Unicorns season: Prince of Persia for the A8!

[oky2000]

Very impressive looking game, nice work.

[+DrVenkman]

And now for whatever reason, running on exactly the same hardware as my prior 2+ hour test, the demo just crashed after <1 hour. No dancing mouse this time, just static garbage.

 

[rensoup]

1 hour ago, DrVenkman said:

And now for whatever reason, running on exactly the same hardware as my prior 2+ hour test, the demo just crashed after <1 hour. No dancing mouse this time, just static garbage.

Here's the optimized mouse version (with its tail chopped off by TIX)... I'm hoping this should be good enough.

 

If not, you can move the mouse (not the princess) near the left or right screen border and it will stop animating, taking no CPU time at all. So if you still get a crash with the mouse animating, just reboot and move it outside the screen.

pop_DD_fastestmouse_notplayable.zip

[rensoup]

5 hours ago, Nickolasgaspar said:

I loaded "pop_DD_fastishmouse_notplayable.atr" demo twice. It loaded without  any issues!

In both attempts there is only one retry at the same sector ([Disk 1] Read Sector 385 ($181) #7 in track 21 ($15) [x2])

It's a little strange because at this point, it's XBIOS.COM which has control, and is just about to load my launcher.

 

For some reason it's reading the directory twice ? And then issuing retries on reading the first sector of my launcher.

[+DrVenkman]

1 hour ago, rensoup said:

Here's the optimized mouse version (with its tail chopped off by TIX)... I'm hoping this should be good enough.

 

If not, you can move the mouse (not the princess) near the left or right screen border and it will stop animating, taking no CPU time at all. So if you still get a crash with the mouse animating, just reboot and move it outside the screen.

pop_DD_fastestmouse_notplayable.zip 147.68 kB · 0 downloads

Okay, I've got the newest demo above running now - I'll let you know how it goes.

[+DrVenkman]

So I tried the new demo and oddly, the demo stopped self-running with the character just standing there. Could move him around but only one step at a time. Literally one step or jump per joystick movement. Despite the prompt, after I intentionally made him die, I could not get out of this screen. 
 

 

So since the demo didn’t self-run repeatedly, I couldn’t do a time test to see if it would crash like before. After this, I hit reset and let the game reload. We’ll see what happens next. 
 

And as I was typing, the same thing just happened. The demo ran once, then after the title screen/text card, it dropped back to the guy just standing there as if to play a game but he only moves one step at a time.

 

[rensoup]

3 hours ago, DrVenkman said:

So I tried the new demo and oddly, the demo stopped self-running with the character just standing there. Could move him around but only one step at a time. Literally one step or jump per joystick movement. Despite the prompt, after I intentionally made him die, I could not get out of this screen. 

ooops.. sorry... I messed up... here's the corrected version

 

pop_DDfastestmouse_notplayable2.zip

[+DrVenkman]

14 hours ago, rensoup said:

ooops.. sorry... I messed up... here's the corrected version

 

pop_DDfastestmouse_notplayable2.zip 147.68 kB · 14 downloads

This version has run for 2 hours and 10 minutes on my 1200XL without locking up with the Blue Mouse of Death. I'll try on other hardware now. My 576NUC+ with integrated FujiNet is out and handy, so it'll be the next target.

[+DrVenkman]

Well, that version crashed on my 576NUC+ at about an hour. Trying it now on my Incognito 800. 

[+DrVenkman]

And the new version crashed in <30 minutes on the Incognito 800. I'll try again to see if crashes as quickly a second time. 

[tjlazer]

It might be tough to make this work 100% on 1,000 different elaborate setups.  Just staying.  Maybe stick to basic setups and finish the game...  This is getting kinda out of hand.

[+DrVenkman]

29 minutes ago, tjlazer said:

It might be tough to make this work 100% on 1,000 different elaborate setups.  Just staying.  Maybe stick to basic setups and finish the game...  This is getting kinda out of hand.

Conversely, it's also Exhibit #1 for why emulator development should go hand-in-hand with real-world hardware testing at every stage of the process. The fundamental issues with XBIOS on anything but xxl's personal machine would have come to light very long ago.

 

Crashed again on my 800 in less than 30 minutes. Solid black screen this time, mouse is visible but not animating. 

[rensoup]

8 minutes ago, DrVenkman said:

Conversely, it's also Exhibit #1 for why emulator development should go hand-in-hand with real-world hardware testing at every stage of the process. The fundamental issues with XBIOS on anything but xxl's personal machine would have come to light very long ago.

 

Development on Altirra has proven to be very effective in my case. it has caught everything so far.

 

There's also no proof that there's a fault with XBIOS, it could well be an issue with the software driver that's sending the data to the machine. That would be my bet... it shouldn't fail on 3 retries, I don't think this would happen with a 1050 drive (hence no problem in Altirra)

 

I'm just trying to dodge that potential issue caused by the software driver.

 

17 minutes ago, DrVenkman said:

Crashed again on my 800 in less than 30 minutes. Solid black screen this time, mouse is visible but not animating. 

Can you try again but move the mouse to the side of the screen with the joystick until it stops animating ? This will ensure that the VBi is very fast.

 

[rensoup]

@CharlieChaplin didn't you mention it worked with your 1050 ? could you try the above version:

 

 

and let it loop for a while ?

 

[rensoup]

22 minutes ago, DrVenkman said:

Conversely, it's also Exhibit #1 for why emulator development should go hand-in-hand with real-world hardware testing at every stage of the process.

Don't forget that the respeqt software is emulating the atari disk drive hence subject to bugs as well.

[+DrVenkman]

1 minute ago, rensoup said:

Don't forget that the respeqt software is emulating the atari disk drive hence subject to bugs as well.

I'm testing via FujiNet, which handles timing-sensitive copy-protected ATX files basically perfectly; such games can and do run for hours straight on any of my machines. The ATR versions of PoP are the ONLY software that trips up any of my Atari's, vintage or modern. The CAR versions from June 10 all work great from my SIDE3 cart now that Jon has fixed the Loader issues previously discussed.  ¯\_(ツ)_/¯

 

6 minutes ago, rensoup said:

Can you try again but move the mouse to the side of the screen with the joystick until it stops animating ? This will ensure that the VBi is very fast.

Once the game has crashed, it's too late to do anything. I'll try that on the current run, since it hasn't crashed yet. Just to be clear - the mouse animates everywhere I place him all across the screen, The only way to get him to stop animating is move him completely off the screen to either side. Is that what you mean?

[rensoup]

6 minutes ago, DrVenkman said:

Once the game has crashed, it's too late to do anything. I'll try that on the current run, since it hasn't crashed yet. Just to be clear - the mouse animates everywhere I place him all across the screen, The only way to get him to stop animating is move him completely off the screen to either side. Is that what you mean?

it doesn't need to be totally off, just very near the edges of the screen, you'll see it stop animating

[+DrVenkman]

4 minutes ago, rensoup said:

it doesn't need to be totally off, just very near the edges of the screen, you'll see it stop animating

Sorry but that's another example of emulator-based development. On a real 4:3 CRT, there's no side-to-side overscan area. He animates from one edge of the visble screen to the other. 

 

EDIT: on my C1702 monitor, if I use the H.Pos control and adjust the screen to push the visible area off the left edge of the screen, I can expose enough of the normally-invisible area on the right edge and I can see where he stops moving. But when I re-center the screen controls properly, that spot is not normally visible. 

[rensoup]

10 minutes ago, DrVenkman said:

Sorry but that's another example of emulator-based development. On a real 4:3 CRT, there's no side-to-side overscan area. He animates from one edge of the visble screen to the other.

ah... that's a bit annoying, perhaps an LCD would display more of the overscan area... or a capture card ?

 

(You can change the overscan to anything you like in Altirra, so it can be as close to the real thing as you want and it's also helpful for debugging color changes,...)

[Level42]

1 hour ago, tjlazer said:

It might be tough to make this work 100% on 1,000 different elaborate setups.  Just staying.  Maybe stick to basic setups and finish the game...  This is getting kinda out of hand.

As if everyone has a "basic" set-up in 2021.

 

What _is_ a "basic set-up" ? Atari 600XL with 1010 cassette drive ? 

Edited June 27, 2021 by Level42

[+DrVenkman]

8 minutes ago, rensoup said:

ah... that's a bit annoying, perhaps an LCD would display more of the overscan area... or a capture card ?

 

(You can change the overscan to anything you like in Altirra, so it can be as close to the real thing as you want and it's also helpful for debugging color changes,...)

Yeah, I know that. But since I am still a dinosaur and all my 8-bit Atari systems are connected to 4:3 CRTs it's never been something I've bothered with. Having said that, I've had the mouse moved entirely off-screen to the right and nothing has crashed yet.

However, and I haven't gotten a photo of it yet, this current version is leaving some graphical garbage on the screen from the right-most flickering torch at the end of the battle with Jaffar on what should be a black screen. I was fiddling with H.Pos on my monitor and happened to have it moved full-left when it last occurred, and there was sine additional garbage off-screen to the right. And as I'm writing this, the same thing just happened again. I managed to get a crappy, badly-exposed photo but you can see what I mean. 
 

The screen wasn’t adjusted as far left as it had been the first time I saw it so the noise at right was only partly visible. This seems to occur every time through the demo. 

[rensoup]

1 hour ago, tjlazer said:

It might be tough to make this work 100% on 1,000 different elaborate setups.  Just staying.  Maybe stick to basic setups and finish the game...  This is getting kinda out of hand.

True but the retries error were a mistake on my part causing the game to load much slower so at least now it should load at 1x SIO again...

[rensoup]

15 minutes ago, DrVenkman said:

The screen wasn’t adjusted as far left as it had been the first time I saw it so the noise at right was only partly visible. This seems to occur every time through the demo.

I'll check that though it's low priority...

 

btw I found some comments in the respeqt github, I don't know if they're still accurate:

Quote

diskimagepro.cpp

 

// Many PRO images are working well with this implementation but some titles refuse to run.
// These titles do not work either when using 810 firmware emulation.
// The non working titles are
//   Ballblazer
//   Boulder Dash
//   Living Daylights
//   Rescue on Fractalus
//   Spy vs Spy 2
//

 

[+DrVenkman]

4 minutes ago, rensoup said:

I'll check that though it's low priority...

 

btw I found some comments in the respeqt github, I don't know if they're still accurate:

 

That's only valid for people using RespeQt. It has no bearing on my situation, using FujiNet on all my ATR attempts lately. 

[+DrVenkman]

Blue (NO MOUSE) of Death again on the 800 after about an hour, even with the mouse off the screen completely to the right.