A800i (Incognito): Bringing back the RIGHT-Cart port...

[Faicuai]

...Let's face it: two cart ports is better than one, four joystick ports are better than two, five (5) motherboard slots are better than NONE... and with Incognito, yes, almost everything is better on the A800 than anything out-there...

 

...BUT, as you may contemplate all this A800 goodness, there's unfortunately a catch... In fact, there is a price to pay when upgrading to Incognito, as you may later understand. Once everything is said and done, your RIGHT-Cart port (CartB) will become unusable... sort of an engine that somehow manages to work with one cylinder while the other is pretty much dead...

 

Well, no more! Introducing OS-XL load with "rC" (right-Cart) extensions, so you can bring back to life that dormant right-port and boot your favorite right-carts reliably, and in a plug-and-play basis!

 

 

 

 

 

plus some critical improvements and optimizations NOT present on native OS-b

[flashjazzcat]

What the Hell happened to the slot name in the BIOS menu?

[Faicuai]

What the Hell happened to the slot name in the BIOS menu?

 

Because of the reverse characters? You did not know you could?

 

Typed-in on UFLASH.

Edited December 1, 2018 by Faicuai

[Faicuai]

Not showing up properly, so here we go again:

 

 

...Let's face it: two cart ports is better than one, four joystick ports are better than two, five (5) motherboard slots are better than NONE... and with Incognito, yes, almost everything is better on the A800 than anything out-there...

...BUT, as you may contemplate all this A800 goodness, there's unfortunately a catch... In fact, there is a price to pay when upgrading to Incognito, as you may later understand.... because, once everything is said and done, your RIGHT-Cart port (CartB) will become unusable in all XL/XE modes!... sort of an engine that somehow manages to work with one cylinder while the other is pretty much dead...

As you may (rightfully) imagine, contemplating that nice, internal dual-port cart-bay, with the right-slot being pretty much an ornament, can place a big toll on any of us! … Well, no more!

Introducing OS-XL load with "rC" (right-Cart) extensions, so you can bring back to life that dormant right-port and boot your favorite right-carts reliably, and in a plug-and-play basis!

A few highlights worth mentioning:

1. Based off latest, most robust (and space constrained) XEGS (rev4) 1987 OS load.
2. ALL native OS improvements and functionality remain INTACT / untouched! (e.g. XEGS keyboard, Int'l character set, etc.)
3. Includes rC extensions with re-implemented CartAB boot / initialization / start manager. Deeper analysis of OS-B and XL/XE boot logic was involved.
4. Re-coded Cart AB signature detection, validation and initialization, with CartB (RIGHT) priority, as originally intended in OSb.
5. It is now much harder to crash the boot-logic by plugging phantom / invalid 8k carts on either port. Cart-initialization and start during power-up will not crash as easily as in native OSb, and you will still have access to media boot, for further analysis.
6. NON-contiguous $8000-$BFFF memory space is supported, exactly as it is on OSb: you can now plug your right-cart and boot A800i with OPTION key-held, OS will size memory to $8000, and RIGHT-cart will be booted... while $A000-$BFFF will remain as USABLE RAM (!)
7. As an extra bonus, the following features are included on this OS load:
   • Support of 1200XL F1-F4 functions (dormant in OS code): on-the-fly ANTIC on/off for 30% boost of CPU time, International Character set ON/OFF, key-click ON/OFF, and HELP key for A800-i.
   • Avery's 2K Floating-Point Package, from Altirra OS, which runs at 2.2x-to-4x speed, and provides SUBSTANTIALLY HIGHER rounding precision and integrity than Newell's 2K package (passing ALL of AcidOS FPP tests).
   • OS load is signed with correct checksums for LOW and HIGH rom sections (8K+8K). It will pass built-in and external checksum tests.

Attached is 180K image with OS ROM image (16k) and documented source code. It is on its fifth iteration of coding and testing, good enough for a try on your A800-i, as well under ALTIRRA by using "attach Secondary Cartridge / RIGHT" feature. You can ALSO try it on XL based machines, as long as you manually join UNALTERED 8K-left and 8K-right images as a single 16K ROM image.

SDX-Scratchpad-II.atr

Enjoy!

Edited December 1, 2018 by Faicuai

[flashjazzcat]

 

Because of the reverse characters? You did not know you could?

 

Typed-in on UFLASH.

 

I knew you could, but never realised anyone would. Thanks: will fix.

[Faicuai]

 

I knew you could, but never realised anyone would. Thanks: will fix.

 

I think that is just fine! What you may need to fix, though, is the endless bottom-top option-looping that happens when returning back to menu with Atari key...

[_The Doctor__]

use browse to attach and then the forum will take your upload filename

after that you click attach

then you have to click add to post...

 

If that fails you need to tack an accepted extender to your upload...

 

it doesn't always show the error for not accepting upload types in flash 9

Edited December 1, 2018 by _The Doctor__

[flashjazzcat]

 

I think that is just fine! What you may need to fix, though, is the endless bottom-top option-looping that happens when returning back to menu with Atari key...

 

Reverse video is intended for highlighting or shortcut keys below the status line.

 

What bottom-to-top option-looping? I've never seen it and no-one mentioned it until now. How do I reproduce on my Incognito?

[_The Doctor__]

hey all that text in your post should be in the .doc or readme.txt on your disk

[Faicuai]

 

Reverse video is intended for highlighting or shortcut keys below the status line.

 

What bottom-to-top option-looping? I've never seen it and no-one mentioned it until now. How do I reproduce on my Incognito?

 

Been trying to find the exact repro-sequence.... so you can debug it.... but seems random... I will work further on it, and will try to send you meaningful info.

 

In essence, when you go BACK to Incognito menu, by pressing Atari+Reset key, the selection cursor starts rolling bottom-top (cycling) through all items on main menu, endlessly, like if someone kept the up-arrow pressed indefinitely.

Edited December 1, 2018 by Faicuai

[flashjazzcat]

In essence, when you go BACK to Incognito menu, by pressing Atari+Reset key, the selection cursor starts rolling bottom-top (cycling) through all items on main menu, endlessly, like if someone kept the up-arrow pressed indefinitely.

OK. Well, I have never heard of or seen this, but a) do you have any joysticks or other input devices connected, and b) if so, do you have joystick control of the BIOS UI enabled or disabled?

[Faicuai]

OK. Well, I have never heard of or seen this, but a) do you have any joysticks or other input devices connected, and b) if so, do you have joystick control of the BIOS UI enabled or disabled?

SORRY for the ridiculous / accidental double-topic here (no idea how it happened).

 

Anyhow:

 

1. NO joysticks connected in any port.

2. "Joysticks 3,4" option enabled.

3. "Joystick on Port #1" option enabled.

 

You are right, it may be that upon return to Incognito BIOS, that joystick port gets "stuck" somehow, and reads falsely in the "up" direction.... Maybe... it does happen independently of the OS / load you select, it is related to the i-BIOS.

Edited December 2, 2018 by Faicuai

[_The Doctor__]

the other question is what do sticks 3 and 4 map to when enabled in the other OS, When enabled how does it get around the memory upgrade selections, I couldn't readily or easily see where they went ... or did the new modified os have to resurrect that in order to work. I was not able to use that previously and I think Kyle may have had the same experience.

Is there a different hardware register that maps into the shadow or vice versa so as not to mess with memory upgrades? Just a muddy thought at this point, some clarity would be nice.

[flashjazzcat]

SORRY for the ridiculous / accidental double-topic here (no idea how it happened).

 

Anyhow:

 

1. NO joysticks connected in any port.

2. "Joysticks 3,4" option enabled.

3. "Joystick on Port #1" option enabled.

 

You are right, it may be that upon return to Incognito BIOS, that joystick port gets "stuck" somehow, and reads falsely in the "up" direction.... Maybe... it does happen independently of the OS / load you select, it is related to the i-BIOS.

 

Interesting. I'll have to see if enabling ports 3 and 4 introduces any unwanted input on ports 1 and 2. If the problem does not occur when ports 3 and 4 are turned off (does it?), then that's our answer and the obvious (and probably only possible) fix would be to disable joystick input if you don't have any joysticks plugged in.

 

the other question is what do sticks 3 and 4 map to when enabled in the other OS, When enabled how does it get around the memory upgrade selections, I couldn't readily or easily see where they went ... or did the new modified os have to resurrect that in order to work. I was not able to use that previously and I think Kyle may have had the same experience.

Is there a different hardware register that maps into the shadow or vice versa so as not to mess with memory upgrades? Just a muddy thought at this point, some clarity would be nice.

 

Had to Google it:

 

http://atariage.com/forums/topic/220641-incognito-stick3-and-strig3/

[Faicuai]

the other question is what do sticks 3 and 4 map to when enabled in the other OS, When enabled how does it get around the memory upgrade selections, I couldn't readily or easily see where they went ... or did the new modified os have to resurrect that in order to work. I was not able to use that previously and I think Kyle may have had the same experience.

Is there a different hardware register that maps into the shadow or vice versa so as not to mess with memory upgrades? Just a muddy thought at this point, some clarity would be nice.

I can tell you (for sure) that I have the same questions (what exactly does enabling 3/4 ports mean), and modified / upgrades OS-loads do no touch in any form or shape such functionality or registers.

Edited December 2, 2018 by Faicuai

[flashjazzcat]

You'd need a modified OS which maintained the shadows using the non-standard hardware registers (see the other topic), but sticks 3 and 4 would still be quite useless for any software which reads PORTB directly and expects it to hold joystick input.

[_The Doctor__]

according to that thread the cpld issue was fixed but the actual behind the scenes hardware id's were not specifically checked and named as to what they finally were and presented to most of us (?). It could be any of what was mentioned or not even though it should end up in the shadows and avoid perturbing memory and be useful with a modified OS. So it looks like it was fixed and it required a from the ground floor update. The final write up never came that the rest of us were aware of. It looks like it should work based on the outline as presented. A modified XL/XE OS should see the joysticks like an 800... but when read from the shadows. I would try with 800XL modes at 64K, then move to different memory upgrades and see how they fare(fair). Currently the XL copies the Joy1 to the Joy3 shadow and the Joy 2 to the Joy4 shadow on a quick look. But it solved very little on the software front... seems folks didn't use the shadows... so a patched OS and corrected games?

 

This could be fun in that maybe four player four stick games like The Tail of Beta Lyrae could work on the i800 in XL mode with all four sticks... pretty slick. There was more than one version of TToBL. I am not sure if they used the shadows or not, and I only ever played it on the 800 and XL constantly, I forget if yet another fixed version was needed for the XE... As the fog is heavy around my thoughts today... I seem to remember somewhat vague that I had three versions and the original was for the 800, there was an official XL modified version, and a possible hack for the XE. not at all clear on that right at this moment.

Edited December 2, 2018 by _The Doctor__

[Faicuai]

You'd need a modified OS which maintained the shadows using the non-standard hardware registers (see the other topic), but sticks 3 and 4 would still be quite useless for any software which reads PORTB directly and expects it to hold joystick input.

CORRECT, but that would not be the main goal, at this point.... It is about ensuring a viable way to work, and that would be great progress.

 

We should let the crew at Sails of Doom know that there is finally a way to read joysticks ports from A800-i in XL mode, which means 4 players can go at it WITHOUT the need of the Multijoy frankenstein.

 

Better to use native (and superior) hardware, in my humble opinion.

[flashjazzcat]

Absolutely, and have at it. The information's been out there for a long time (and the Incognito's been out for six years or so now?), and the (at least theoretical) availability of four stick inputs has been general knowledge for that entire time. I know of absolutely no serious attempt to make use of it, though.

[flashjazzcat]

BTW: You didn't tell me whether the 'looping' problem occurs with sticks 3 and 4 disabled, but I've just checked the source code and the only way I can see joystick input getting stuck is if one of the pins was pulled low for some reason. As soon as all bits read '1' (as they should with nothing connected), the stick is assumed to be in the neutral position.

[_The Doctor__]

Check both incognitos you have to see if maybe one does it and the other does not, as that might have been part of the fix as well.

[Kyle22]

I used a memory utility, I believe it was called EYE. This would copy a selected area directly to video RAM. This is a visual real time tool. I set it on the $D000 area and looked at every page, all the time I was moving the stick(s) on ports 3 and 4.

 

I never saw anything change.

 

Of course, ports 1 & 2 made a difference, but I couldn't find 3 or 4 anywhere.

[+Larry]

Not showing up properly, so here we go again:

 

 

...Let's face it: two cart ports is better than one, four joystick ports are better than two, five (5) motherboard slots are better than NONE... and with Incognito, yes, almost everything is better on the A800 than anything out-there... (snip...)

A800-rc_OS_Extensions-1.jpg

A800-rc_OS_Extensions-2.jpg

SDX-Scratchpad-II.atr

 

From your pics... What devices are you using for your nice display setup? S-video upscaler? HDMI upscaler? Or?

 

-Larry

[Faicuai]

 

From your pics... What devices are you using for your nice display setup? S-video upscaler? HDMI upscaler? Or?

 

-Larry

Considering that the setup is pretty well tested and proven at this point, here's are the ingredients:

1. The 800-i has essentially an untouched / OEM video output circuitry other than one (1) resistor to bring down the y-signal (Luminance) component, as its native voltage ranges are a bit over-driven (as shown on the scope-meter). The 800, in principle, does not need any video upgrade other than this resistor (R189 if my memory is correct).
2. Monitor is ViewSonic VP950b 5:4 (1.25x aspect ratio, which handles 4:3 very well), set in fast / high-speed mode, and auto-scale.
3. In the analog-to-digital path, the first appliance is the DVDO iScan HD/HD+ video processor. It receives ALL analog output from the Atari (either composite or Y/C, selectable with remote control) and coverts analog inputs to digital signal, as well as applying aspect ratio conversion, overscan processing, frame de-interlacing if needed (video or cinema), desired top/sides edge masking to show whatever area of the image you want, Luma, Chroma and Phase adjustments (digitally), input-to-output frame-rate correction or adaptation if needed (set to 59.92 fps unlocked), and preliminary sharpening and final upscaling, which is now set to XGA at the DVDO's DVI output.
4. Finally, the monitor performs final upscaling to 1280x1024 and additional sharpening in the digital domain, because this dual-stage image scaling and sharpening ends up outputting a more pleasant, pixel-fine image (when evaluating Atari's output in 320x192 hi-res mode, free of aliasing / beating patterns, when dots are close enough to each other). The monitor sharpens for higher spatial frequencies than the DVDO, whereas the latter sharpens more around the edges and coarse details. Both combined, produce a superb image, in both 40-columns and emulated 80-columns in SpartaDOS SDX.

Edited December 3, 2018 by Faicuai

[flashjazzcat]

BTW: You didn't tell me whether the 'looping' problem occurs with sticks 3 and 4 disabled...

 

^^Bump.