server and did automatic shuffling between GameData and folder called GameData-cached). Some people may use multiple installs or gamedata folders for this purpose. If there's a mismatch, you get this: http://godarklight.info.tm/data/kmp/darkmultiplayer/darkmultiplayer17-failed-modcheck.png

@StarDrive85: In about 1-3 hours after I get my hands on 0.25 until there's a version that "works" for dev. As for specific 0.25 bugs, that could take up to a week. However, I'll have to encourage people to use dev for the time being, I'm not releasing 0.1.6 until I nab all the features I want in DMP, and I'm estimating the end of october (Plus, releasing the "finished" version of DMP the day that KMP was first release would be nice). Since the media group has released the KSP 0.25 videos, I suspect it will be out within 1-3 days if we look back at KSP's history . @Dacan: Rotations are currently set, but they aren't interpolated. I'm not familiar enough with quaternion interpolation with angular velocity to accurately guess how that would work. Would probably be a worthy goal to figure out though. TCP isn't faster than UDP because of the speed-of-light / processing delay you still have. TCP sucks when you try to stuff a lot of things down the pipe, but you'll get the same thing if you implement TCP-on-UDP which is what most games / libgren does . I try my best to keep the pipe empty - DMP only queues 1 message at a time, and there's injection points for high priority messages every 8kb, which shouldn't be too much lag for most people. @SeventhArchitect: Most mods should just work, but you'll need to edit (or generate with Options -> Generate DMPModControl.txt) a DMPModControl.txt file for the server that includes the mods / mod parts. If you have the same game data folders and this is a private server, you can also turn mod control completely off. By default, vessels with modded parts will not get synced to the server. There's also an option that completely prevents launching. @tetryds: Most of the mod support burden actually lies with DMP - As DMP mostly works like single player (except other vessels spawn in and move around...), then most mods should (and actually do) work. The only place it gets iffy is when 1 client sees something that didn't happen in the other clients frame of reference (physics are calculated client side still - but just updated from other players) - Which is exactly what happens with weapon mods, as they likely kill individual parts from a ship, when the other client never saw them fire in the first place. There is upcoming mod-support for mods to hook into already in dev, but DMP needs a relicence (unlicence is my current proposal) in order to allow non-GPL mods to use DMP's network system. And I also need to write a reflector for a soft dependency on DMP @snashkille: While Senshi is completely correct about port forwarding, that's not what your error log is telling me. "System.Net.Sockets.SocketException (0x80004005): Only one usage of each socket address (protocol/network address/port) is normally permitted" tells me something is already running on that port. If you're already running DMPServer, that explains it and it likely was a port forwarding problem, otherwise you'll need to pick a different game port in DMPServerSettings.txt.

@Dacan: I assume you're mainly talking about surface updates (<10km altitude) with respect to lag - While in orbit DMP uses the orbital parameters where position/velocity can be accurately calculated for the future (KSP's methods do this for me), but DMP also syncs the engines and time which helps keep spectators seeing the same thing as the pilot. As for surface positioning, DMP does not yet interpolate along the velocity path which is why you see planes jumping back and forward in flight - This would be a very easy feature to implement and should help things a great deal, so long as I don't go interpolating people into the ground . Disappearing is likely caused by KSP killing vessels in atmosphere as they get put on rails, which isn't something I have a lot of control over, and although the "pack" range can be increased, it makes the game lag quite badly when mods do this. Vessel explosions can be for various reasons, DMP's positioning isn't quite perfect yet and does cause clipping with the ground quite often. The other reason things could feel slightly "spikey" is DMP's exclusive use of TCP, rather than UDP. TCP isn't a terrible choice due to the way KSP works, it's not significantly more laggy than UDP is, and after I correctly sync angular velocity and interpolate the position things should mostly be ok *crosses fingers*. Also, 200ms is not a large amount of lag when you live in Australia . darklight@chrislinux:~$ ping forum.kerbalspaceprogram.com -c 4 PING forum.kerbalspaceprogram.com (85.236.103.14) 56(84) bytes of data. 64 bytes from hosting01-ksp.multiplay.com (85.236.103.14): icmp_seq=1 ttl=48 time=378 ms 64 bytes from hosting01-ksp.multiplay.com (85.236.103.14): icmp_seq=2 ttl=48 time=369 ms 64 bytes from hosting01-ksp.multiplay.com (85.236.103.14): icmp_seq=3 ttl=48 time=377 ms 64 bytes from hosting01-ksp.multiplay.com (85.236.103.14): icmp_seq=4 ttl=48 time=356 ms --- forum.kerbalspaceprogram.com ping statistics --- 4 packets transmitted, 4 received, 0% packet loss, time 3102ms rtt min/avg/max/mdev = 356.464/370.596/378.693/8.931 ms

@Harrekin: It was a rather simple logic error, vessel/vessel dockings that didn't involve the controlled craft have always been broken, but generally that doesn't happen in stock which is why I missed it @X.C.M: As soon as groups / shared science / permissions get pushed to my master, it will instantly be available in the dev builds. There's still quite a long way to go with 0.1.6 though (as in 0.25 will definitely come out before 0.1.6), so when KSP updates people might have to jump on dev for a while until I finish the feature set for DMP. And as always, everyone is welcome in #DMP on espernet, but being Australian I might be a little hard to catch. The more people we have on dev the more eyes we have to spot bugs, which means the more stable 0.1.6 will be

@inigma: The KAS bug should be fixed in DMP-dev, DMP was not picking up dockings that didn't involve the active vessel, so the old vessel was never removed from the server. KAS just makes the problem more apparent as when you hook pipes together, that's a docking between two other vessels, not involving the active vessel (which would be the kerbal). If you lined up 2 vessels and switched to a nearby craft, the same bug would have happened - So this was definitely DMP's fault .

@Yalin Hawk: I'll have to poke the universe creator, I think it's alright though. It's possible that revert might have incorrectly killed vessels if you're using 0.1.5.5, I'll bring it back for 0.1.6. @Harrekin: DMPServer doesn't have a backup utility, I assume you're talking about the server GUI from masteryx (which is something I haven't really used) @Rekrahttam/MaxDeath10x: DMP disabled revert due to the way KSP works - If I can detect a revert correctly then I should be able to go back in time and play through the updates (which means syncing *back in time* might become possible again ) @inigma: I'll check out KAS today, I have a day off. @kasperhangard: That's kind of the idea of MCW_LOWEST, but I never implemented it. I'll go over the warp system for 0.1.6. So long as you leave the safety bubble, you should see all the other players from the past or present. You can sync to the future in the player window. @PurpleBlurpL: 1 can be easily enforced with a no warp server . Might be fun to try as a stream event... @dcompart: Currently, players in the past can sync with players in the future. Syncing the past will be possible if I bring back reverting. I should be able to, just needs some out-of-the-box thinking to resolve some paradoxes @spiritplumber: Thanks for pointing that out. I should really be testing things more before pushing them into release

@chiefymuc: It seems there's a conflict with window ID's somewhere although this *really* shouldn't be happening as I'm using a unity call to guarantee a free window :-/. I've yet to reproduce it, but when I do I'll poke it some more. As a pre-emptive measure I've added a constant offset to the DMP windows so they may not conflict anymore (but this could be wishful thinking...) @therealcrow999: All that does is copy your saved vessels (the .crafts) to the server, that shouldn't help with this bug, which suggests it strikes randomly and you're experiencing a co-incidence. @Dumdidldum: In 0.1.6 (the next version). Story incoming for everyone! @inigma: For a mod developer, I'm a surprisingly stock player and I actually haven't used those other mods (although I'll give into the rest of the modding world one day ). I do know KAS causes explosions with players around, and TAC-LS is time dependant so it's most likely going to have problems with DMP's subspace mode. @The_Rocketeer: Yes and no, depending on a server setting and who is connected. 'keepTickingWhileOffline' (default online) is a side effect of how subspaces are locked. Subspaces are basically epoches that reference the real server clock time, and lock a universe time to that point. It then counts the difference in seconds to the actual server clock time (figured out by an NTP-style approach) and multiplies that by the subspace speed to get the correct universe time. When keepTickingWhileOffline is disabled, it simply relocks the last locked universe time to the *real current server time* when the first player connects. Short answer: Contracts are currently individual but the time is shared, so that player will get expired contracts if they disconnect/reconnect when someone warps far into the future. Shared contracts are coming for 0.1.6. @inigma: "a DMP duplicate vessel detection auto-removal-of-the-newest-vessel workaround?" - PartKiller was supposed to take care of the bits that break off, but I'll have to check how KAS does things. If it's an actual docking then it's DMP's fault, if it's not, then it's unsolvable :-/

@Steven Mading: The GUI's will need to be cleaned up in a nice way in order to do this still, but it sounds like a great idea. I'd still prefer spectators to follow pilots automatically though, perhaps an option that's enabled my default. @PaidLeber: Probably not, otherwise we'll run into the thousands-of-kerbals-to-sync problem again . KSP's crew generator starts to hang the game with many kerbals as it's a while loop that keeps generating names until it's unique, the VAB and runway also lock up for 30 seconds with ~2000 kerbals in the crew roster... @Kenobi McCormick: Dracofunk is correct - DMP handles mods quite well because it chooses *not* to handle them . DMPModControl.txt on the server is the file that specifies what can or can't be in GameData/ - You can autogenerate it by pressing on the options window in DMP and clicking "Generate DMPModControl.txt", then moving it from the KSP folder to the DMPServer folder. Note: Plugins/ are for server side plugins, of which there are *very* few at the moment. Eventually it's going to give other mods the ability to change DMPServer or DMPClient as they see fit. @dsonbill: I have no idea how you've picked up on this so quickly, but that chat rework looks *very* close to what I had in mind (thanks). Also "COM$" is going to create many questions for DMPUsers, it's probably best to use @Server or @consoleIdentifier up the top right side of the tabs next to "Leave" (to seperate them from PM's), or use no prefix and push the server console I/O through it. I'll need to do that rework though . Also player names are authenticated, so any ClientObject that is authenticated = true *is* who they say they are. We'll have to change that authentication system though, it's just a *tiny* bit dodgy atm . Also, before I accept your PR I'll need to make you aware of something first, DMP needs a licence change because any mod that tries to use DMP-dev's plugin system *has* to be either a GPL project, or considered a GPL project (due to linking), which is simply ridiculous. I'm leaning towards LGPL as it does what I want (keeps forks open so I can steal good bits back , I don't care about attribution), although there's also arguments for MIT/BSD, but I'm not as fond of that licence. When I write the plugin reflector that will be public domain so people can incorporate it if needed. Also, definitely come hang out in #DMP, I usually reply much faster there . (Everyone is welcome - #DMP isn't just for me, other devs or server admins)

@dsonbill: Commented on your PR, it's a good start, but DMP code is by no means fantastic or neat . It gets a little crazy in DMPClient where i have to work around KSP weirdness, and I'm starting to suspect KSP weirdness is a direct result of unity weirdness (positions / velocities / rotations are all float instead of double, hence KrakensBane and FloatingOrigin) @longbyte1: Thanks for helping @Korkin @Steven Mading: Forcing yourself into spectate mode seems like a good idea, as it also gives the ability for pilots to release control and give it to a spectator. Adding this into the 0.1.6 todo list. Slaving the clock sounds like an important feature to have, spectators should automatically slave to the pilot as well... The warp system will get some love (a near-rewrite perhaps?) for 0.1.6. @inigma: Thanks for pointing that out, however instead of shipping a persitent.sfs like KMP, I should just be able to create a new Game() object in KSP and save that to a persistent.sfs file. I'll knock that one over today as it's beyond easy to fix. @uncle natzer: As far as I know there isn't any hosting companies that actually host DMP (I'd be very surprised as there's a ton of competition already with the public servers), but maybe one of the other public server admins may host one - join IRC (link in sig) and ask DMPServer should be very light on ram and CPU (it should constantly look idle), but bandwidth has always been a near-exponential thing as it has to relay nearly every message to everyone else.

@Steven Mading: 1) Sounds like a good idea, but I'd preserve the player name click -> PM functionality. I'll clean up the GUI's for 0.1.6 too I guess. I'd also like to implement the flightgear style chat while the window is minimized (basically press ~ and what you type will come up in the yellow text). 2) Hadn't noticed this one, thanks for letting me know . Also, I will also save the window positions and also make sure they can't be dragged off screen, some people have ran into that one before. @Everyone, 0.1.6.0 is going to take me a while, I basically want to get DMP feature complete and kind of go into bugfix only mode. but I've still got my seasonal job which is why my github history has taken a massive nose-dive. It seems to be a constant month or two away at the moment . After 0.1.6 comes to life and proves stable, DMP 0.2.0.0 (I can't count to 7) will be a re-factor / cleanup and will be tagged as 'beta'. After that, I'll update DMP for new versions of KSP until squad releases KSPM . EDIT: 0.1.5.6 is available, I'm tagging this now so I can make a real start on 0.1.6.

Inigma, Thanks for letting me know . I personally support the idea of modpacks as it makes modded servers for DMP so much easier, players generally like the ideas of mod packs, but other mod developers are generally against this. The major reason is an issue with support - Mod packs won't update until after a mod does, and other mod developers may not like hunting down bugs that don't exist . Also I believe DMP has an incompatiblity with KAS when things are connected together, but that's likely due to the position / rotation kick that DMP currently has. Explosions, explosions everywhere!. After 0.1.6 comes out everything should be good (*crosses fingers*) Also, prepare yourself for when #kspmodders gets hold of this, hopefully this thread doesn't end up like http://forum.kerbalspaceprogram.com/threads/91939. EDIT: Just to make it clear, I will provide support for DMP even in a mod pack, and it's almost certain that a mod pack that includes DMP will likely show up on the public server list sooner or later .