discovered a problem with sio2USB

[sup8pdct]

I have discovered that a couple of my hard drive images that i placed on sio2usb stick from the pc are not what they should be when read by the sio2usb interface. a whole block of approx $1300 sectors of one image appears in another image. it is fine if the files are created and never removed or moved from the stick using sio2USB but copies put onto pc for backup will be corrupted. Not all files will be affected. only those that cross particular boundary's. I have emailed Thomas Grasel about my findings and the boundary's that i have discovered so far.

The first couple of hundred of floppy images should be ok how ever.

 

James

[Marius]

I have discovered that a couple of my hard drive images that i placed on sio2usb stick from the pc are not what they should be when read by the sio2usb interface. a whole block of approx $1300 sectors of one image appears in another image. it is fine if the files are created and never removed or moved from the stick using sio2USB but copies put onto pc for backup will be corrupted. Not all files will be affected. only those that cross particular boundary's. I have emailed Thomas Grasel about my findings and the boundary's that i have discovered so far.

The first couple of hundred of floppy images should be ok how ever.

 

James

 

Hi James,

 

There are really lots of troubles with sio2usb. I hope all the users post their problems with the device, which are going to result in a perfect device in the future. I have mailed a few times with THomas Grasel about the device, and I hope he or his team are going to improve it.

 

Today I mailed him that my experience so far is that if you are setting up the USB stick once, and you leave it in your sio2usb and only play the games on that stick... it works. Still not perfect, but it works.

 

But the use of a SIO2USB is the fact it is portable... an USB stick is made to carry data with you whereever you go, and to plug it in and out the computer as much as you want.

 

The fact that the SIO2USB is not so compatible with all the regular ATR's around is a pain in the ***. I'm a mac user so the existing tool to fix atr's is not available for me. Besides that: I do not want to use a tool to fix an atr. When APE and Sio2IDE/Sio2SD are able to handle an certain ATR. Sio2USB should be able to handle it too. And so are there another few problems I hope they will fix.

 

A much better idea would be they release the firmware and tools in open source. I am sure there are better programmers around to make this baby to a better/higher level.

 

I really appreciate their work, and I know: this is all a hobby. But Sio2USB is not a cheap interface, and it is an interesting idea... so it is not to piss anyone off I write all this critical things... I want a sio2usb that does what it should do.

 

thanks

Marius

[jdh]

The fact that the SIO2USB is not so compatible with all the regular ATR's around is a pain in the ***. I'm a mac user so the existing tool to fix atr's is not available for me. Besides that: I do not want to use a tool to fix an atr. When APE and Sio2IDE/Sio2SD are able to handle an certain ATR. Sio2USB should be able to handle it too. And so are there another few problems I hope they will fix.

 

I primarily use a Mac too, I will ask the author of the ATR repair tool if the source code could be made available and if so make a Mac conversion.

[Marius]

I primarily use a Mac too, I will ask the author of the ATR repair tool if the source code could be made available and if so make a Mac conversion.

 

That is very kind of you, but not the best option. It would be nice to have that tool on Mac, and even on Linux. But I would prefer a SIO2USB that simply accepts all kind of atr's around. When I'm using a certain -preferred - cross assembler which creates ATR's that are not compatible with SIO2USB I have to use that tool EVERY time I compile and try my program on the Atari.

 

And yes... they are right... it would be better there are only 'valid header' ATR's around, but come on... if APE and Sio2SD and Sio2IDE are able to handle all ATR's... SIO2USB should be too.

 

By the way I found a few other problems with SIO2USB... and now I really doubt if their opinion about 'valid atr headers' is the right one

 

but thanks for your idea.

Marius

[Shawn Jefferson]

The ATR header is an odd beast... I don't know why they didn't just use 3 bytes to describe the density and number of sectors in an ATR and that's it.

 

I'm assuming you are describing the ATRs floating around that are less than 720 sectors in length. I don't have an USB2SIO device, so can't say for sure that's what you are talking about, but as far as the header of the ATR is concerned, those small ATR images are legit. There is a whole other argument about the utility of those small ATRs, but that's sort of beside the point I think.

[Heaven/TQA]

Marius... I am using MADS plus XBOOT and my ATRs have variable size and I haven't discover any compatibility issue with my own generated ATRs.

 

What kind of cross assembling tools are you refering to?

 

ps. did you updated the firmware?

Edited February 27, 2008 by Heaven/TQA

[sup8pdct]

I have discovered that a couple of my hard drive images that i placed on sio2usb stick from the pc are not what they should be when read by the sio2usb interface. a whole block of approx $1300 sectors of one image appears in another image. it is fine if the files are created and never removed or moved from the stick using sio2USB but copies put onto pc for backup will be corrupted. Not all files will be affected. only those that cross particular boundary's. I have emailed Thomas Grasel about my findings and the boundary's that i have discovered so far.

The first couple of hundred of floppy images should be ok how ever.

 

James

I have discovered more which is a very serious problem. Any write to an ATR image gfrom the atari to the SIO2USB that goes over sector 56636 on the usb stick will overwrite the boot sector and start writing over the 1st fat The sector size is 512 bytes on the stick so the first 32meg worth of atrs is ok. That is a lot of floppy images..........

 

James

[+Larry]

I primarily use a Mac too, I will ask the author of the ATR repair tool if the source code could be made available and if so make a Mac conversion.

 

That is very kind of you, but not the best option. It would be nice to have that tool on Mac, and even on Linux. But I would prefer a SIO2USB that simply accepts all kind of atr's around. When I'm using a certain -preferred - cross assembler which creates ATR's that are not compatible with SIO2USB I have to use that tool EVERY time I compile and try my program on the Atari.

 

And yes... they are right... it would be better there are only 'valid header' ATR's around, but come on... if APE and Sio2SD and Sio2IDE are able to handle all ATR's... SIO2USB should be too.

 

By the way I found a few other problems with SIO2USB... and now I really doubt if their opinion about 'valid atr headers' is the right one

 

but thanks for your idea.

Marius

 

I'm curious about this -- what kind of ATR headers exist? (How do they differ one to the other?) What does your incompatible header and ATR look like -- can you post one?

 

IIRC, there are "short" ATR's, there are ATR's that some versions of APE gave authenticity warnings about, there are protected ATR's (ATX's?) that indicate they have protection in them. Are there others? Is a "valid header" one that conforms exactly to the original SIO2PC specification?

 

As I said, I'm just curious, because this is an interesting issue, and I haven't seen this problem with my SIO2SD.

 

-Larry

[sup8pdct]

I primarily use a Mac too, I will ask the author of the ATR repair tool if the source code could be made available and if so make a Mac conversion.

 

That is very kind of you, but not the best option. It would be nice to have that tool on Mac, and even on Linux. But I would prefer a SIO2USB that simply accepts all kind of atr's around. When I'm using a certain -preferred - cross assembler which creates ATR's that are not compatible with SIO2USB I have to use that tool EVERY time I compile and try my program on the Atari.

 

And yes... they are right... it would be better there are only 'valid header' ATR's around, but come on... if APE and Sio2SD and Sio2IDE are able to handle all ATR's... SIO2USB should be too.

 

By the way I found a few other problems with SIO2USB... and now I really doubt if their opinion about 'valid atr headers' is the right one

 

but thanks for your idea.

Marius

 

I'm curious about this -- what kind of ATR headers exist? (How do they differ one to the other?) What does your incompatible header and ATR look like -- can you post one?

 

IIRC, there are "short" ATR's, there are ATR's that some versions of APE gave authenticity warnings about, there are protected ATR's (ATX's?) that indicate they have protection in them. Are there others? Is a "valid header" one that conforms exactly to the original SIO2PC specification?

 

As I said, I'm just curious, because this is an interesting issue, and I haven't seen this problem with my SIO2SD.

 

-Larry

I would like to know as well

[Defender II]

I hope the bugs are worked out soon. Is it just a sdoftware problem?

I'd like to get one.

It sure would make transfers easier. My PCs & Ataris are in differen't rooms.

[sup8pdct]

I hope the bugs are worked out soon. Is it just a sdoftware problem?

I'd like to get one.

It sure would make transfers easier. My PCs & Ataris are in differen't rooms.

 

The creators have replicated the problem I found and are at this moment tracking down the bug.

It only shows it self when going over every 32meg boundary on the USB stick. How many single density floppies can you fit into 32meg?

I calculate 364.

or 182 double density floppy images.

That is a lot of data.

Otherwise it works well.

 

James

[jdh]

Got an email that a new firmware was released today.

 

The ATR length issue was annoying me so I've also written a small command line utility (only tested with g++ on my Mac OS X Tiger but written in 'standard' C++) that will correct those ATR lengths on single density images - http://www.puisoft.co.uk/atari/utils/