Trajectory program event handling

Question

Nobody else responded to my question on the subject "Trajectory" in Java Programming forum. See my post history for problem description and code. Perhaps the subject is more appropriate here as an event handling issue. Who has the technical savvy to attempt a solution?

Answer 1

(2) I wouldn't have problem with separate classes as
long as I have no inner classes and the program
essentially works. Sure, one 10,000 - line class may also work fine. However, it is reccommended do not make classes
(as well as source files) very large. The reason is that long file is more difficult to debug.
Second reason why i advised you to do that is that you mix in one class Trajectory (which is about 2000 lines long!!!) all - interface, event handling, calculations, parts responsible for 2D panel processing....
Well, this is not Java is supoposed to be used for.
(3) Creating object of class Falling object is
something I'll consider. If you could show me a brief
code example, I'll understand more.
class fallingObject
      private double mass;
      private double speed;
      Position[] trajectory = new Position[1000];  // Create class Position each object of which holds only
      // 2 variables - x and y
     //all initial data related to this object
     fallingObject(double mass, double speed , /* the rest ........*/)
          this.mass = mass;
           this.speed = speed;
          moveObject();
          //and so on
    // some methods you need
     private moveObject ()
         // perform all calculations
         // fill array
     // etc
Now, because all except trajectory array and constructor is private you cannot change this object outside
class. This means that if something goes wrong with calculations, you check only this class and nothing else. In your case you can change some variable at lines 1352 and 122 - you have to check all program to find bug. Here you have to check only this class.
You may also supply this class with mutators and accessors
public double getMass()
   return mass;
public void setMass(double mass) // only if you really need this
   this.mass = mass;
The time interval
is a tricky parameter since not all calculations work
properly with input data; check input
the logic would require more
complexity and possibly cause run time problems;
automatically testing for the appropriate time
interval could also reduce program efficiency. Dont think so. You can calculate time of falling and divide it by 1000, for instance.
It will take very short time.
I don't
want to set a limit for array points to 1000 or less
since the precision becomes limited and the power of
the program can be hurt severely since the time
duration would be restricted to less than 16 2/3
minutes( 1 second per array index increment). Dont use real time simulation. There is no sence in it. You just want to show how object is falling.
Also, I
don't think the array size can't be changed during a
program run although I have set a much higher maximum
value for array size. Theres no need to change array size.
(4) Erasing data from textfields is something I desire
as a calculation reset feature since it should signal
user that it expects new data and not looking at old
data; however, I might consider additional variables
to hold the data. Well, man, it was quite irritating to type them in. So many fields. User is not stupid,
he knows that he can input other data. Almost anybody will leave this page and wont want to
fill fields again. You have to think about user, not about what you want.
(5 )BTW, what I have in the array doesn't have the
same precision as with the calculation mode. The 2D
simulation uses less precise data to become visually
practical for comprehension of users. But perhaps I
need more than one type of array, but then again,
performance could be a problem.
(6) I thought the program already does take parameters
from text fields and start again. This is what I
desire anyway so I'm not sure why I need new start
feature.
(7) Restart animation seems unaccepatble since I
prefer to treat each simulation as something based on
new calculation data.
I would have initially set up the program with OOD
emphasis, but the original intent was a java
conversion of a a C++ program that needed only a
structured program design and the conversion worked
fine. This Java program has evolved with the 2D
simulation enhancements.
When you did testing the other day, did you use
appletviewer? I looked at it on your site using Safari and run it as application also.
Appletviewer will not detect the problem
so don't rely on Appletviewer to interpret the
problem. Oddly, I wish the program behaved the same
way as what happens when using appletviewer. In fact,
if I could change 2d reset button logic to terminate
and restart the entire applet again via html, then my
problem would be solved. But I don't know if this type
of applet restart is possible or how to effect the
code change if feasible.