Altirra 3.00 released

[tebe]

Atari800Win 4.1

 

[ijor]

Floppy operation is non-deterministic on real hardware, yes. Until someone takes your floppy image or program within the floppy image and bakes it into a cart loader, or something hits init mode along the way. Then you're back to square one with a deterministic RANDOM. Even loading off disk isn't necessarily a guarantee of this as the "disk" may be a solid state emulation device with timing good enough to significantly reduce the range of random seeds you are going to get. This is becoming more common as use of actual floppy disk drives decline.

 

I wasn't arguing if it is good practice (or not) to rely on this non deterministic behavior. Just mentioning the fact that some protections on other platforms do depend on this, and conceivable there might be cases on the Atari 8-bit as well.

[+JAC!]

 

http://www.virtualdub.org/beta/Altirra-3.10-test28.zip

http://www.virtualdub.org/beta/Altirra-3.10-test28-src.zip

 

• Rewrote cart mapper dialog. It now has a checkbox to display technical details for each cart type to assist when trying to find something vaguely similar to an unknown cart.

Would be good, if double click on the mapper type closed the dialog.

[Fox-1 / mnx]

Finally I found something to whine over!

 

A bit of an inconvenience in Altirra is the forced DOS 2.x filename limitation.

 

Underscores, and a number as the first character, are not allowed in filenames when adding them with the image explorer. However this is in line with the DOS 2.x naming convention this is a bit annoying as almost any other DOS 2.x format-compatible DOS allow these by design. Is there a solution that won't clutter the menu's, like an override flag or something like that?

[phaeron]

Atari800Win 4.1

 

 

So what?

 

In case you didn't know, Atari800 doesn't implement the RANDOM sequence reset in initialization mode. You can write SKCTL=0 and then SKCTL=3 without restarting the random number sequence, which is not accurate. You're going to need a more compelling argument based on an actual hardware configuration rather than just what another emulator does.

[phaeron]

Finally I found something to whine over!

 

A bit of an inconvenience in Altirra is the forced DOS 2.x filename limitation.

 

Underscores, and a number as the first character, are not allowed in filenames when adding them with the image explorer. However this is in line with the DOS 2.x naming convention this is a bit annoying as almost any other DOS 2.x format-compatible DOS allow these by design. Is there a solution that won't clutter the menu's, like an override flag or something like that?

 

Yeah, whether _ and @ are allowed is currently driven by whether the filesystem code detects a MyDOS filesystem, which is probably not the best way to drive this. Do you have an idea of what the generally accepted character set and format are for other DOSes for 2.x filesystems? Altirra currently allows [a-z][a-z0-9]{0,7}(|\.[a-z0-9]{0,3}) for DOS 2.x and [a-z][a-z0-9@_]{0,7}(|\.[a-z0-9@_]{0,3}) for MyDOS.

 

As for space in the UI, that's not an issue. I've been considering adding a menu to the disk explorer anyway for options like automatic name mangling and correcting bad VTOCs.

[Fox-1 / mnx]

A quick overview of a few DOS'. (not necessarily DOS 2.x format compatible).

 

• Dos 1.0: As Dos 2.x (1st char. NOT a number, "_" and "@" NOT allowed at all)
• BiBoDos 6.4: 1-0, A-Z (1st char. may be a number, "_" and "@" NOT allowed at all)
• DOS 3: 1-0, A-Z (1st char. may be a number, "_" and "@" NOT allowed at all)
• DOS 4.0: 1-0, A-Z (1st char. may NOT be a number, "_" and "@" NOT allowed at all)
• DOS XE: 1-0, A-Z (1st char. may be a number, "_" and "@" ALLOWED anywhere)
• DOS XL 2.30: 1-0, A-Z (1st char. NOT a number, "_" and "@" NOT allowed at all)
• DOS XLE: 1-0, A-Z (1st char. may be a number, "_" and "@" NOT allowed at all)
• HappyDOS / DOS II+D v4.5M: 1-0, A-Z (1st char. may be a number, "_" and "@" NOT allowed at all)
• K-DOS: 1-0, A-Z (1st char. may be a number, "_" and "@" NOT allowed at all)
• MachDOS 3.7a: 1-0, A-Z (1st char. NOT a number, "_" and "@" NOT allowed at all)
• SuperDOS 5.1: 1-0, A-Z (1st char. may be a number, "_" and "@" ALLOWED anywhere)
• TopDOS 1.5a: 1-0, A-Z (1st char. may be a number, 1st char may NOT be "@". "$", "#", "&", "_" and "@" ALLOWED)
• TurboDOS 2.1EX: 1-0, A-Z (1st char. may be a number, "_" and "@" NOT allowed at all)
• XDOS 2.4: 1-0, A-Z (1st char. may be a number, "_" ALLOWED anywhere. "@" NOT allowed)

Note: Not all DOS' return an error at attempts to use "_" or "@". Some will just ignore that one character and skip it in the new file name. In some cases this will lead to problematic double file names or blank file names.

[phaeron]

http://www.virtualdub.org/beta/Altirra-3.10-test29.zip

http://www.virtualdub.org/beta/Altirra-3.10-test29-src.zip

 

• Fixed bug in PAL high artifacting engine that caused odd hires pixels to be ignored. This was due to an optimization accidentally being turned on too often. When the emulator detects that there is no hires in a scanline, it uses a faster routine that only samples luma at half rate and uses merged pair filter kernels. For PAL high artifacting, it was accidentally running this all the time.
• PAL high artifacting engine no longer has artifacting hue and saturation hardcoded. You may need to reset your PAL artifacting color settings as the default is likely to be pretty blown out.
• There are now combined NTSC/PAL artifacting modes that automatically select the appropriate variation of artifacting for the current video standard. Quick review: NTSC standard artifacting emulates color from even/odd pixel patterns, PAL standard artifacting emulates chroma averaging, and the high artifacting modes do full video signal simulation.
• Added a button to the System page of the configuration dialog to quickly toggle between NTSC and PAL. This command already existed, but you could only get to it by adding a keyboard binding.
• Added auto-detection for 810 revision B firmware. (Thanks Nezgar for tirelessly tracking this down!)

[Mclaneinc]

Just tried and reset my PAL colours but there seems no difference between no artifacting and standard artifacting?

 

High looks great...

 

I did read the change log a few times but if there's artifacting present then should there not be 'some colour'?

 

Question, not critique....

 

Also, the system swap between PAL & NTSC does not change the palette, design? Seems to just change the artifacting?

 

Not sure..

Edited June 4, 2018 by Mclaneinc

[phaeron]

It's possible you may have the NTSC and PAL color settings crossed -- check Adjust Colors. The default is to use separate NTSC and PAL palettes, and the one you change depends on which video standard is currently set. If you had the artifacting mode crossed though, like NTSC with PAL artifacting, then you may have gotten the palettes mixed up.

 

PAL standard artifacting only emulates the chroma blending between adjacent scanlines. Since most standard screens don't have a lot of vertical color variation -- none for the GR.0 text screen -- the only artifact you'll see is a single line of reduced color at the top. This effect is most important for APAC screens, where it injects color into the grayscale lines.

[Wrathchild]

I did notice the other day that the move to profiles and the debug crash dialog have some overlaps.

If you have a PAL 64K XL profile setup then when the crash dialog's options are used, e.g. "Change Video Mode" to change to NTSC or "Change Memory Configuration" to 320KB, the profile is updated and persisted for the next time you run Altirra.

Will this include / move to a 'try a different profile' style of approach?

[Mclaneinc]

Checked my config and I've kept NTSC and PAL separate (or thought I had) and testing with Runabout II in XL PAL mode, left PAL quirks on and used XL rather than Linear..

 

1. No artifacting is B&W

 

2. Standard artifacting is B&W

 

3. High = colours expected in the test part of the program..

 

Is this right regarding 2.? (even after 2nd test as per message)

 

Even reset my settings just in case..(which fixed the NTSC / PAL swap button)

 

As for the button swap for NTSC and PAL, yes, works now 100% colours change as per video system (must have zero'ed or whatever the palettes as you say while playing around in profiles. Because that had changed I redid the artifacting test and that was still the same for that test ie 2. was B&W in PAL

 

Its purely to make sure I'm running correctly with the new changes rather that to query the quality of the changes..I know better than that

 

It does not take a genius to realise my knowledge of the PAL and NTSC systems goes as far as number of lines on screen and that is it, beyond that you might as well talk to a chimp and teach him using treats, he'll learn, I'll just get fatter!!

 

Paul..

Edited June 4, 2018 by Mclaneinc

[k1w1]

I just updated Altirra so I could add the "810-RevB.bin" and noticed a vacant spot for the "IS Plate Disk Drive Firmware".

 

I have searched and found the article referencing this but can find no link to the rom.

 

Can someone point me in the right direction.

[Mclaneinc]

I just updated Altirra so I could add the "810-RevB.bin" and noticed a vacant spot for the "IS Plate Disk Drive Firmware".

 

I have searched and found the article referencing this but can find no link to the rom.

 

Can someone point me in the right direction.

 

 

Here you go, there's also a 4K version of the same rom, originally it was an 8K rom made up of 2 4K segments which only needed one of those with the notion that the other 4K was there for any future expansion of the IS plate..

 

I'll hunt out the 4K one for you as completeness, both are support by Avery in Altirra..

 

Just manually add it to the firmware manager..

 

Paul..

 

4K added now

 

 

ISPlate.rom

ISPlate (4k).rom

Edited June 5, 2018 by Mclaneinc

[k1w1]

Thanks Paul.

 

My Altirra firmware folder is complete again.

[Mclaneinc]

Then my work here is done...(for now)

[phaeron]

I did notice the other day that the move to profiles and the debug crash dialog have some overlaps.

If you have a PAL 64K XL profile setup then when the crash dialog's options are used, e.g. "Change Video Mode" to change to NTSC or "Change Memory Configuration" to 320KB, the profile is updated and persisted for the next time you run Altirra.

Will this include / move to a 'try a different profile' style of approach?

 

No, too difficult to do this without dictating too much of profile setup, which I don't want to do since it should be mainly user driven. I've been considering paring down the crash dialog as it's a bit overwhelming.

 

Checked my config and I've kept NTSC and PAL separate (or thought I had) and testing with Runabout II in XL PAL mode, left PAL quirks on and used XL rather than Linear..

 

1. No artifacting is B&W

 

2. Standard artifacting is B&W

 

3. High = colours expected in the test part of the program..

 

Is this right regarding 2.? (even after 2nd test as per message)

 

Yes, this is correct. Standard artifacting is meant to accommodate the most common and cheap effects; PAL color blending is the more commonly expected effect and is a lot faster than the full filtering needed for false color generation.

 

I just updated Altirra so I could add the "810-RevB.bin" and noticed a vacant spot for the "IS Plate Disk Drive Firmware".

 

Note that all that was added recently was autodetection; the rev. B firmware works on older versions, it just isn't automatically recognized and named.

[Mclaneinc]

Thanks Avery, advice and work always appreciated..

[+JAC!]

Feature request: Would is be possible to use the scroll wheel/page up/down to scroll the memory windows of the debugger up/down (i.e. inc/dec the start address by $10)?

[phaeron]

http://www.virtualdub.org/beta/Altirra-3.10-test30.zip

http://www.virtualdub.org/beta/Altirra-3.10-test30-src.zip

• The .basic_rebuildvnt command now has a -t switch to enable TurboBasic XL label support. This allows rebuilding the VNT for protected programs that have it wiped and use TBXL labels (which otherwise would be interpreted as Atari BASIC arrays).
• The disk explorer now has options to relax filename rules for DOS 2 compatible volumes, and will attempt to encode the names of files that don't work with the rules when copying in new files.
• Double-clicking on a mapper in the cartridge mapper auto-confirms.
• XEL-CF3 with the swap button is now emulated.
• The emulator now attempts to detect if a binary program being directly loaded overlaps enabled internal BASIC, and if so will prompt to disable it. This is detected if internal BASIC is enabled and the program has a segment loading over $A000-BFFF before the first init segment.
• Build changes: localconfig .props work again, compiler bumped to VS2017 15.7.3.

 

Feature request: Would is be possible to use the scroll wheel/page up/down to scroll the memory windows of the debugger up/down (i.e. inc/dec the start address by $10)?

 

Seems reasonable and should be doable, would take some work since there's no scrolling support in that window yet.

 

[ivop]

A question: did anyone ever attempt to, instead of running Altirra under wine (Linux/macOS), recompile it with winelib?

[+DrVenkman]

 

• XEL-CF3 with the swap button is now emulated.

 

This may be silly question, but how is the swap button physically implemented (e.g., what key on the keyboard or what menu button in what window is clicked to tell the system to swap disk images)? Could something similar be incorporated for SIDE2?

 

(All of this reminds me, I still gotta buy some parts and build a -CF3 version and figure out how to mount the button on my 1088XEL),

[flashjazzcat]

This may be silly question, but how is the swap button physically implemented (e.g., what key on the keyboard or what menu button in what window is clicked to tell the system to swap disk images)? Could something similar be incorporated for SIDE2?

Highlighted is the SIDE2 button menu item, but I've circled the one you'd select for the XEL-CF swap button:

 

 

You can bind the SIDE menu button to a key via Tools->Keyboard shortcuts, but I didn't yet try the same thing with the XEL-CF swap function.

Edited June 10, 2018 by flashjazzcat

[+DrVenkman]

Highlighted is the SIDE2 button menu item, but I've circled the one you'd select for the XEL-CF swap button:

 

swap.png

 

You can bind the SIDE menu button to a key via Tools->Keyboard shortcuts, but I didn't yet try the same thing with the XEL-CF swap function.

 

Very cool. I had no idea that menu was even there.

[+Nezgar]

When I drag a file into a disk image using Disk Explorer to a SpartaDOS disk, the current time stamp is applied, not the time from the file that I dragged in. And likewise if I drag a file out of a disk image, the resulting file on the windows side also receives the current time, not the original time of the file in the disk image.

 

it would be nice have the timestamps preserved for filesystems that support them... The only way I have been able to maintain them so far is using RespeQt/SDX PC-LINK.

 

Thanks,

Ryan

 

edit: I tested this using 3.0.0 / Dec 28,2017 - I should have tested this with the latest beta, maybe it's been addressed already...

Edited June 13, 2018 by Nezgar