a simple boxe timer (code included)

[T.Jackson]

p.s. is this code more faster? (no IF needed)

LOOKDOWN2 W0,>[255,242,218,192,166,144,124,105,80,59,49,40,0],B0
LOOKUP2 B0,[60,600,180,120,60,60,40,30,20,15,10,45,30],timerAc
LOOKUP2 B0,[60,180,60,60,60,30,40,30,20,15,10,15,15],timerBc
LOOKUP2 B0,[6,1,2,3,4,5,6,7,8,9,10,11,12],timerN

At a guess I'm going to say much slower. LOOKUP / DOWN is a function, a group of operations, probably containing many IFs inside it. I think the protocol behind that function would be lengthy, well protected and complex.

I place no bets though, never implemented it.

Trent Jackson

[T.Jackson]

The main reason "Select Case" is faster is because it allows you to structure your code. As with an IF statement there's still a "compare" operation being preformed, but it's usually far better optimized.

The biggest problem with your original code is that every IF will be executed, which consequently wastes time.

Say for example that b1 = 245, the first IF is tested, the result returns true because b1 is within range of (255 - 243), so the block of code associated with that IF is executed, timerN is set to 1 etc ...

Now here's the problem -- all subsequent IFs are executed even though b1 is clearly out of range and the result will always return false while ever b = 245.

    If (b1 <= 255 And b1 >= 243) Then ' Comparison always performed
        timerN = 1
        timerAc = 600
        timerBc = 180
    End If
    If (b1 <= 242 And b1 >= 219) Then ' Comparison always performed
        timerN = 2
        timerAc = 180
        timerBc = 60
    End If

Select Case can be configured in a way so that only one comparison is ever performed, only one block of code is ever executed.

Trent Jackson

[precision]

Great,... Yes.

IF AND THEN --- compile to 159 words.
Select case --- compile to 58 words.

I will have to compere all statements. eg. FOR/NEXT loops. while/Wend, GOTO, GOSUB/RETURN, HI/LOW, TOGGLE.
For time of Execution speed and code space.


.

[T.Jackson]

I will have to compere all statements. eg. FOR/NEXT loops. while/Wend, GOTO, GOSUB/RETURN, HI/LOW, TOGGLE.
For time of Execution speed and code space.

while / wend (same as "do / loop" but without a condition) -- will always be faster than "for next loops" Main reason is because there is no counter, there's no constantly incrementing or decrementing after each iteration.

But I encourage you to conduct your tests and post the results here.

Trent Jackson

[T.Jackson]

Another thing also, the "AND" operator -- try to avoid it at all costs.

if b1 = 100 then
   if b2 = 200 then
      Thank Mr Jackson for these invaluable tips
   end if
end if

is much better than ...

if b1 = 100 and b2 = 200 then
  do something ...
end if

You might save a few bytes of code space, it'll run faster too.

Can anyone tell me why?

Trent Jackson

[skimask]

IF you're actually asking a question vs. just talking out your.............................
then...

if b1 = 100 then
   if b2 = 200 then
      Thank Mr Jackson for these invaluable tips
   end if
end if

2 comparisons, 3 jumps

if b1 = 100 and b2 = 200 then
  do something ...
end if

3 'nested' comparisons (2 + 1), couple of extra 'state' bits to save in the middle of it all, plus 4 jumps

IF you're posing a question for the class from the podium, then nevermind...

[hozone]

here the 0.3 version.

contains some bug fixes, the most important:
in 0.2 timer values are contained in BYTE VAR, big values could not be showed (in 0.3 WORD type is used).

regards,
hOZONE