allista
-
Posts
2,208 -
Joined
-
Last visited
Content Type
Profiles
Forums
Developer Articles
KSP2 Release Notes
Everything posted by allista
-
Nope. I'm sorry, but I simply lack the resources to support multiple KSP versions that differ considerably. MJ still does a lot of things that TCA cannot do. Especially many of the orbital maneuvers
-
Yep. Thanks for the video! Will be fixed in the next update.
-
Neither FAR, nor RO are explicitly supported in that no tuning has been specifically made for them. In the past they worked together with TCA fine, but that was in ksp-0.90. What you describe seems to be the result of overtuning of TCA's attitude control, which I'm currently reimplementing completely. So there's a chance the problem will go away with the update. This does look like a bug to me. I'll only see the video as soon as I'm home, though. Then I'll be able to tell more. Edit: looks rather strange. First of all, I don't see any of the engines to fire up, as if there was no fuel. Prior to separation, boosters should have been ignited and fly off by themselves. Also, main throttle goes up after first stages were activated; but even then the remaining engine does not work (at least I don't see it does). Which version of TCA do you use? And can you share the .craft file of this rocket so that I could test it?
-
There's a way to do it, but some text file editing is required, as there's still no GUI for it, sorry Here's a part that describes what needs to be done:
-
So far there isn't. But I'll do it: https://github.com/allista/ThrottleControlledAvionics/issues/43
-
Technically that's possible. Even half dynamically, using part tags. But I have a gut feeling that if I just recalculate the landed state, things will go the wrong way. For instance, I will have to recalculate effective vessel weight to exclude landed parts; and ship bounds to get a meaningful vessel height; and I'm sure there are things i haven't thought of yet.
-
This question was recently asked, so I made a video tutorial on the matter: In short: VTOLs to thrust & maneuver, the rear to manual, and then you need to configure action groups as described in the video. Cheers! Good question! Unfortunately, nothing could be done about it right now. Winch is a part of the vessel, it splashes and the whole vessel is declared splashed by KSP. TCA just reads the flag. To have this correctly TCA needs to know the difference between the winch and other parts. This could get awfully complicated...
-
Using tags is a great idea, thanks! I thought I make a dedicated PartModule with which TCA could communicate, or even a small part that would tag a vessel as a landing site... but a part tag may indeed be enough
-
Stupid me, really Thank's a lot, that did help. -
It does not, because linux version of Unity is still a beta with versions like 2017.1.0b. But there's no errors in Editor.log, either. I'll try to setup the 5.4.0p4 in POL and see how that works...
-
Nope, checked it first of all. It does look like SIGSTOP or breakpoint. Editor connects, process stops (black screen, no CPU activity); Editor disconnects, everything resumes. The only notion of a similar problem I found was about an Android player that similarly freezes/unfreezes on connection, but that was an old thread with the "update version of Unity" kind of solution. Edit: debugging with MonoDevelop works alright, though.
-
Having problems with the Profiler on Unity-on-linux. As soon as the debug LinuxPlayer (renamed to KSP.x86_64, etc.) runs and the UnityEditor connects to it, the Player stops (as if at breakpoint). If I then switch (Profiler tab) the Connected Player from LinuxPlayer(localhost) to Editor, the Player resumes and runs as if nothing happens. Choosing it again freezes it again. Have someone any experience with the Profiler on linux and, perchance, knows a fix/workaround?
-
1. you're trying to use an old version of TCA. 2. and that's the real cause of the problem, you're trying to use an even older version of AT_Utils which is incompatible with KSP-1.3 and with the version of TCA you use. You should install the latest version TCA from the SpaceDock archive: https://spacedock.info/mod/198/Throttle Controlled Avionics/download/3.4.2 Corresponding version of AT_Utils is included.
-
As you can see, the exception occurs inside KSP.IO.PluginConfiguration class, to which I have no access (I don't directly handle file IO). But since it's an old bit of code which hadn't caused problems before, I would guess IO is handled correctly there. Also, as everything is handled in the main Unity thread, there can be no simultaneous access attempts. The only other explanation is that the lock on that file is held by another process. Maybe previous instance of KSP is still running? I know that's hardly probable, you never can tell. In any case, a reboot should either solve the issue or show us that there is indeed some nasty problem.
-
Waypoints do account for the water (and are always set above it) and for static objects (buildings, etc.); and you can always edit a point manually -- either by mouse-dragging (left-click+drag changes location, middle-click+drag changes altitude), or via editor window (Edit button in the waypoint list, or right-click on a waypoint on the surface). But, I would guess that's not what's happening in your case. The boat you're trying to land on is still a vessel, not a static; and TCA has the Collision Prevention System (CPS) module that tries to avoid collisions with other vessels by automatically correcting your course. It's CPS pushes your craft sideways, away from collision with the boat. Currently, there's no automatic way to detect if a vessel is actually a landing target (or just sits on the way). So to land, you will need to go to Advanced Tab and disable CPS (using corresponding button). Don't forget to re-enable it afterwards, though It did come about: in Editor, when you build a new ship, you can open TCA window and push the Modules button; there you will see a set of tiers of modules with their dependencies on each other shown. You can deselect any of them; dependencies will be disabled automatically; and controls (and even whole tabs) of disabled modules will be absent in flight window.
-
A minute of narcissism Random Automatic Tests of remastered orbital autopilots URL: https://www.twitch.tv/videos/154384791 Program: Jump To a random surface target Place: Mun Ship: simple rocket with 4 side-attached engines and a reaction wheel Running time: 10h 23m, 119 repetitions Results: 0 failures/crashes Yes, but are they strong enough to handle its moment of inertia + drag on wings and tail? Without enough forward motion control surfaces are no help. Anyways, I don't know about the second autopilot and their co-existence with TCA, but I'm sure it's a bad idea to enable both at the same time. And TCA needs static control authority for hovering, no way around it.
-
What pitch-torque providers do you have there? Because it looks like it tries to combat aerodynamic forces and fails due to insufficient pitch authority.
-
Honestly, I don't know. The last try to fix your problem was published two versions ago, prior to KSP-1.3; I've tested and retested it, it worked for me. But for you it still, incomprehensibly, did not. Then I stopped developing for KSP for two months, then 1.3 came out and I had to change some things, then the last release. The only thing I can suggest now is to test it again and see.
- 1,632 replies
-
- 1
-
-
- part count
- storage
- (and 1 more)
-
Should work by itself. You did arm it, right?
-
This time I don't see any straightforward problems, apart from the presence of OneTimeResourceConverter.dll in GC/Plugins folder, which shouldn't be there, but also shouldn't cause trouble. Remove it, and if it won't help, try a clean install with only GC(AT_Utils included)+CommunityResourcePack+ModuleManager. If that crushes, something's wrong with my dlls; if not, there's a mod conflict of some sort.
- 1,554 replies
-
- 2
-
-
- ship construction
- launchpad
- (and 1 more)
-
Logs?
-
You're missing ConfigurableContainers folder with core config files that are distributed with AT_Utils. Reinstall AT_Utils: https://github.com/allista/AT_Utils/releases/download/v1.4.4/AT_Utils-1.4.4.0.zip
-
Yes, that's my photo (somewhat old though), but I'm no German and I don't watch TV; none at all
-
Hey, nothing's late about it: a hobby is a hobby Ok, so I'm leaving everything concerning mass/cost balance as is. And I'll add the :NEEDS[!KolonyTools] stanza to the ISRU patch (is it the correct way to check for MKS, though?). RE spawning: this is still the most troubling part for me. The ground is simple, as basically it is already implemented. The attachment: in orbit is prone to blow things up if the attach node becomes obscured by other parts of the station. Or if something flies by. on the ground has the same flaws + gravity will force to always attach the container on top, vertically. The dry-dock method is, essentially, what the Hangar does; so I will need to either move some things from it into AT_Utils (which is tedious), or we'll have to bundle Hangar.dll too. Except that Hangar stores and launches ProtoVessels, not ShipConstructs, so I'll need to adapt to that too. And that will require, as you said, several dry docks of different sizes; and some larger kits will still be restricted to ship-from-Kerbin only. Transport: agreed.
-
First question bothers me as well. Probably so. It would also make sense logistically: you'll have a hatch from inside the AssemblyLine into the docked kit container and would simply put everything in it until it's full, then undock it. Second: correct. The reasoning goes like this: Kit production takes much time, and completed kit takes some space. A kit may be well bigger than the AssemblyLine. Kerbals would work on part-kits in the line, but where would they put the products? So you need a storage for parts of incomplete kit, and a place for assembling them into the box. When you spawn something, it's best to be sure that nothing gets in the way. As with the kits themselves, inflatable parts are the best to achieve this, as they can physically push objects away. Unlike the kit, AssemblyYard would be a very small and flat thing when you spawn it, but then it can inflate to the size of the future kit. So it's also a matter of aesthetics. Third: for the same reason of respecting the kraken's slumber
