AlonzoTG

... seeing lots of: [EXC 14:34:44.641] NullReferenceException: Object reference not set to an instance of an object PipeLines.CivilianPopulationRegulator.applyCalculations (Double dt) PipeLines.CivilianPopulationRegulator.OnFixedUpdate () Part.ModulesOnFixedUpdate () Part.FixedUpdate () [EXC 14:34:44.810] NullReferenceException: Object reference not set to an instance of an object PipeLines.CivilianPopulationRegulator.applyCalculations (Double dt) PipeLines.CivilianPopulationRegulator.OnFixedUpdate () Part.ModulesOnFixedUpdate () Part.FixedUpdate () [EXC 14:34:44.847] NullReferenceException: Object reference not set to an instance of an object PipeLines.CivilianPopulationRegulator.applyCalculations (Double dt) PipeLines.CivilianPopulationRegulator.OnFixedUpdate () Part.ModulesOnFixedUpdate () Part.FixedUpdate ()

ok, I went to my base today, same one mentioned earlier... I had left the farms disconnected from the rest of the base because I was thinking about removing the part from my install, just to keep things clean on my end. When I went to the base today, one of the farms had been rescaled to the larger size and now overlapped a bunch of other equipment causing a cascade of explosions that basically trashed most of the base. =( I dunno... -- goes ahead and removes the part and see if the save loads before the explosion.

... After thinking about this for the day, investigating a few things I think the way forward for me, and for a lot of people, is for me to write a compiler for PlanetFactory configs. This should satisfy the author's assumed desire not to have other people taking over their stuff and it should make transitioning to this mod much easier for the users. I think it should be possible to borrow PF code to read in the existing configs, then filter the configs against the game as it exists today, discarding all obsolete fields, ineffective settings, then generate MM configs for Kopernicus... So what I needs from youse is a recipe for a "well formatted" Kopernicus config, basically documentation, so that I can make sure my code generates the cleanest configs possible.

Hey, I am about ready to give Kopernicus another try becuz: 1. Planet Factory's latest update doesn't exactly work -- spews about 13,000 error messages (actual count!). 2. It appears that Kopernicus has reached a level of maturity that rivals, and in places, surpasses PF. 3. The general direction of the footprints is in the direction of Kopernicus. Okay... Now my problem is that I've pack-ratted an unbelievable amount of content for the PF mod, much of which I'm quite fond of, only a tiny fraction of which appears to be available for Kopernicus at this date. Any suggestions?

yeah, I'd like that, -- see if I can merge it with the changes I've made. Thanks for explaining pipelines, sounds like a great concept for multi-biome harvesting.

okay, kewel.... Here's a snapshot of what I had built: In somewhat related news, I've been trying to fix up that lab from ur StanfordTorus kit because it's one of my faves, I got the exterior texture back... Things look great from the EVA view except for the kerbal's rotations but the IVA view is foobar, I had it working before but it got messed up when I started using Unity to work on this. IVA kerbals are in position but the floor and everything else is rotated 90 degrees and pointed toward the north on the KSP pad. Wishlist for that part: Raster Prop Monitor live orbital display over monitor screen. Get the damn airlock to work again, I totally botched it... Radiation-proof glass like you find on airplanes, real spaceships. general fit and finish improvements. I'm not sure what your workflow is... By the way, notice the exception report in that screenshot, I'm seeing thousands of "PipLines.[C.P.R].ApplyCalculations exceptions...

I am wrapping up my first major city build, design population is around 304 LGM. I am on a mod-planet that is in Kerbin's orbit but no civies have shown up. I accidentially built with the orbital dock but swapped in the ground dock by editing the save, maybe that caused a problem... Anyway, a universal algorithm to detect whether u are orbiting something is: boolean recursive_is_orbiting(local_body, tgt_body) { if(local_body == tgt_body) return true; if(local_body.parent == null) return false; return recursive_is_orbiting(local_body.parent, tgt_body); } 2. Getting agriculture bootstrapped was kinda tricky because I didn't have any bioMass, I had to ship some up with kerbin using a dinky little multi-function container I had. My life-support tower is overbuilt by a factor of 100 because I wanted to have time to fix things before running out of supplies. Lag is pretty horrible at the base. =P

Had to look up Bob Duff. Properly implemented, OpenCL pragmas should be completely transparent. A standard debugging procedure is to simply turn them off, test, then turn them back on. The program should behave exactly the same except faster. This actually works, I've even used it! =P There are some tricks in that you need to define what is thread local and what is shared but otherwise the pragmas are invisible to the program logic!

i'm having more fun with base building...I managed to build phase 1 of my base with karb and ore. =) Ok, that's out of the way... Phase 2 calls for six Netherdyne farms. The game is already starting to lag... =\ These farms are of satisfactory size but have completely flat bottoms, which makes perfect sense for something intended to be built, and remain, on the ground.... The game is already starting to lag, and I want this thing FINISHED so I'm going to skimp on parts as much as possible. So I drop the farm down with only a handful of KAS ports on it... When I complete the build, it intersects with the survey stake and is sent a few dozen M into the vacuum. So I immediately switch to MC and delete the stake, and then switch back to the base proper. When I get back, the farm is now about twenty M to the East of where it was, or overlapping with major parts of my base, needless to say there were some rather spectacular fireworks, parts being thrown 27km into suborbital trajectories.... Jeb was amused, I am not. =\ This is bizzare because when the Kraken possessed the 20M diameter storage tank previously, it caused the thing to walk westward towards my base to crush it... So, clearly the Kraken works against me whether I build my crap to the east or the west... =( The other parts of my base worked because they were standing on landing frames, ranging 2.5m to 10m, one colossal piece weighing, for some reason, 8,000 tons sitting on five HUUUUUUGE landing legs. I suppose I could put legs on the farm but that doesn't make sense... =\

I've been asked OOB to do a bit of a data dump on what I know about this mod. Keep in mind that I haven't really gotten down and dirty with it, I have copies of several hundreds of planets that I, at my most ambitious moment, used to populate a galaxy with three additional stars... (Star Systems, which I have been trying to fix myself, plus KSP's severe resource limitations ended that). 1. File types: The system file that lists the planets in the current pack and whether the user has activated the system. eg: System { enabled=true planetName=planet1 planetName=planet1a planetName=planet1b planetName=planet2 ... } The problem with this sistem is that the menu on the loading screen does not have a scrollbar!!!!!!!! The second file describes a specific planet. Some planets are defined internally to the mod, which is bad legacy code but well... Minimally, a planet must have a PFBody description (lesson learned = PF* is processed directly by planet factory) Here's a simple one at random: PFBody { name=Freezer templateName=Jool flightGlobalsIndex=700 } CelestialBody { bodyDescription=An icy, cold gas giant. } Orbit { inclination = 2.0 eccentricity = 0.2 semiMajorAxis = 32850237728 //46000000000 LAN = 0 argumentOfPeriapsis = 0 meanAnomalyAtEpoch = 0 epoch = 0 referenceBody = Dolas } Notice the "CelestialBody" This is an important class internal to KSP. Here we replace the data from the template planet with a new description. Basically, once the template is loaded, we use the statement above to overwrite the data we wish to change. The fields available change with each version of KSP so there is no substitute for checking with the interface descriptions. =(((( Here, I'm explaining the general principle. Now the orbit.... As a user, this is what you will have to be fixing up. It is very important that the SOIs of celestial bodies NEVER cross. If they do cross, the physics engine breaks in profound ways. -- you can fly through planets! =P I made up a spreadsheet that uses titius bode-like methods to generate appropriate orbits for the major sattelites of a given planetary system. As you can see, I was monkeying with the above planetary orbit... Now, the data files. 1. Planets are always spherical and always have spherical coordinants. However, planets can be deformed a number of ways using PQS settings. Planets are too big to define so most of the data that goes into a planet comes from these PQS settings. A planetary map and mesh provided by the user is a "broad-brush" outline of what the planet looks like. Macro-scale features such as equatorial bulges and stuff are set up in configurations. Also, the ground textures and procedurally generated fine details are generated on the fly as you approach the planet. You must also specify how this will be done for your planet. (most people just live within the limitations of a template planet....) Another feature is decals, where you can apply additional geographical features such as volcanoes to your maps. The planets which come with Planet Factory were carefully designed to illustrate a broad, if not exhaustive, suite of examples of each of PF's features. Sido's Urania planet pack is probably the finest example of planetary artwork I know of. The orbits are completely broken (see above) but the planet designs are outstanding in all respects! I do not know much about the .bin files, they specify some kind of mesh data for the planet, poke around in the source code for more information.

reading the original post for a change, some interesting things come up... 1. OpenMP -- u can tag loops and stuff for multi-processing. 2. C++AMP which might be cheap or free way to use OpenCL-like computing, I hear that is a feature in Clang. But I definitely hear you about solving the maf first. =)))) GJ!!!!!!!! In other news, I have some fairly perverse planetary systems in my customized install... On top of that there's StarSystems -- GALACTIC Kerbal. =P Anyway, being a moron, I'm using things like Titius-bode to make plausible planetary arrangements...

hmmzers! I should bookmark this! Ideas that have probably already been suggested: OpenCL Approximations of distant objects (At distance D, jool and it's moons becomes a single object jool + moons, obviously this would happen for pol and bop before laythe and tylo) Even though, techincally, ksp is multi-threaded, the cpu utilization is horrible, leaving 2.8 to 6.8 unused CPUs on any reasonably game-worthy PC...

Here's a brief explanation of what starlight switcher does, it's really very very simple. The issue is that the game has only one slot for t3h star, in the Psystem class. 1. Keep a list of light-emitting stars. 2. set up an event handler to check for the nearest star. 3. On trigger, re-calculate the nearest star and put that in the single star slot, tweak a few light settings here and there... (issue = compute color and saturation values for stars. ie close up all stars are white but far away they take on their characteristic hue....)

Regarding tutorials, that's kinda tricky. I've been using this mod for a long time, mostly as a user who sometimes needs to shuffle planets around so everything fits... Unfortunately you need to get into the source code a little. Get the free version of MS visual studio then look at "assembly-csharp" in the ksp-data directories, From there use the object browser to find the "pq*" classes. Basically PF lets you send parameters to these classes to set their properties. I think there's also a PQS editor out there that lets you play with this stuff in-game, note the settings, and then configure your .cfg accordingly. The process is a bit more involved than what I can fully explain here.

To re-cap earlier in the thread, the issue is that StarSystems currently does most of it's work far too late in the game setup process. (Planet Factory is aware of this and even sets a flag to prevent additional planets being loaded on it's end after this point!) Basically what happens is that the planet moves but the ship does not. Anything in a solar orbit (for any sun) should be OK because stars are loaded long before planets are moved. I don't have SS running on any of my games right now, so plz test this.

hmm, I'm still a Planet Factory fan myself. I guess we should start by evaluating what value Star Systems brings and see what we can do about improving how that is delivered. In rough order of importance: 1. The starlight switcher. -- without it solar pannels become useless; this is the only mod I know of which provides this. 2. Easy and modular configuration of glactic-scale organization. 3. Easy and intuitive way to design stars, ie by mass and fundamental properties rather than back-configured from surface gravity or some other secondary property that doesn't really make sense on a star.

yeah, and the drill does not work when placed right up next to the deposit if it is KAS connected to a base that is in a different biome. =\ -- made 20M procedural parts storage tower to hold teh waterz. The waterz are in a 60 degree cliff face...

That science lab has some SERIOUS geometry problems. =( The airlock is so broken right now that kerbals die the instant they attempt EVA, I have no idea what's wrong. =( The thing has pervasive object placement issues. The most serious of which is that the IVA view is rendered after the EVA view so, in this model, you can see right through the central core... maybel i'll just move that to the other model... Tragically, the artwork shown is not visible in IVA -- ...? (guess where I got that!) Anyway, those kerbals seem to be having a bit too much fun. =P I can see why the author didn't want to maintain this anymore... =\

I tried to overhaul the object structure in the part and things totally blew up on me. =(

I think I have an "And" problem... I'm trying to make the IVA AND EVA models fit together in blender AND the game so that both models work in flight AND IVA mode... Obviously, I have too many ANDs... Is there a guide for setting up transforms and dealing with coordinant system issues?

For a 25M torus, I'd suggest just adding a spin animation to the model, like a part in LLL does. now for the BIIIG torus, the spin-up would be accomplished by two stock nuclear rockets rotated 90 degrees in pinwheel formation, that will, after maybe 20 minutes, achieve the desired spin. I have a radial attachment port that should do the trick there. Then you would need a spin decoupler to attach a docking rig... Preserving the spin through scene changes/save-load might be possible with a plugin applied to the spin decoupler.

good news this time, the construction of a 40Mx20M consumables tower worked great...

I started working on a different part, "Station Habitat" from stanford Toruses, The model has some -- problems --, I cleaned up a lot of stuff, re-exported, put it in the game, looked better, wanted to come back today to continue the job but it was metza-foobar, I'm getting a bunch of index out of range errors on the importer and there does not appear to be a way to cut and paste the stupid blender pop-up, or even keep it visible for a non-miniscule amount of time. =\ goal of project was to remove all interior components from the exterior, clean up all sloppyness. then remove all exterior portions from the interior view and make sure all the objects are properly nested (the interior view loaded as rotated 90 degrees) then go back and polish everything.

Y0zers, I've been playing around with the StationHabitat module. I managed to do quite a bit of damage to it... The models are now roughly half as complicated as they used to be. More remarkably though, the kerbals are now visible from flight view!!! Numerous glitches and problems with the model's encoding remain. =( The airlock is broken because I had to clean up a bunch of detritus around it. I'm extremely noob with Blender and there are a lot of things that need to be fixed and cleaned up before this new version can be released, if you want to help me with this, I can send you my WIP. (Requires Raster Prop Monitor).

I didn't accomplish much today, I moved a few things around in Star.cs and that's about it. I found a very telling flag in Planet factory "too late to add planets" and it is set right before StarSystems does most of it's work... =P This is going to be hard. =\