APT Hard Disk Preparation and Utilities

[+Roydea6]

My Incognito upgrade went well, My Ultimate has an AMD PLCC and the program just asked to quit. The Incognito had to be turned off for the Drives to be set to normal.

The FDISK.COM program went well and looks very nice ... Koodos..

[Kyle22]

I was hoping to keep the cart port open for BASIC-XE. I have already wired my 1200 as an XEGS, and made both ROM sockets wired in parallel. I broke a pin or two off the MyIDE, and lifted some traces while replacing the pin header on the MyIDE board. My eyes aren't as good as they once were, but I'm going to try and repair it. I can't afford a new one or an SIDE at this point.

 

I'll let everyone know if / how it works.

[flashjazzcat]

My Ultimate has an AMD PLCC and the program just asked to quit.

Asked to quit when - before the flash? It should have reported the Manufacturer and Device ID in hex if it didn't recognize the chip:

 

Unrecognized flash media: $xx $xx

 

 

Can you tell me what those hex numbers are, Roy?

 

 

[flashjazzcat]

I was hoping to keep the cart port open for BASIC-XE.

Yeah - fair point. Keeping the cart port free is the main reason I added a PBI port to my 1200XL, and then I went and put an IDEa interface inside it anyway so the PBI port's become redundant. Moreover, I don't currently use any application cartridges. Well - at least I've got flexibility.

 

Regarding external carts: perhaps someone could hack a build of Basic-XE to run from the upper 32 banks of SIDE? You can replace the XEX loader up there with pretty much anything you like.

Edited January 5, 2014 by flashjazzcat

[+Roydea6]

UM1 flash reports -unrecognized flash media: 0x01 0xA4

With the AMD chip PLCC

[flashjazzcat]

UM1 flash reports -unrecognized flash media: 0x01 0xA4

With the AMD chip PLCC

 

I make that an AM29F020 Ray - a 256KB AMD flash ROM. Should be 512KB (AM29F040) ???

Edited January 5, 2014 by flashjazzcat

[+Roydea6]

I have been using it since I installed and It has excepted nearly every flash I have installed on it

It is a (AM29F040)

Can I swap out for an SST 39SF040 chip

[flashjazzcat]

I have been using it since I installed and It has excepted nearly every flash I have installed on it

It is a (AM29F040)

Can I swap out for an SST 39SF040 chip

 

No problem Roy - I just thought that was the ID for a 29F020 - obviously I'm wrong.

 

Leave it in - we need to make sure it gets recognized properly. I will investigate the datasheet...

[Kyle22]

Regarding external carts: perhaps someone could hack a build of Basic-XE to run from the upper 32 banks of SIDE? You can replace the XEX loader up there with pretty much anything you like.

I have the 5 in 1 OSS cart. It has a 27c010 in it. Unfortunately NOT socketed I was hoping that someday there will be a way to re-do some of the logic in the U1MB board so we could flash the OSS languages in, (except Writer's Tool) and select among them with the menu. I'm sure it'll be a long wait, though.

 

BTW, I forgot to tell you I have 39SF040's in both my Incognito and U1M.

Edited January 5, 2014 by Kyle22

[flashjazzcat]

I have been using it since I installed and It has excepted nearly every flash I have installed on it

It is a (AM29F040)

Can I swap out for an SST 39SF040 chip

 

Turns out the AMD IDs were wrong. Try this:

 

PBI Update 0.4b.zip

 

I have the 5 in 1 OSS cart. It has a 27c010 in it. Unfortunately NOT socketed I was hoping that someday there will be a way to re-do some of the logic in the U1MB board so we could flash the OSS languages in, (except Writer's Tool) and select among them with the menu. I'm sure it'll be a long wait, though.

 

BTW, I forgot to tell you I have 39SF040's in both my Incognito and U1M.

 

Yeah, it would have been really nice to have OSS supercart emulation in Ultimate. Funnily enough, I need the 64KB GUI bank to boot like a banked supercart, so maybe you'll be able to repurpose that space when Candle gets that fixed up for me.

 

Thanks for confirmation on the flash chip type too.

[Kyle22]

The update 0.4b also works fine on my 39sf040.

[flashjazzcat]

The update 0.4b also works fine on my 39sf040.

 

Good - should do: all I did was correct the AMD IDs.

 

Anyway: thanks Roy and Kyle and anyone else who's tested the flasher. I'll probably release the final version tomorrow, along with FDISK, and then the rest of the tools after that. I'd expected to release everything all at once, but this has proved rather impractical and would probably be ill-advised. There's such a large quantity of tools and drivers, it's probably best to deal with things a little bit at a time.

Edited January 5, 2014 by flashjazzcat

[+Roydea6]

Yes 4b work good thank you..

[flashjazzcat]

Yes 4b work good thank you..

 

Great - thanks for letting me know Roy. That's the last tick in the final box.

[AtariGeezer]

Here's my results:

Incognito:

-> Flash from SIO2SD: Good

-> Flash from Fat32: Good

 

Ultimate1MB: (Both chip types)

-> Flash from SIO2SD: Good

-> Flash from Fat32: Fail (maybe due to needing cpld update?)

[flashjazzcat]

Here's my results:

Thanks Jay. I'll try and flash my Ultimates from FAT32 tomorrow. The program has a test for a "JED Bug", but I'll have to ask Candle if this also applies to U1MB. I didn't really understand the CPLD issue at the time, so I'm not sure what the problem would be.

[AtariGeezer]

I tried FDISK on Incognito and it came up with message saying it needs to move the data, I said okay. Rebooted after writing partition and drives C: to O: are no longer readable. Reformatted drives C: to O:, but for drive O: which was 65535 sectors @ 512bytes, Format only reports 698 sectors @ 256 bytes...

[Kyle22]

If you change partition table scheme, your drives will not be usable because the pointers to the "virtual sectors" will be different.

[AtariGeezer]

No changes were made to the table...

[Kyle22]

Earlier, you said it needed to "move the Data"... I don't know for sure, better ask Jon.

[+Roydea6]

I tried FDISK on Incognito and it came up with message saying it needs to move the data, I said okay. Rebooted after writing partition and drives C: to O: are no longer readable. Reformatted drives C: to O:, but for drive O: which was 65535 sectors @ 512bytes, Format only reports 698 sectors @ 256 bytes...

Same here but I powered down and rebooted with the power switch and my partitions were ok.... see my previous post about the Incognito..

[AtariGeezer]

Earlier, you said it needed to "move the Data"... I don't know for sure, better ask Jon.

I was As FDisk opened and recognized the card, it popped up with that message...

 

Same here but I powered down and rebooted with the power switch and my partitions were ok.... see my previous post about the Incognito..

Hmmm, I wasn't so lucky... No Biggie, I can re-copy the files over

[Kyle22]

I spent a long time today re-copying files.... But, my re-format was intentional.

[flashjazzcat]

Don't forget to save and activate the partition table. It isn't automatically re-read from the disk on reboot unless you hold shift. This is because mounted ATRs etc need to persist in RAM. If you use the activate option in FDISK, not only is the table written to disk, it's also refreshed in RAM.

[flashjazzcat]

After giving it some thought, I decided that having two different options for saving the partition table was a waste of time which merely invited confusion. The attached version does away with the "Activate" option. Partition->Write now writes the partition table to disk, and issues a media change request to the driver, causing the edited table to immediately become active in RAM.

 

Regarding metadata (i.e. partition names), I've added an attribute flag (M) to the partition display, which conveniently signifies whether a partition has a metadata leader. Because toggling this attribute can have a destructive effect on partition alignment, you still have to go into Partition->Properties to change it, however. Still - it gives the user a visual cue when editing partition tables created with older versions of the program.

 

The problem with metadata arises simply because Konrad and I introduced the "global metadata flag" to APT. Previously, FDISK created partitions with metadata sectors whether you wanted them or not. This created issues with partition tables converted from the IDEa layout using Konrad's converter tool, because such partition tables necessarily did not have metadata sectors (because if the converter had appended them to the front of each partition during the conversion process, the data in the partitions would have been progressively knocked out alignment). The idea there was to painlessly convert old media to the new standard, so then FDISK had to cater for existing tables without metadata sectors. So what I did was disable them globally if the partition table already existed and didn't have metadata sectors. If you added new partitions to that table, no metadata sectors were added. Conversely, when creating a new partition table from scratch, metadata sectors were always added.

 

Of course this was entirely unsatisfactory long-term, since it offered the user no control over what was going on. Hence the global metadata flag. The user can choose - at the point of initialising the media - whether subsequent editing sessions will add metadata to newly created partitions. The flag is saved in the APT header, the default (for each new partition table) being "with metadata", unless the user toggles the flag off. The trouble is that now - when editing partition tables created with much older versions of FDISK - if the application sees that the global metadata flag is off (as was once the mandatory state of this previously "reserved" bit in the APT header), additional partitions will be created without metadata sectors - until the user remembers to toggle the global flag and write it to the disk.

 

So - there you go. If you're creating new partition tables afresh on newly initialized media, you'll automatically get space for partition names. If you're editing partition tables created with older versions of FDISK: check the state of the global metadata flag, and change to suit if need be.

 

FDISK41.zip