This is WIERD

[J-G]

EDIT  -  What happened to @WP's post ??

I had deleted it after seeing emersonline's message which told me that my idea of the "optimized-away" expression with unused result was nonsense.

Shame  -  without seeing that, I wouldn't have realised that the semi-colon was the 'Key'.

[wp]

No, it's not the key. You just have to press F8 one or two more times so that the red line is at the end of the procedure - then D has its correct value (*). Or - without the debugger - try this: use only the compact line, and put the result into variable A rather than D:

Code: Pascal  [Select][+][-]

1.   Procedure Find_A;
2.   begin
3.     if sin_b = 0 then
4.       begin
5.       {
6.         A := b*b + c*c;
7.         A := A - (2*b*c * cos_a);
8.         A := Sqrt(A);
9.         }
10.         A := Sqrt((b*b + c*c) - ((2*b*c) * cos_a))
11.       end
12.     else
13.       A := (B * sin_a) / sin_b;
14.   end;

(*) Of course, you can ask why this is necessary ... But then ask Martin_fr, the debugger magician.

[J-G]

Missed an opportunity to optimize  Tan_X  as  Sin_X / Cos_X 

I've been considering this option and have come to the conclusion that using the existing [Tan] Function is the better method.

If the angle is 90° (which, when talking of Triangles, is often likely!)  then Cos(90) is zero so dividing by zero will create an error.   I have to assume that the in-built Tan function takes care of such so using that I don't need to take the precaution.

I'd appreciate a counter argument

[emersonline]

No, it's not the key. You just have to press F8 one or two more times so that the red line is at the end of the procedure - then D has its correct value (*). Or - without the debugger - try this: use only the compact line, and put the result into variable A rather than D:

Code: Pascal  [Select][+][-]

1.   Procedure Find_A;
2.   begin
3.     if sin_b = 0 then
4.       begin
5.       {
6.         A := b*b + c*c;
7.         A := A - (2*b*c * cos_a);
8.         A := Sqrt(A);
9.         }
10.         A := Sqrt((b*b + c*c) - ((2*b*c) * cos_a))
11.       end
12.     else
13.       A := (B * sin_a) / sin_b;
14.   end;

(*) Of course, you can ask why this is necessary ... But then ask Martin_fr, the debugger magician.

I had already come across this peculiar behavior with the Lazarus debugger before (where the cursor passes over a statement without executing it, reaches the end of the block, and then re-enters the block to finally execute the instruction), but I didn't remember it during my first test. Anyway, your observation about including the missing semicolon was not only effective but also reinforces a good practice. Even though the compiler allows it, it can lead to strange results/behaviors, as we just saw in this case.