On-Calc handwriting

[KermMartian]

I didn't bother doing those optimizations because they only optimize for size, but slow it down. I had an extensive discussion about this very topic on my forum (0-->N vs. DelVar) and we determined via batch tests that 0--> is faster. This makes sense, since if you DelVar, you have to wait for the variable to be created next time you try to do something with it.

[kalan_vod]

Did you guys time it? But what you are saying makes sense.

[DarkerLine]

Is there a slow-down factor in these optimizations?

Code: Select all

ClrDraw
{0,0->L1
{31,47->L2
{61,94->L3
Repeat K=105
getKey→K
If K>45 and K<96
Then
int(.1K
{Ans-7,K-3-10Ans -> L1
End
max(0,min(L1+L2,L3->L2
L1not(L2=L3 or not(L2 -> L1
Pxl-On(L2(1),L2(2
End

[kalan_vod]

I will try it later, but that looks smaller and bigger at the same time.

[CoBB]

Here's the whole program, let's you sketch with the touch keyboard method:

But this is a completely different program! You don't calculate a difference from the previous point (let alone use a timer) but basically introduce a set of slower and faster cursor keys.

[KermMartian]

Is there a slow-down factor in these optimizations?

Code: Select all

ClrDraw
{0,0->L1
{31,47->L2
{61,94->L3
Repeat K=105
getKey→K
If K>45 and K<96
Then
int(.1K
{Ans-7,K-3-10Ans -> L1
End
max(0,min(L1+L2,L3->L2
L1not(L2=L3 or not(L2 -> L1
Pxl-On(L2(1),L2(2
End

Indeed, lists are always slower than vars.
And we did time that, kalan.

[DarkerLine]

Ah well, much of the execution time is actually spent waiting for a key, I think the size improvement is worth it.

[KermMartian]

Ah well, much of the execution time is actually spent waiting for a key, I think the size improvement is worth it.

Fair enough, that's always the trick, figuring the correct ratio and balance of speed and size.

[Mike K]

Okie, haven't loaded it onto my calc yet, but heres the compiled version (asm):

www.mikeknoop.com/upload/test.8XP

-Mike

[KermMartian]

nostub?

[kalan_vod]

I couldn't get it to work.