Denial

The inevitable happened: it got a MINIGRAFIK viewer and the C64 conversion of the splash screen went in.

This is my best worst hangover ever. Woke up with Arkanoid music and sound effect routines for the VIC in my email folder… More on that later with all the credits.

Just downloaded. I hope I can use paddles.

Update

Only tried joystick, but there is no sound. Is it me, or by design? seems to be on the 'to do list':

pixel wrote: Missing:
* Enlarging bonus
* Breaking bonus (opening a door out to the next level)
* 29 levels and the DOH
* Floating enemies
* Scores
* Music & sound

Update 2

Paddles support, graphics (the bat) becomes very flickery.

mrr19121970 wrote: Only tried joystick, but there is no sound. Is it me, or by design? seems to be on the 'to do list':

Nope. No sound. Just got the routines and need to port Arukanoido to ca65 – using my own assembler doesn't make much sense where team work makes the dream work.

There's a new version: https://github.com/SvenMichaelKlose/aru ... anoido.prg

Good news:
* Music and sound by beamrider (not fully utilised yet).
* Dynamic frame rate.
* Runs on NTSC.at the same speed as PAL does.

Bad news:
* Screen is too high for NTSC.
* Ball/vaus collision detection disappoints at times.

Sweet!

Very nice! Tried it for a little while, here's my bug report:

- joystick detection does not work unless run/stop-restore is done before "RUN" when loading from disk. This is probably due to comparing to numerical values instead of set bits? This is a common problem - disk access sets (or erases?) some bits in the VIA-registers that are also used for the joystick and if you compare to numerical values the joystick will never work. Mike could explain in more detail

- the color of the "pipes" at the border of the screen seems to switch from cyan to pink depending on when you start the game from the title screen. Maybe this has do to with $900e (volume register and auxiliary color)

- when getting the "Laser"-extra it does not fire for a few times until it finally works. Maybe a counter needs to be reset? (could not test this extensively).

tokra wrote:- joystick detection does not work unless run/stop-restore is done before "RUN" when loading from disk. This is probably due to comparing to numerical values instead of set bits? This is a common problem - disk access sets (or erases?) some bits in the VIA-registers that are also used for the joystick and if you compare to numerical values the joystick will never work. Mike could explain in more detail

That just worked on the real machine here. Did you use VICE?

Yes, but just checked on the real machine. It works even worse there. Most of the times it does not work at all, even if I stop/restore or sys65017 before starting. Some other times it works. I tried other games and all these work fine. Definitely something fishy going on here...

Edit: Just tried with the old 2015-version. No issues there.

tokra wrote:Yes, but just checked on the real machine. It works even worse there. Most of the times it does not work at all, even if I stop/restore or sys65017 before starting. Some other times it works. I tried other games and all these work fine. Definitely something fishy going on here...

Edit: Just tried with the old 2015-version. No issues there.

Will have to clean everything up a bit. It's been a long break. The paddle is used as soon as there's a change in register $9008 btw. See https://github.com/SvenMichaelKlose/aru ... rs.asm#L22

pixel wrote:Will have to clean everything up a bit. It's been a long break. The paddle is used as soon as there's a change in register $9008 btw. See https://github.com/SvenMichaelKlose/aru ... rs.asm#L22

I made a quick peek at the file - at least two issues:

* there's no initialisation for 'old_paddle_value' and 'is_firing':
- this locks the joystick if whatever is in RAM @old_paddle_value is different from what the paddle register returns,
- 'is_firing' might need to be counted down from a much higher value than 9, which corresponds to tokra's observations.

* The comparison 'old_paddle_value' with $9008 should allow for a small tolerance band, as on real h/w the paddle values jitter.

...

Otherwise the joystick registers are read out correctly with masks, so no complaint here. However, it is not necessary to set port A of VIA #1 repeatedly to input. It is already on input as per OS default, and if you want to go by the book, it's sufficient to write to the DDR once at the start of the program.

Fixed a lot of things. Joystick works and it got a little bit faster (for NTSC at least). Sounds got prioritised.

Do real paddles work now?

Mike wrote: * there's no initialisation for 'old_paddle_value' and 'is_firing':
- this locks the joystick if whatever is in RAM @old_paddle_value is different from what the paddle register returns,
- 'is_firing' might need to be counted down from a much higher value than 9, which corresponds to tokra's observations.

* The comparison 'old_paddle_value' with $9008 should allow for a small tolerance band, as on real h/w the paddle values jitter.

old_paddle_value is set in main.asm. is_firing is now initialised there as well. Thanks!

The joystick/paddle code got mixed up too much and had a bunch of regular bugs. I hope that worked.

Looks a little better, but still does not work when I drag and drop the file into VICE. The bat immediately goes to the left and stays there.
It works if I do a manual stop/restore before starting.

Same behaviour on my VIC-20 with Mega-Cart when I start through the file browser. Somehow the VIA-registers are not initalized right.

Hrmph.

Now nothing touches the VIA anymore but the IRQ timer and joystick reads.