• Content Count

  • Joined

  • Last visited

  • Days Won


Everything posted by satguru

  1. @BitOfGold Is your PR in? I find that I still need to do "shadowGenerator.bias=-0.3;"
  2. InstancedMesh id is not serialized. name is.
  3. A 3rd person character controller for BabylonJS source : docs: demo : git download: npm: Originally developed as part of my Vishva project
  4. due to chrome v63 bug my BabylonsJS 3.0 app broke on chrome to fix I did engine=new Engine(this.canvas,true,{"disableWebGL2Support":true}); ...
  5. @Deltakosh Serialization/deserialization is not happening properly now. After Serialization/deserialization ellipsoid behaves as before.
  6. I think our assumption about how the ellipsoid is positioned in the mesh might be incorrect. See the discussion below, on how I think the ellipsoid is actually positioned in the mesh Also, as noted in that discussion, collision between two meshes is not between ellipsoid of one colliding with the ellipsoid of the other . It is between ellipsoid of one colliding with the surface of the other.
  7. New Release 2.5.0 (12/03/2017) added translation and scaling bounds. client can limit translation and scaling between a minimum and maximum value contributed by @focomoso
  8. satguru

    gltf/glb issue

    can somebody check why we cannot load the attach glb file. If I drag and drop this into our sandbox at I get "Cannot set property '0' of undefined" error I am able to load this in other gltf viewers or vincent2.glb
  9. @aWeirdo Nicely drawn Based on the above discussion the offset is the distance of ellipsoid top from mesh center So
  10. @Deltakosh nice. Have few more questions but will ask them in the question forum. Thx @aWeirdo Not fishy anymore When you make ellipsoid.y=0.5, you are changing the ellipsoid height to 1m (same as that of box) and when you make ellipsoidOffset.y = 0.5 you move the ellipsoid's top to the top of box thus moving the ellipsoid completely inside the box
  11. So if I move the ellisoid up by 1m I should then be able to move box1 down by 1 m. But that does not happen. It only moves down by 0.5m. See So it would seem the ellipsoid is positioned as follows The top of the ellipsoid at the center of mesh? And thanks for indulging all my questions.
  12. Sorry to be a bother but I still don't get it. I do not see why we have to move the ellipsoid up by 1m. As you can see from, my very professionally drawn, sketch below there is already a gap of 0.5 meter between the ellipsoid and box2. What am I missing?
  13. Regarding a. If the height of the ellipsoid is 2m and is centered on the box and height of the box is 1m then that means the ellipsoid should only extend 0.5m outside the box, at the top and bottom. That means the gap between the bottom of box1's ellipsoid and top surface of box2 is 0.5m. Should I not be able to move the box1 down by atleast 0.25m?
  14. Yes I was wondering about the ellipsoid property. Can you clarify few points below? a. In the above example the gap between the boxes is 1m. Does box1's ellipsoid extends outside the box1 bounding box by 1m? b. If I move box1 with moveWithCollision then I am assuming the collider system uses its ellipsoid as the impostor. But what about box2? It is NOT being moved by moveWithCollision? Does the collider system uses its ellipsoid too or does it use its bounding box as impostor?
  15. Thanks. Keep forgetting about compute world matrix Here is another issue though Now I am trying to move box1 down towards box2 by 0.25m Gap between them is 1m. But box1 does not move down at all. Infact it moves up by 0.01.
  16. FYI I was watching Siggraph 2017 sessions on YouTube and came across this nice talk on BabylonJS by Sebastien
  17. see Here I have 2 boxes box1 and box 2. box1 is way above box2 I move box2 down by 1m using moveWithCollision(). I print the before and after position on console. Everthing works as expected. But now if i turn on collision on box1 (uncomment line 21), box2 does not move !!
  18. New Release 2.4.5 (09/04/2017) see for details with this release the edit control is now also available from the NPM repository
  19. @MackeyK24 don't see whats wrong with doing "direction=velocity.normalize()"
  20. When I used the term "displacement" I did not realize it then but now when I google it up I find it has a very precise definition "Displacement is a vector. This means it has a direction as well as a magnitude and is represented visually as an arrow that points from the initial position to the final position. " from I would think perfect for moveWithCollision I suspect those physics lessons from school are still floating around somewhere in my subconscious mind.
  21. @Wingnut I remember coming across that link about a year back and making a mental note to revisit that later on but forgot all about it. Thanks for providing the link. I definitely want to understand that better and @RaananW writes well. I also remember reading somewhere on the forum that Dad72 originally came up with the idea of moveWithCollision. Nice of him. Anyway the impression I get is that WebWorker just deals with collision and does nothing as far as velocity is concerned. @Deltakosh am I wrong in treating velocity as just displacement?
  22. @Pryme8, Yes it would seem it is units/frame but you have to keep calling that function every frame. It's not as if you call it once and it keeps moving along that vector henceforth. If you remove collision from the picture then it same as just setting the position of the mesh every frame. Calling it velocity is stretching it a bit. @Wingnut, the code you point out , seems to be just setting the position. Do not see time anywhere. Funny!
  23. how is this velocity? velocity is speed and direction. What is the speed measured in ? units/s ? if I do moveWithColiisions(new BABYLON.Vector3(200,0,0)); The mesh moves 200 units immediately and then stops. So don't think the speed is 200 units/s.