Pokemon Purple

Delnar_Ersike · Post by **Delnar_Ersike** » Mon 29 Jan, 2007 3:51 pm

I remember running your Pokedex program on my 83+ Silver, yet it only had a minor speed increase in drawing.

trevmeister66 · Post by **trevmeister66** » Mon 29 Jan, 2007 6:24 pm

Ah, but think how it will run on my 84+ SE

Delnar_Ersike · Post by **Delnar_Ersike** » Mon 29 Jan, 2007 10:47 pm

trevmeister66 wrote:Ah, but think how it will run on my 84+ SE

FYI, the 83+ Silver and the 84+ Silver both have 15 Mhz processor speeds, so there is no difference.

tifreak8x · Post by **tifreak8x** » Mon 29 Jan, 2007 11:13 pm

Even though the 83+SE, 84+, and 84+SE have the same speed processor, I find my 83+SE seems to be faster than my 84+. I do not yet have an 84+SE to test with, so I am unsure about it. I think keeping track of the clock slows the processor down slightly...

I tested this by drawing Bulbasaur on both calculators at the same time, the 83+SE beat the 84+ by 5 or 6 seconds.

necro · Post by **necro** » Tue 30 Jan, 2007 3:38 am

its to bad you aren't making any of this asm, something like ion's large sprite routine would work well for such a use (assuming you copy and paste the code from it to the asm prog).

thegamefreak0134 · Post by **thegamefreak0134** » Tue 30 Jan, 2007 7:18 am

True, but he does want to make this a BASIC only version. I think he can pull it off too. His will be BASIC, mine will be Asm. We are striving (and helping each other out in the process) to make each the best it can possibly be for the language in question. Thus, I will certainly (and obviously) be using nothing but asm, and he will be using nothing but BASIC, up to and including Basic Helpers written in asm like zsprite and Omnicalc.

I do still want to see that routine by the way. Oh, and I do want to confirm that the 84+ is indeed slower in BASIC operation than the 83+, I did a bunch of counting tests on it. Not quite sure what the speed difference is, but I do know that the new one is slower.

-gamefreak

Post by **DarkAuron** » Tue 30 Jan, 2007 12:47 pm

tifreak8x wrote:Even though the 83+SE, 84+, and 84+SE have the same speed processor, I find my 83+SE seems to be faster than my 84+. I do not yet have an 84+SE to test with, so I am unsure about it. I think keeping track of the clock slows the processor down slightly...

I tested this by drawing Bulbasaur on both calculators at the same time, the 83+SE beat the 84+ by 5 or 6 seconds.

If you want to find the true difference, figure out the time difference it takes to update the screen. I know for a fact that the ti-84+SE OS updates the screen slower, which becomes rather annoying when scrolling through 8kb of BASIC code. Anytime you draw a line, a pixel, or characters on the homescreen in BASIC the screen gets updated, and that's overall is going to be much slower on the 84+ series. Try it.

tifreak8x · Post by **tifreak8x** » Tue 30 Jan, 2007 4:16 pm

thegamefreak0134 wrote:True, but he does want to make this a BASIC only version. I think he can pull it off too. His will be BASIC, mine will be Asm. We are striving (and helping each other out in the process) to make each the best it can possibly be for the language in question. Thus, I will certainly (and obviously) be using nothing but asm, and he will be using nothing but BASIC, up to and including Basic Helpers written in asm like zsprite and Omnicalc.

I do still want to see that routine by the way. Oh, and I do want to confirm that the 84+ is indeed slower in BASIC operation than the 83+, I did a bunch of counting tests on it. Not quite sure what the speed difference is, but I do know that the new one is slower.

-gamefreak

Yes, thank you for reminding me. And the only asm program I am now using is Resource, for the archiving capabilities.

I did some checking, the version that you saw was the older version. I am going to resend the programs with the new stuff, so you can see how much faster the new system is.

But here is the code for the newer system until I can get the new screenshot:

:I->C:J->U
length(Str4->N
For(/theta/,1,N
inString("123456789ABCDEF",sub(Str4,/theta/,1
If 7<Ans
Pxl-On(U,C
If 3<8fpart(Ans/8
Pxl-On(U,C+1
If 1<4fpart(Ans/4
Pxl-On(U,C+2
If fpart(Ans/2
Pxl-On(U,C+3
C+4->C
U+(C=(I+32))->U
If C=(I+32):I->C
If sub(Str4,/theta/,1)="G":Then
U+1->U
I->C
End:End

thegamefreak0134 · Post by **thegamefreak0134** » Tue 30 Jan, 2007 4:56 pm

:I->C:J->U
length(Str4->N
For(/theta/,1,N
inString("123456789ABCDEF",sub(Str4,/theta/,1
If 7<Ans
Pxl-On(U,C
If 3<8fpart(Ans/8
Pxl-On(U,C+1
If 1<4fpart(Ans/4
Pxl-On(U,C+2
If fpart(Ans/2
Pxl-On(U,C+3
C+4->C
U+(C=(I+32->U
If C=(I+32):I->C
If sub(Str4,/theta/,1)="G
Then
U+1->U
I->C
End:End

This is some severely optimized stuff already, I'm quite impressesd. I just noticed a couple of little nit-picky things here and there, bolded. Of course, you used some weird nesting on the first line and I don't have a real cal to test in front of me, but the second one (moving the Then after the "G) saves you a byte and a fraction of a second I believe.)

-thegamefreak

*EDIT* Just noticed, the line in-between them can become

If C=(I+32:I->C

to save another byte.

Saibot84 · Post by **Saibot84** » Tue 30 Jan, 2007 5:34 pm

Although this would probably exponentially increase the amount of work you'd have to do, what about changing that to use some sort of scripting language such that each "sprite" is saved as a series of inputs for either Line(, Pxl-On(, or Pxl-Off... In my tests, I once tried to create a Zelda game in Basic (using real sprites instead of ASCII) and the sprites could be displayed at a decent speed because I used a GIANT subroutine program that read something to the effect of: If A=1:Then:Line(...):Line(...):Line(...):...:Pxl-Off(...):End:If A=2:Then:Line(...):Line(...):Line(...):...:Pxl-Off(...):End... But since this took an IMMENSE amount of space, maybe an interpreter that parses a list as a series of Line(, Pxl-On(, or Pxl-Off commands might be faster.

____________________
Reaktix Software (formally XDG Kat-Productions)
-Michael

MyBlog ChatWithMe

kalan_vod · Post by **kalan_vod** » Tue 30 Jan, 2007 6:33 pm

Saibot84 wrote: Although this would probably exponentially increase the amount of work you'd have to do, what about changing that to use some sort of scripting language such that each "sprite" is saved as a series of inputs for either Line(, Pxl-On(, or Pxl-Off... In my tests, I once tried to create a Zelda game in Basic (using real sprites instead of ASCII) and the sprites could be displayed at a decent speed because I used a GIANT subroutine program that read something to the effect of: If A=1:Then:Line(...):Line(...):Line(...):...:Pxl-Off(...):End:If A=2:Then:Line(...):Line(...):Line(...):...:Pxl-Off(...):End... But since this took an IMMENSE amount of space, maybe an interpreter that parses a list as a series of Line(, Pxl-On(, or Pxl-Off commands might be faster.
____________________
Reaktix Software (formally XDG Kat-Productions)
-Michael
MyBlog ChatWithMe

Faster yes, much more space needed though.

tifreak8x · Post by **tifreak8x** » Tue 30 Jan, 2007 9:05 pm

thegamefreak0134 wrote:
:I->C:J->U
length(Str4->N
For(/theta/,1,N
inString("123456789ABCDEF",sub(Str4,/theta/,1
If 7<Ans
Pxl-On(U,C
If 3<8fpart(Ans/8
Pxl-On(U,C+1
If 1<4fpart(Ans/4
Pxl-On(U,C+2
If fpart(Ans/2
Pxl-On(U,C+3
C+4->C
U+(C=(I+32->U
If C=(I+32):I->C
If sub(Str4,/theta/,1)="G
Then
U+1->U
I->C
End:End
This is some severely optimized stuff already, I'm quite impressesd. I just noticed a couple of little nit-picky things here and there, bolded. Of course, you used some weird nesting on the first line and I don't have a real cal to test in front of me, but the second one (moving the Then after the "G) saves you a byte and a fraction of a second I believe.)

-thegamefreak

*EDIT* Just noticed, the line in-between them can become
If C=(I+32:I->C
to save another byte.

While normally that is the case, I was getting some really wierd errors upon execution, when I put the ) back in there, it worked just fine, so for 3 bytes, I am not going to complain.

Saibot84:

As Kalan said, that would take massive amounts of memory, and as it stands the sprite data already takes 25,960 bytes, though that was optimized from the 40,432 bytes they used to take up. o.O

necro · Post by **necro** » Wed 31 Jan, 2007 3:50 am

cool, it is to bad this doesn't use asm though (even if just for things like that)

tifreak8x · Post by **tifreak8x** » Wed 31 Jan, 2007 4:40 am

Well, see, We would expect ASM to be able to handle this, since it was built on a variation of Z80 asm. But it makes it more impressive (imo) to be made in BASIC, and be fully playable.

kalan_vod · Post by **kalan_vod** » Wed 31 Jan, 2007 5:13 am

tifreak8x wrote:Well, see, We would expect ASM to be able to handle this, since it was built on a variation of Z80 asm. But it makes it more impressive (imo) to be made in BASIC, and be fully playable.

More so, anyone who completes a massive project as such. Does not really matter what language, as this has been attempted countless times. I would be amazed just to see a complete port.