Ticker 2.0Beta release.

Information about latest Revision level of the software
and current Download Links.
BillM
Old Timer
Posts: 191
Joined: Wed Jan 06, 2016 10:12 am
Location: Mystic CT

Re: Ticker 2.0Beta release.

Post by BillM »

Art

I've been following the conversation between you and Brian & decided to do some comparisons of files. I've been saving older images of both the Gearotics msi files and the coinstall.exe files by appending the version number to each file name. I downloaded V2.00B on 3/9/23 and today I downloaded the file via:

https://gearotic.com/GearoticMotion.msi

I then downloaded the version that worked for Brian:
https://gearotic.com/GearoticMotion1.msi

using comp /N file1 file2 I found that the version I downloaded on 3/9/23 is identical to that downloaded via GearoticMotion1.msi.

What surprised me is that the https://gearotic.com/GearoticMotion.msi download produced a file ~10Kbytes larger than the one I had downloaded on 3/9/23.

When I install a new version I run both the GearoticMotion.msi file as well as the coninstall.exe file although this is really not necessary. My version is installed on a different drive and folder specifically F:/Design_Graphics/Gearotics. I always make sure that the coinstall.exe loads files to the same directory.

I use firefox to download the files and I'm running Windows 10. In firefox I have it setup to clear cache, history, & website date when I close firefox. I also only sync Bookmarks across platforms.

The larger filesize mentioned above is really a puzzler.

Bill Michael
User avatar
ArtF
Global Moderator
Global Moderator
Posts: 4586
Joined: Sun Sep 05, 2010 6:14 am
Contact:

Re: Ticker 2.0Beta release.

Post by ArtF »

Bill:

It could be they are slightly different files, I probably quickly uploaded a slightly newer version as GearotionMotion1. I usually only announce when
something changes. I am still working on force issues so I haven't announced that 2.01 is actually online. It adds a button that creates a
Anchor escapement with a couple of clicks. I am trying to get it all to fire properly before I did a video or announced it is all.

I have discovered a couple things.. (experiment with the Anchor if you like..), HULL creation is now 4 times faster.

1) A structure like the Pallet arm on the Anchor benefits from having more hulls, I override such a structure to tell it a maximum of 512 hulls
during hull creaton. It usually makes about 250 - 280. More hulls doesn't slow one down much unless you are in constant collision. This added
hull number makes the sliding pallet lift faces slide easier and are harder to get to hang up on edges.

2) You need to add a center of mass about half way down the pendulum arm, so that gravity properly affects it. This is part of a larger discussion
to be had about Center of Mass of objects. The anchor will tick properly and speed up or slow down as you change gravity. A 300mm
pendulum does give about 1 second period at 9800mm's/sec gravity, so the math adds up.

3) Do not orient lock any item unless it shares a shaft with something else. (preferable with a bit of mass. Orient lock the added gear
or item if you wish. This added mass allows the re-orientation forces to balance much easier so performs much better.

Overall, the physics is working quite well and more robust than many sims I've seen..but.. and there's always a but,
escapements are really hard for sims to do, the sliding and force transfers are small on clock escapements, so getting it
right is hard, and I'm not quite there as yet.

Honestly , its so close I can taste it.. but I can describe what the issue really is..

In Bullet, the center of mass (COM) of an object is at its origin. When you set a shaft hole in an object ( the main shaft I mean ),
the objects origin is placed there as all shafts originate at mesh origin. This creates a problem as gravity is applied at origin
as its assumed to be the COM. Placing a COM object onto the pendulum fixes the problem in most ways,but one..
The pendulums lock to its shaft is "slippy", as if a grub screw is loose. This is caused because the solvers see the force needed to
rotate the pendulum to match the shaft as incorrect as the COM isn't properly taken into account. This causes tooth slippage at times
during ticking. I'm pretty happy with most things, and how its all working, ( I do have a list of corrections needed on various fronts),
but overall its all very good and getting better..except escapements begin still too picky...

I added an Anchor button because ostensibly the Anchor is the easiest and most reliable escapement as long as the math is correct
on its creation, so for Kinetic artists its a great energy system. ( So the button will try to ensure all is good for a test clock..)

Im working on a solution to what you'll see is a slipping shaft error caused by the mass COM offset.


Art
Post Reply

Who is online

Users browsing this forum: No registered users and 5 guests