I'm looking to replace the following snippet with something that doesn't use decimal mode. It should increment TRIES if its lower nibble is under 9, or add 7 if that nibble equals 9. All other cases are "don't care".


The replacement doesn't have to be in situ (that would surely be impossible). Speed is not important. Of course I know how to write something straightforward that works. But I'm hoping there's an amazingly clever bit trick out there that is just begging to be used.

Context: I'm trying to achieve a slightly more faithful execution of this number guessing game on a certain system without decimal mode. https://www.applefritter.com/files/Appl ... ermind.pdf

This is my best shot. First put an escape in place with this JSR:


Then do the work in these 13 extra bytes:


Testing works. The game's try counter wraps from $09 to $10:

I came up with this 15 byte routine using an ADC designed to roll over if 9+ is input, with a BBC basic wrapper for testing. If the target byte were in zero page it would shave off a couple of bytes, but it's still too big.



The output is as expected for 0 to 9 (the ~ forces BBC basic to output numbers in hex)

Sorry, I missed this part:

I am not aware of much beyond masking the nibbles and testing for overflow method for generic BCD math.

If the goal is to count to 10, then this will work.



Actually, this will count to 19 correctly.

Hope that helps!

Daryl

_________________
Please visit my website -> https://sbc.rictor.org/

Kind of slow, but smaller than the other patches and works through [edit: up to] 99 (allegedly):

You can turn the middle three instructions into a subroutine if you don't want to affect your original labels.

_________________
Got a kilobyte lying fallow in your 65xx's memory map? Sprinkle some VTL02C on it and see how it grows on you!

Mike B. (about me) (learning how to github)

That is 10 bytes. The original has 7, so you need 3 bytes to avoid SED. That passes as amazingly clever with me.

[Edited out some of my earlier confusion]

How far up does it need to count? Is 0-19 ok, or does it need to go further?

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

Certainly past 19. A bad guesser needs many tries. But it doesn't really matter what happens after 99...

Nice work Mike! Since speed was not an issue, this is a tidy solution.

Daryl

_________________
Please visit my website -> https://sbc.rictor.org/

This seems like a straightforward way to do it, testing the lower nybble directly and conditionally forcing it to carry through: This is 16 bytes, 22 cycles (without the RTS since you can inline this), and avoids the looping. It could be 15 bytes, 19 cycles if specialised for a CMOS 6502, which I'm pretty sure got used in the Tamagotchi toys:

I replaced my solution with yours. By keeping it inline, the original program grows with just 2 bytes to a grand total of 172. Bravo!

https://github.com/kervinck/gigatron-ro ... ae0d1307f6