Hmm...no one posted here at all for 2009, I hope this isn't a forbidden topic or something now. Anyway here's a few routines that I thought were kinda helpful



Code:



This one could probably be optimized. Also, ld a,48 could be changed to ld a,c to center text at a manual x co-ordinate


Code:

Does that rand routine produce uniformly-distributed numbers?

It probably depends on when you call it.

r is a 7 bit value, dependent on the number of opcodes that have been read in.

Indeed it is, but a cursory glance at the routine posted makes me worry that the results would be low-biased.

I've used it between user input to keep r fresh. Its only truly random for up to 256 combinations between inputs but when you don't need to use it that often, its smaller, faster and doesn't kill as many registers as bcall(_random) and whatever code would be put around it to keep it in a certain range.

I guess if you had an interrupt that used a bunch of T-states (like grayscale), then you could add something like this to the end of it to make r more random


Code:

Here's a floodfill routine for small spaces. Try a large space, and you'll overflow the hardware stack. It also doesn't like reaching the top or bottom of the screen.

Code:

[/code]

Input: hl=object name eg. .db 6,"PRGM",0
Output:a=page hl=datastart de=dataend bc=datasize
Find an object regardless of whether or not its the the flash

Code:


Input: a=getcsc keycode
Output: z=is pressed
Converts a getcsc keycode to both a keygroup and keymask, then checks the key with direct input allowing you to check multiple keys

Code:

Nice ones, Anakclusmos. There's a very similar routine to that first one in the Ion libraries; it's called iDetect and is documented here:

http://dcs.cemetech.net/index.php?title=IDetect

I know I just thought people would rather check by name than a string.

Ah, of course, that's the main difference. Very nice, I'm sure that they would.

Calc84's Optimized Version of My BC-A->BC


Code:

To clarify why I did a cpl/scf/adc instead of a cpl/inc/add or neg/add, is that it handles the case of A=0 properly. Typically, SUB N and ADD A,-N give opposite carry outputs, but SUB 0 and ADD A,-0 both reset the carry flag. On the other hand, SCF \ ADC A,255 will set the carry flag like we want it to.

Indeed, thanks for that. Needless to say, this code segment will work equally well for HL-A and DE-A as well as its original BC-A.

EDIT: Cleaning up and updating (bug fix and faster)
FindSym

This routine has four entry points with different inputs, but they all have the same documented outputs as rFindSym:

Input

Code:


Output
Same as the documented outputs of rFindSym.

Code

Code:

I created a routine which creates a menu just like it's basic counterpart. Although this is very easy, and doesn't include complex coding, I hope it'll be of any use for anyone.

Code:

That's quite good, thanks for sharing!

No problem!
I'm still planning to make a scrollable one, with multiple menu's at the top of the screen.

Cool, I look forward to seeing that one as well. You should check out the DCS GUI API, too.

Hey,

I updated my MenuProcedure a bit. It will now automatically store the result into A, instead of saving the place where it's stored in Asm_flag1. So here it is, enjoy!
Edit: I forgot I added a second call, so the codestream didn't work anymore. But now it does. Everything comes at a price though, register IX is destroyed aswell, as it wasn't in the previous version.

Code: