mirrorbird (23:47:48)
JR 1 then

mirrorbird (23:47:43)
hmm actually maybe JR 0 would be an infinite loop. that's like 'goto self'. again: don't know, never writ z80 except tiny copy routines n stuff

Agemixer (23:46:32)
Hmm, yeah the list of jump pointers is another approach for it

mirrorbird (23:46:16)
or better JR 0 (relative)

mirrorbird (23:46:05)
seems one possibility is an optional jump, you could 'poke' (arf) the same address if you just want to continue on.

mirrorbird (23:45:40)
he mentioned the term "branch and link", this seems to be -- something

Agemixer (23:45:01)
to make a jump instead of subroutine call and then return of it, takes much less cycles.. only if it "returns" to the same prog counter most of time

mirrorbird (23:42:16)
but i think he's actually self-modifying jump-addresses, to avoid doing RET (to save a pop??). that sounds insane. but that's how i read it

mirrorbird (23:41:55)
Galaxian dude has 'blogged' at me a bit about his z80 micro-optimisation, which of course i barely understand bc i have written very little asm in my life

Agemixer (23:41:52)
if i don't do the recordings in a well planned sequence and the little post processing, it could mean 30min..1hour per (3 min) tune

Goodbye Moon

Erk (Eric Kernin)

mirrorbird

Time Left: 3:54