New (alt) BIOS for Ultimate 1MB/Incognito

[flashjazzcat]

Damn... That thought briefly occurred to me too. Glad you got it sorted, and sorry for any confusion.

 

Actually this was kind of useful anyway, since it made me realize that if the flasher detects the Ultimate banking scheme but can't put the ROM into autoselect mode, it should now report "unidentified flash ROM or flashing disabled". So thanks for that.

[fujidude]

Well in that case I'm glad it happened, so it makes it better. This stuff just keeps getting better and better. Sometimes I wonder how well supported some of the other classic 8-bit computing scenes are. I feel like Atari 8-bitters are spoiled rotten with all this goodness, but maybe there is crazy awesome stuff like this for other platforms too. I could go look, but I wouldn't know what I was looking at necessarily not being familiar with much outside of A8.

[flashjazzcat]

[OT]Check out SymbOS 3 with network support on the MSX for crazily awesome on other platforms. [/OT]

[lemiel]

Maybe uflash needs to show banner with version or something like that - path maybe?

[fujidude]

Maybe uflash needs to show banner with version or something like that - path maybe?

 

Version would probably be enough.

[fujidude]

Same here. I never understand why people emulate u1mb when the emulator itself has all those features too.

 

Because it is quicker, easier, and more convenient to manipulate a lot of that stuff from within the emulated Atari rather than changing Altirra settings.

 

But modifying a8 hardware goes simply the best on real a8, so u Flash is my ultimate choice.

 

Not everyone shares the opinion that real hardware is more fun to use than emulation. There are a lot of advantages that emulation offers. Yes, there are also reasons to prefer real hardware as well. Reasonable people can differ on this preference. In my case it's a moot point anyway because I no longer have my real hardware.

 

I think most of the people who are stuck so much to pc tools never got at the point of setting up a good and smooth working a8 setup and file system.

 

You may very well think that. It may even be true (or not), but I know for a fact it isn't true for me.

 

I actually do like the UFLASH program quite a bit. I think it is a very good example of well written Atari8 software. It is less convenient for me to get all the files needed for the flashing process on a .ATR (after converting the long (and descriptive) filenames I use on the Windows host system to 8.3 format etc.), but I find that once I have it all put together, I actually like the UFLASH program better than the Windows tools in some ways. I may start doing it that way all the time.

Edited July 4, 2015 by fujidude

[Marius]

 

but I find that once I have it all put together, I actually like the UFLASH program better than the Windows tools in some ways. I may start doing it that way all the time.

 

But that is EXACTLY what I mean. As long as people do not have a working and up to date file system /WITH/ files (content) on their atari 8bit, of course they complain about the unhandy thing of first getting it on an ATR, and then moving it to the Atari HD and then finally being able to flash… in that case: yes a PC solution is probably handier.

 

I'm glad you agree about the fact that once it is all put together, it might be even handier to use than PC/Windows tools.

[danwinslow]

Well in that case I'm glad it happened, so it makes it better. This stuff just keeps getting better and better. Sometimes I wonder how well supported some of the other classic 8-bit computing scenes are. I feel like Atari 8-bitters are spoiled rotten with all this goodness, but maybe there is crazy awesome stuff like this for other platforms too. I could go look, but I wouldn't know what I was looking at necessarily not being familiar with much outside of A8.

 

Yeah, for whatever reason, the Atari scene is packed with people doing amazing stuff. I showed this forum and had a conversation with a staunch Apple 8 bit fan, and while he was not ready to admit his lifelong passion had sadly been misspent, I could tell he was jealous.

[flashjazzcat]

Version would probably be enough.

Help -> About shows the version number.

 

But that is EXACTLY what I mean. As long as people do not have a working and up to date file system /WITH/ files (content) on their atari 8bit, of course they complain about the unhandy thing of first getting it on an ATR, and then moving it to the Atari HD and then finally being able to flash… in that case: yes a PC solution is probably handier.

 

I'm glad you agree about the fact that once it is all put together, it might be even handier to use than PC/Windows tools.

Heh... I'm puzzled by the point about having to put ROMs into an ATR, since you can run uFlash from the SIDE loader and load the ROMs direct from the FAT on the CF card. And you can load the ROMs direct from a mirrored PC folder in the emulator and using AspeQt or similar. These methodologies have been amply documented in manuals, readme files, and in video walk-throughs. Maybe they're just not conspicuous enough.

Edited July 5, 2015 by flashjazzcat

[AtariGeezer]

These methodologies have been amply documented in manuals, readme files, and in video walk-throughs. Maybe they're just not conspicuous enough.

Some people just prefer to use PC tools

[voy]

Here is my setup of U1MBv2 with both BIOS 2.4 and PBI 1.3 modules. It contains SpartaDOS X 4.47, Q-Meg+4.04, XL OS 1.3, XL OS 1.3 with Hias' patch, OS B PAL version, Atari BASIC C and Alien Ambush, Galaxian, Gridrunner games in the other BASIC slots). It has been tested on the real hardware w/o problems.

Ultimate1MBv2 with BIOS v2.4 and PBI v1.3 (by Voy).bin

Edited July 6, 2015 by voy

[Marius]

OK this amazing new BIOS is out now for a few days, and I'm very curious what the user experiences are.

 

I can't imagine I ever used U1MB without this new BIOS, and I will never go back.

 

For me the Profiles are still one of the best features. I have a blue and a red profile. The write protect bit is very convenient and the OS independent high speed SIO routines where you can enable/disable HSIO for D1:-D4: independent of each other, is one of those features I consider as sugar on top.

 

uFlash was already the best invention since sliced bread… so this new Bios is the best invention since someone did come with a wheel out of the forest

[flashjazzcat]

Here is my setup of U1MBv2 with both BIOS 2.4 and PBI 1.3 modules.

Thanks for this.

 

OK this amazing new BIOS is out now for a few days, and I'm very curious what the user experiences are.

I find the relative silence (aside from the usual protests) encouraging.

Edited July 7, 2015 by flashjazzcat

[phoenixdownita]

One thing I have to ask, does it support side loader or alike to load xex, atr or is it for now only geared to SDX?

 

Also I venture it does work with MyBIOS. If not is there intent to cover that case (using MyBIOS + MyIDE2 + U1MB [mem exp only is fine])?

Edited July 7, 2015 by phoenixdownita

[Roydea6]

I would like to see it come out of Beta testing... XEGS game slot opened up.

Kind of flashy and cumbersome just to get option rich content to do what the original did without the extras.

[flashjazzcat]

One thing I have to ask, does it support side loader or alike to load xex, atr or is it for now only geared to SDX?

Yes - everything should work as before with regard to the SIDE loader.

 

Also I venture it does work with MyBIOS. If not is there intent to cover that case (using MyBIOS + MyIDE2 + U1MB [mem exp only is fine])?

I'm not a MyBIOS user, so I'll require some test material to look at. Happy to investigate.

 

I would like to see it come out of Beta testing... XEGS game slot opened up.

Under what circumstances was it closed, and under what circumstances did it "open up"? I'm assuming you have the board jumpered for XEGS functionality? I'm a little unclear about the problems you were having, and would welcome more information.

 

Kind of flashy and cumbersome just to get option rich content to do what the original did without the extras.

Many thanks. I'll look again at the original in an attempt to locate the hard disk menu, OS independent High-Speed SIO, ability to set the RTC, flash write lock, dual profiles, etc, etc, which had obviously eluded me for some years.

[Marius]

I haven't tested it lately, but I see no reason why MyBIOS won't work anymore, since MyBIOS is an OS and the 4 OS slots are still supported.

 

The New BIOS in U1MB is not an ATARI OS (!!!) … for end users it is the part that is shown when you enter the Ultimate 1MB menu. You simply can change more settings with the New U1MB BIOS and you get some more powerful features.

 

So just like the old Ultimate 1MB setup, you still can toggle the 4 OS slots, and one of them could still be MyBIOS.

[danwinslow]

Many thanks. I'll look again at the original in an attempt to locate the hard disk menu, OS independent High-Speed SIO, ability to set the RTC, flash write lock, dual profiles, etc, etc, which had obviously eluded me for some years.

 

Jon, I hope you've learned a lesson from this, and will be more assiduous in future research before writing a brilliant piece of software.

Edited July 7, 2015 by danwinslow

[TheNameOfTheGame]

 

Jon, I hope you've learned a lesson from this, and will be more assiduous in future research before writing a brilliant piece of software.

 

Lool this got me laughing. Nice retort!

[morelenmir]

I am quite intrigued by the "TurboFreezer" compatibility as this is a piece of hardware that has interested me in the past. I do not quite understand how the "Ultimate1MB" emulates or offers compatibility with it though. Can anyone explain that to me?

[flashjazzcat]

I am quite intrigued by the "TurboFreezer" compatibility as this is a piece of hardware that has interested me in the past. I do not quite understand how the "Ultimate1MB" emulates or offers compatibility with it though. Can anyone explain that to me?

Having read somewhere that problems were assured when using the two together (Ultimate and TF), I was surprised when Hias told me that only a few software factors were causing some issues, and preventing (for instance) the Freezer debugger from displaying and working correctly when the machine was frozen while inside the BIOS. What's different now is that:

 

• A copy of the Atari OS ROM font is present in the BIOS ROM at the usual address ($E000), which the Freezer's debugger requires
• OS compatible NMI (DLI and VBI) vectors are used, which allow the TF debugger's keyboard input to work correctly
• The Freezer's cart emulation is no longer broken by the BIOS clearing out CCTL

Although I don't own a Turbo-Freezer, the designer of the modern version of TF (Matthias Reichl, aka Hias, who also wrote the High-Speed SIO code found in the new PBI BIOS) has tested one with the new Ultimate BIOS and assures me the changes have had the desired effect.

[fujidude]

I'm looking for some advice regarding the accelerated SIO. My setup is Altirra, SDX 4.47, and emulated U1MB. What is the most advantageous way for me to get accelerated SIO?

1. Altirra SIO acceleration
2. U1MB
3. SDX

Of the 3, I'm guessing SDX is the least desireable? But what is most, and why?

[phaeron]

I'm looking for some advice regarding the accelerated SIO. My setup is Altirra, SDX 4.47, and emulated U1MB. What is the most advantageous way for me to get accelerated SIO?

 

Having the emulator handle the SIO request directly is always the fastest choice as it effectively does the read in zero emulation time. You won't be able to beat it with any 6502-based path. For speed, in order of preference:

1. Direct SIO hooking (SIO Patch or System > Acceleration >D: patch). This will get you instantaneous disk reads and writes. Downside is that it won't work at all when custom SIO routines are used or if you've got something trying to remap the disks, i.e. PBI disk emulation. Use MAP in SDX to force use of SIO, i.e. MAP 1 OS for D1:.
2. PBI-based transfers to/from emulated IDE device. This is the fastest possible 6502-based transfer as it simply involves hammering a data port, and it's reasonably compatible. Downside is that you need to go through a doubly-emulated disk....
3. Burst transfers (System > Disk Drive > Burst or System > Acceleration > SIO Burst Transfers in Altirra). This uses the normal SIO routines but forces the data through as fast as the transfer routine can handle, which most of the time will still exceed a real-world SIO divisor 0 transfer in speed. In versions of Altirra up through 2.60, you'll have to choose the appropriate burst mode for whether the transfer routine is interrupt-based or polling-based. In 2.70, this is no longer necessary and burst writes are also supported. Unlike #1 and #2, this method will work with most custom loaders.
4. Use a disk emulation mode that supports fast SIO and a DOS or patched OS ROM that supports that same mode. Altirra supports both XF551-style and US Doubler-style fast transfers, depending on the emulation mode. The upside is that this is the same as you would get on real hardware, but the downside is that this will only work if the DOS or OS ROM supports it, which the stock ones don't.

The H: and PCL: (PCLink) devices are also viable alternatives if you only need file-level I/O instead of sector-level I/O. H: is good for DOS/BASIC, and PCLink is for SDX. PCLink can be accelerated if the driver is launched with the correct switch, either /A or /S (I forget). Altirra supports burst I/O on H:, so large reads and writes will be very fast.

[fujidude]

Phaeron, thank you very much for writing that up! I really appreciate that you not only answered the question, but provided additional useful info that can answer the same question but with different configurations. It has helped me learn I have been doing it wrong, well not wrong but not completely optimal either. I have set Altirra to accelerate SIO for D: and PRT:, and also have enabled burst. But what I did not do, was put a /A onto the DEVICE SIO so that Sparta would not use it's own high speed SIO, and rather go with the standard OS ones. With Altirra acceleration turned on, I'm assuming it is those OS ones which are intercepted and replaced with host level speeds? You're so right, nothing in the 6502 (or even 65C816) path is going to beat my host level speeds.

[fujidude]

BTW, with PCLINK.SYS, it is the /S switch that causes it not to use SpartaDOS SIO acceleration for anyone who is wondering. The MAN page for PCLINK did not document the possibility to use the /S switch. It just stated that by default it uses the SD accelerated SIO. The SpartaDOS X 4.47 User's Manual does document the switch.

 

Attention DLT (if you're reading), might be good to show the option in the MAN page. Also, consider making the switch consistent between the SIO and PCLINK drivers, since it performs the same function in each. For now, with SIO.SYS I will think of /A as meaning "(A)tari OS" and /S in PCLINK.SYS as "(S)tandard Atari SIO."