TimePhys - An experiment

TimePhys

Please note that while I do plan to continue this project, TimePhys is not its name. That's just what I call it.

That being said, this is an experiment. It is an experiment in custom GUIs, Functions, Dynamic Timescaling (Throw stuff into other stuff really hard, it goes slomo 90% of the time - needs a total velocity of 250 or so) and also an experiment in usage of UIDs. For instance, picking the objects for dragging is done via storing the effected object's UID in a global variable.

Feel free to rip the UI functions out of this. You use functions to create stuff, cause it to do stuff, etc. and yes the load file function under UI is for loading a UI from a file. In fact it'll load it all into a window titled the same as the file, with offsets and the like.

Screenshots

It's a bunch of gray sliders and boxes with crosses on them, not much to look at hence not much point screenshotting it.

Controls

Double-click to spawn a box. This is a placeholder object, it's actually the close button I made for the UI.

Escape to close the window.

Click and drag boxes to throw them about.

Move the slider to adjust how closely the boxes follow your cursor when dragged (it multiplies the difference in velocity by the slider's value, so 16 is good close following and 32 can get motion blur going at 0.1 timescale)

Clear to remove all physics objects, obviously.

Download

If you have any suggestions, speak up. I want to hear them.

On the list already:

• More UI objects: Checkbox, radio button sets, windows, and possibly even a custom textbox although I'm inclined to think that will be very much impossible.
• Object system with good old Phys-Box's gibbing objects - but with randomized gibs.
• Custom saving which saves all physics objects, their velocities and associated variables etc. (once I learn some file stuff)
• A console, letting you run functions with parameters (I always wanted one of these!)

The .exe gives me an error when I try to run it. I looked at the cap though and it seems interesting enough. One thing though and I used to do it too. You dont need to divide event sheets up so much Most of that stuff can probably go in the main event sheet and you can cut down on the even sheets. there is not point in having event sheets with 2 or 3 events in them just use a group. just a suggestion.

The .exe gives me an error when I try to run it. I looked at the cap though and it seems interesting enough. One thing though and I used to do it too. You dont need to divide event sheets up so much Most of that stuff can probably go in the main event sheet and you can cut down on the even sheets. there is not point in having event sheets with 2 or 3 events in them just use a group. just a suggestion.

This will end up a lot bigger. I have done the event sheets with groups only for a larger project - not huge, but not small either - and it gets annoying. I'd rather have everything clearly divided from the first than have to spend time dividing it all up later when it gets too much to work with.

Also, I have no idea what causes it, but the .exe errors for me too when I download it, but the local copy I added to the zip works perfectly.

Edit: Works even in zip locally, once uploaded it breaks.

That's a cool use of timescaling. And the best use of motion blur I've ever seen.

I like it

I'd like to point out that this includes an example of how to absolutely move an object with the proper physics reaction.

How about doing this speed-setting with a hidden proxy physics object for a platform object?

I'd like to point out that this includes an example of how to absolutely move an object with the proper physics reaction.

How about doing this speed-setting with a hidden proxy physics object for a platform object?

I was actually going to be doing something more along the lines of a physics-based platform player, if I made a player at all. If I proxied a physics object in there, then it'd drag behind if it got caught on something immovable. On the other hand I could make the player exert force on physics objects it overlaps with, but then I'd have to do some fancy shit with mass to make the player not just push everything.

Easiest way seems to be to make a physical player bounding box that handles all the physics with a player image mounted to that. Possibly bone-movement since I haven't tried that, although definitely not IK solver (on that small a scale? Pointless. The player will only be about 24px tall)

I tried this here:

I set position instead of velocity, which fixes the immovable object problem but lacks proper momentum, bouncing things reacts weird.

I set my proxy to immovable, btw, as it was supposed to follow the mouse pointer.

I tried setting velocity but things went weird, probably because I had immovable turned on.

The tricky bit of a physics platformer is that once you go all-physics, you lose all the benefits of the platformer behavior. There has been talk of making a new physics platform behavior, which would be designed to work like that.

Cool example. Thanks for sharing your .cap so all of us can learn!