From Keller’s 1st report:

5. Performance problem when all charts open with complicated track.

            I have trouble reproducing this problem (perhaps you could email me a track that exhibits this problem?)  Also not sure how much time should be spent solving the situation with 3 charts and a complicated track.

6. Moving the skater below ground (temporarily) by dragging his track.

This seems acceptable to me, since the user must be actively holding the skater below ground.

7. Skater goes underneath the track in double-well

            See Tests->Test Fallthrough for some experiments in which I tried to reproduce this bug.  I did see one fall through, but this was in many (50+) trials.  The screenshot in Keller’s document that indicates falling through actually looks like the red dot is above the track, and the falling-through is just a graphics anomaly.

11. Bounces at an angle, then stands upright.

18. Energy vs Position data covers legend and axes.

25. Energy bar chart has no units.

            I added a Joules label, but were you asking for quantitative markings?  Should we provide these?

31. Measuring Tape has a ruler icon

            The measuring tape doesn’t look much like a measuring tape when iconified.

33. Arrow keys for thrust appear everywhere, but are advertised only on when gravity=0

34. PE=0 line is visible in space

35. Is thermal energy property of skater?  What about resizing pie chart?

36. Same as (35)

37. Pie chart disappears when going underneath PE=0

41. Something preventing characters from being grabbed near the left side of the screen

            I couldn’t reproduce this problem.

43. Frictionless Landing

            This button is poorly worded, and probably shouldn’t be exposed to students.  It refers to whether the normal velocity is converted to thermal during the (instantaneous) landing process.  What should I do about this?

45. Time series misses a single point near the end sometimes

            I tried several techniques to address this problem, and eventually I couldn’t reproduce it any longer (though I wasn’t sure what combination of my fixes actually fixed it).  So there is a possibility it is resolved.



From Keller’s 2nd report:


7. Dragging the skater around should automatically determine the velocity and kinetic energy (as in Moving Man).  I didn’t implement this yet since it would be a substantial feature addition and therefore more time consuming than many of these quick bugfixes, and I’m not certain how vital it is to this simulation.


9 & 10. Chris has recommended that the bar chart contain quantitative information, but we have previously discussed keeping it qualitative.  What do you prefer here?


Issues 3, 12 and 14 all relate to the same issue.  This simulation has 3 different modes of operation “Live”, “Record” and “Playback”.  For example, some difficulties this causes are:

1.       Sometimes “Step” means step to next recorded data, but sometimes it means record new data.

2.       The Energy Vs. Time chart shows Live data, not Recorded data.  Thus the data it displays won’t always match what is seen during playback.

Covertly changing from Live mode to Record mode without the user knowing it has been problematic.  Here are some different ways we might address this problem:

1.       Redesign this part of the simulation to make it obvious to the user that there are 3 modes (for example, the user can see and switch between all modes)

2.       Delete the “Live” mode, so that data is always recorded.

3.       Refine the semantics for covertly changing from live to record

4.       Remove all covert mode changes, so that record mode is not automatically started when opening the Energy Vs. Time chart.

a.       integrate the record/playback controls with the Energy Vs. Time chart.