Jump to content

allista

Members
 View Profile  See their activity

  • Posts

    2,208

  • Joined

  • Last visited

 Content Type 

Everything posted by allista

  1. allista
    Version 3.5.5 for Kerbal Space Program 1.7.0 Released on 2019-05-14 Added ability to change UI color scheme at runtime To access the Color Scheme dialog, right-click the GC toolbar button Advanced Tab: AutoSave and Launcher/Toolbar options are now persistent. Fixed RCS balancing (issue #65) In-editor TCA window is now available for vessels with RCS thrusters only (issue #66) Improved landing trajectory rendering using UnityLineRenderer: Lines are thicker and smoother Deceleration trajectory is colored according to predicted hull temperature Other minor fixes. Download
  2. allista
    Version 2.4.5 for Kerbal Space Program 1.7.0 Released on 2019-05-14 Added ability to change UI color scheme at runtime Added "C" button to the tank manager window titlebar which summons the Color Scheme dialog Download
  3. allista
    Version 2.3.2 for Kerbal Space Program 1.7.0 Released on 2019-05-14 Made docked deployment compatible with auto struts and Kerbal Joint Reinforcement Next >=4.0.1 Added ability to change UI color scheme at runtime To access the Color Scheme dialog, right-click the GC toolbar button Download
  4. allista
    I have advanced tweakables, but no autostruts configured. But, indeed, there's two dlls in the KRJn archive and thus in my GameData: And sure enough, removing the KerbalJointReinforcement.dll fixes the problem: both purple joint from the root part and the green joint from parent disappear when resizing starts, then reappear when it's finished, and this time the box stays where it should be. @Rudolf Meier thanks for the help! Now my code is compatible with both autostruts and KJRn. While you just need to fix the distro, I suppose
  5. allista
    This is how the vessel looks like: And here's a crude analog using arbitrary stock parts: Note that the Service Bay does not have the joint to the root part (the top-most fuel tank). No moving parts, of course. And here's the .craft file: https://drive.google.com/open?id=1XOwWKYeCqBt_1JLcHpwFWEQh8QPaWeOe
  6. allista
    In this test I've noticed that the green and magenta lines seem to z-fight each other, and there's one of each color going from the root part to all the others. Results: Between 3 and 4 the short green line connecting the box to its immediate parent (the docking port) vanishes. But the green+magenta lines from the root to the box remain. After 5 the short green line reappears.
  7. allista
    No such a way now, and with current architecture it won't be easy to implement. You can use a dedicated hangar part, which is resizable, or add the AnisotropicPartResizer module to the mk3 parts.
  8. allista
    I've considered the possibility of integration with KIS awhile ago, but it turned out it doesn't have any public API I could use, so the project stuck As for OSI, my answer is of course the Global Construction It may seem an overkill, but it's already there and can do exactly what you want, namely construct a single part in orbit or on surface and deploy it there just like out of a hangar.
  9. allista
    This is a single Vessel made in editor as is. It's a rigid structure, the only thing that moves is this box, which is a leaf of the part tree of the vessel. The magenta joint links the box with the root part. The box moves as follows: 1. The scale of its root Transform is changed a little. 2. The PartJoint between it and it's parent is destroyed. 3. It's location and orgPos are changed so that the DockingNode transform relative position does not change 4. The PartJoint is recreated.
  10. allista
    1) the PartModule on the box (DockedKitContainer) that performs the resizing 2) it's going to the root part; the box is a child of the docking port that it's attached to 3) yes, the same DockedKitContainer module The problem here is that just after the resizing the box is shifted a little, so that a gap is formed between it and the docking port. If the box were to grow instead of shrinking, this would mean part intersection and consequent explosion. Note, that with KJRn-4.0 the shift was much bigger; and the box was moved to the spot where its center was before the resizing. @Rudolf Meier also note that during the resizing each physical frame the PartJoint between the box and the docking port is destroyed, then part.partTransform and part.orgPos are updated, than the PartJoint is recreated and part.ResetJoints() is called.
  11. allista
    While the box resizes, IsJointUnlocked returns true, and each physical frame the vessel.CycleAllAutoStrut(); is called. Here's what happens, though I'm not sure what that means:
  12. allista
    Thanks for the keen observation. I'll look into it more thoroughly. But the first thing that comes in mind is: Between April 28 when AT_Utils-1.6.3 were released and May 9 when GC-2.3.1 was I've switched between different branches of both mods and recompiled them multiple times. And the script that makes the distro uses files currently present on the file system. So in the end, the AT_Utils in GC distribution is compiled from the same source code as the full one from github, but binary assemblies differ. Though I'm not sure exactly how Monodevelop's compiler works to produce this result. As for the additional .dll, like all others in AT_Utils, it does nothing unless used in other mods or configured in dedicated parts.
  13. allista
    Tried it myself. Even tried to implement the IJointLockState interface as suggested by @Rudolf Meier in the OP of KJRn, but nothing helped save for the removal of KJRn itself. So I guess my way of resizing of a container is too hacky to work with any external joint tweaking.
  14. allista
    I think it's safe then to add the KJRExcluded module to DIY kits in GC itself, so that at least the users of KJRc won't have that problem.
  15. allista
    Don't bother, just zip/7zip it and share via gdrive/dropbox/whatever.
  16. allista
    This surely is a bug, thanks for reporting it. Expanding a docked container in space is one of the trickiest thing I've ever done with KSP, and at the time of development it was very prone to such krakenizations. I've made it work on a slightly modded game high around the Kerbin and retested it at Duna, but there's no telling what a rouge floating point error may do to the thing on some elliptic kerbolcentric orbit. So my first three questions are: what's the orbit of your space dock (so that I could reproduce it) the list of mods; I don't know of any incompatibilities yet, but who knows, something may poke me in the eye output_log.txt (Player.log on Mac/Linux) made during the incident
  17. allista
    Here you go: https://github.com/allista/GroundConstruction/releases/download/v2.3.1/GroundConstruction-2.3.1.0.zip
  18. allista
    Version 2.3.1 for Kerbal Space Program 1.7.0 Released on 2019-05-09 Excluded some pure-technical part modules (like ModuleTestSubject and ModuleOverheatDisplay) from DIY kit complexity calculation, which decreases both SpecializedParts and SKH build costs of many kits. Note for modders: the excluded part modules are listed in the IgnoreModules.cfg and could be added/changed using MM. Download
  19. allista
    Ok, so you clearly have some installation issues. Several mods are not compatible with KSP-1.7 that you're running: UmbraSpaceIndustries/ART UmbraSpaceIndustries/Konstruction UmbraSpaceIndustries/MKS HullCamera FlightPlan GroundConstruction 2.2 (which is outdated, there's an update for KSP-1.7) You have the latest Hangar for KSP-1.7, but you're missing the correct version of AT_Utils, so the Hangar itself does not load properly. The best thing is to reinstall the Hangar from Spacedock (which does include correct AT_Utils version): https://spacedock.info/mod/1000/Hangar/download/3.3.5 Another option is to manually install the latest AT_Utils from Github: https://github.com/allista/AT_Utils/releases/download/v1.6.3/AT_Utils-1.6.3.0.zip
  20. allista
    This turned out to be a regression due to ModuleRCS implementation changes introduced somewhere between 1.5 and 1.7. So thanks for pointing this out: I've managed to fix it https://github.com/allista/ThrottleControlledAvionics/issues/65
  21. allista
    ModuleTCA is the core, the brain that does all the calculations and controlling; that's why it is installed into command modules. TCAEngineInfo is just a metadata storage that associates with an engine part and stores things like TCA engine groups and roles. As for FS engines (and any other custom engines for that matter), adding support for them is (almost) equivalent to reimplementation of the core code of TCA, as it heavily relies on the knowledge of the stock ModuleEngine implementation; namely, its interface and the math behind its functionality. Because to work well TCA often needs to predict in real time engine's behavior.
  22. allista
    Version 3.5.4 for Kerbal Space Program 1.7.0 Released on 2019-05-08 Improved Land (from orbit) and Ballistic Jump autopilots with adaptations for new aerodynamic model of KSP 1.4.5+ Fixed stucking in "Testing aerobrakes" phase of the landing when no aerobrakes are present. Other minor fixes in landing logic. Download
  23. allista
    A question! @lampeh proposed an improvement for the complexity calculation in GC, namely Which resulted in the following config with part modules to ignore: https://github.com/allista/GroundConstruction/blob/cab417babf02477f7e2090caf7dcf5bfdefefddd/GameData/GroundConstruction/IgnoreModules.cfg Before I push it into the next release, would anyone like to add/remove anything there?
  24. allista
    On windows there's the KSP_Data/output_log.txt which is much more verbose than KSP.log On Linux/Mac the same file is called Player.log and is located elsewhere.
  25. allista
    This comment was made ~3yrs ago: https://github.com/allista/ConfigurableContainers/commit/7db0aa83c278a6388220553980b205f70dd030b4 so it was probably true then And the 2x multiplier was an arbitrary compromise between the mentioned values. If everyone will agree on some value I'll be glad follow; but until then I prefer to leave things as is.
×
×
  • Create New...