Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Inconsistent behavior with stick to track disabled. #194

Closed
bryo5363 opened this issue Aug 19, 2014 · 19 comments
Closed

Inconsistent behavior with stick to track disabled. #194

bryo5363 opened this issue Aug 19, 2014 · 19 comments
Assignees

Comments

@bryo5363
Copy link

While testing issues I noticed that there is inconsistent behavior when the stick to track feature is disabled. The most clear demonstration of the inconsistent behavior is using a loop in the playground mode.

Steps to reproduce
-open sim
-click on playground mode
-create a loop
-disable stick to track
-drop the skater from the top at slightly different heights

screen shot 2014-08-19 at 11 49 47 am

@bryo5363
Copy link
Author

I believe this issue is somehow related to #172. Some of the behavior involves the skater falling through the track or spawning on the other side of a track.

@samreid
Copy link
Member

samreid commented Aug 20, 2014

I created the track shown above and was able to demonstrate different behavior when dropping the skater from different heights. Sometimes he will pop out of the loop, and sometimes not. I commented in the commit above about the code that causes the skater to sometimes pass through the track when at shallow angles: the problem is that it is searching for a spline point nearby to the skater's initial location instead of halfway between the initial location and final location, but switching to search based on the centroid causes worse (buggier) problems.

@samreid
Copy link
Member

samreid commented Aug 20, 2014

Reassigning to @ariel-phet to discuss the severity and priority of this issue.

@ariel-phet
Copy link

I think since this behavior depends on a sensitive dependence to initial conditions it is OK for the moment.

@ariel-phet ariel-phet removed their assignment Aug 27, 2014
@samreid
Copy link
Member

samreid commented Aug 29, 2014

@bryo5363 said:

@samreid @jbphet and I were talking about #194 and I believe this issue should be dealt with. Every time I create a loop with stick to track disabled, the skater will shoot through the track. If one had to create a very difficult path in order to reproduce this issue it would not be much of a problem but because the icon for the playground itself is a loop I believe people will be trying to recreate the track and experience this issue. Let me know what you think.

@samreid
Copy link
Member

samreid commented Aug 29, 2014

This behavior looks like it is significantly improved after addressing #207. Should be verified by QA.

@bryo5363
Copy link
Author

bryo5363 commented Sep 9, 2014

@samreid I am currently testing on an iPad and the issue is still present when creating a loop.

@samreid
Copy link
Member

samreid commented Sep 9, 2014

What is the issue exactly? Can you produce the same problem on Win8/Chrome? Can you make a screencast about it?

@bryo5363
Copy link
Author

bryo5363 commented Sep 9, 2014

When creating a loop with stick to track disabled, the skater will occasionally skate through the loop. Pressing return skater will replicate the exact motion but dropping the skater from a different position will create a new point where the skater will fall through.

@bryo5363
Copy link
Author

bryo5363 commented Sep 9, 2014

I reproduced it on OSX 10.8.5 Chrome

@samreid
Copy link
Member

samreid commented Sep 10, 2014

In my experience testing the last few versions, I have seen the skater "skate through the loop" but on further inspection, the skater is going off of a tiny jump and moving beyond the track. It would be good to see the situation you are describing to see if that is the case.

@bryo5363
Copy link
Author

@samreid I noticed this issue on 10.9.4 Safari 7.0.5. I created the same type of track that I have posted on this issue. I do not think that the skater is going off of a tiny jump; I dropped the skater from a higher point and did not observe the skater falling through but at a critical point the skater exhibits this behavior.

@bryo5363
Copy link
Author

Also the skater did not fall through the loop until she completed a full completion of the track; it occurred the second time she went around.

@bryo5363
Copy link
Author

bryo5363 commented Oct 9, 2014

I did not notice the issue on Safari 9537.85.10.17.1 but I did notice that the skater would still fall through the loop at various heights on Chrome 37.0.2062.120 on 10.9.5. I will test this issue on other devices before closing the issue.

@bryo5363
Copy link
Author

bryo5363 commented Oct 9, 2014

Just noticed the issue is still present in the latest Safari. I created a track like this

screen shot 2014-10-09 at 11 27 13 am

The skater would usually go through the loop but occasionally would hit the end of the loop and bounce back up.

@bryo5363
Copy link
Author

Here is another example of a track that will produce inconsistent behavior. When dropping the skater from a certain height, she will fly through the track when she hits the bottom.

screen shot 2014-10-16 at 11 39 30 am

This is the same track from my post in #297

@bryo5363
Copy link
Author

I also noticed some strange way to produce this behavior. Drop the skater from the top of the track (try to recreate the track I posted above). Switch tabs immediately, wait about 3 minutes then switch back. The skater will always fly off the track when hitting the point of least PE

This behavior was noticed on OSX 10.9.5 Chrome 37.0.2062.120

@samreid
Copy link
Member

samreid commented Oct 24, 2014

I also commented on this issue in #297

@ariel-phet
Copy link

I would be concerned if this were happening with "stick to track" enabled. But since stick to track is disabled, I think some funkiness is allowed. In the real world to skate around a look generally requires a pretty smooth loop. I think we can leave as is. As long as energy is conserved in this situation we are OK. As much as I would like the behavior here to be perfect, we must remember that the goal of this sim is to focus on energy conservation, so a little bit of odd behavior is acceptable as long as Energy is still behaving properly.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants