KryoFlux FAQ

All questions about how to use KryoFlux go here.
Post Reply
User avatar
mr.vince
Posts: 1873
Joined: Tue Oct 05, 2010 5:48 pm

KryoFlux FAQ

Post by mr.vince » Wed Oct 06, 2010 2:19 pm

Here are the Frequently Asked Questions for KryoFlux which are intended to grow over time. Posts to this topic are made by moderators only.

User avatar
mr.vince
Posts: 1873
Joined: Tue Oct 05, 2010 5:48 pm

Re: KryoFlux FAQ

Post by mr.vince » Wed Oct 06, 2010 2:37 pm

Q: My KryoFlux board appears to be unresponsive. DTC complains about not being able to talk to the device. Pressing RESET on the device does not help.
A: This can happen due to underpowering the board via the power supplied by the disk drive. Please unplug both. Now plug in KryoFlux to USB first, then plug in floppy drive power. Always power KryoFlux (connect to USB) first.

Q: When dumping disks with KryoFlux, transfers are erroneous ("device reported a streaming error") or don't work at all.
A: Your KryoFlux board is connected to an USB hub, or your computer already has an internal hub. Please connect the device to a real USB port.

User avatar
mr.vince
Posts: 1873
Joined: Tue Oct 05, 2010 5:48 pm

Re: KryoFlux FAQ

Post by mr.vince » Sat May 14, 2011 3:36 pm

Q: Is it true that at present, the only way to make IPF is to send it to SPS as the current software does not have the function to make IPF? Will the functions to make IPF images and write these back to disk be made available in the near future i.e are you guys working on enabling these functions for public use?

A: KryoFlux is an imaging solution. It was made to ingest a broad range of source media and directly transfer standard formats into sector dump images.

IPFs are made by preservationists that have a strong knowledge of disk coding, formats and protections. The software used for this, the CTA Software Preservation Analyser, is a highly specialised product that is available to institutions, archives and libraries. It is also available to private users, but its high price tag is very likely to be beyond what a private person is willing to invest.

Preserving software is not a one-click process although the Analyser has advanced features built in that help working on a disk very fast. Nevertheless, in the hands of an untrained person, the analyser will not be of much use.

Because of this, making IPFs is not available in DTC. These are two products with different functionality.

It is planned to bring writing of stream and draft to DTC, but please understand that this is nor a priority, neither does this mean that such files can always be successfully written to disk. IPFs are not only analysed, they contain information on how the data is going to be written (instead of all other formats known to us which store data how it was read).

It is easy to tell how data will be read back after it was written, but not how it was written, if you only know how it reads back.

The Software Preservation Society is working closely with archives around the world and does offer inspection of data and generation of IPFs as a free service. This service is limited by the time available for this task.

User avatar
mr.vince
Posts: 1873
Joined: Tue Oct 05, 2010 5:48 pm

Re: KryoFlux FAQ

Post by mr.vince » Sun May 15, 2011 5:56 pm

Q: Why isn't the Analyser available to the public for free, too?

A: Because it's a very special product that was created over the last decade. Many many many man-hours of work went into the Analyser. In the hands of a preservationist it's a powerful tool, but will be totally unusable for the unexperienced user. Think of an IDE for C++ or similar. If you can't write programs (=understand MFM and other variants of disk coding) even the most advanced IDE will be useless. It is very easy to produce broken IPFs if you do not know what you are doing. The result would be a world full of broken IPFs or IPFs made from modified sources that would be "disguised" as genuine.

Therefore it requires a huge amount of support to train new people on how to use it. This takes much time, time we would spend on playing with our kids, going to the cinema or having a good time with friends. It is not meant to sound ignorant or arrogant, but it is a high class product for a very small niche and people needing it can afford buying it. This will not only compensate us for the time spent, but also bring some revenue for further development and preservation in general.

Because of the above the product does come with a protection device (dongle) that does contain the actual licence. We therefore can't give out demos easily. But: If you think you should have it because you want / can help please do not hesitate to contact us. Asking is always allowed.

User avatar
mr.vince
Posts: 1873
Joined: Tue Oct 05, 2010 5:48 pm

Re: KryoFlux FAQ

Post by mr.vince » Tue Oct 04, 2011 4:32 pm

Q: What's a "flippy disk"?

A: A "flippy disk" is a 5.25" disk that has side 1 written with the disk flipped over. Early 5.25" drives were single sided only, which means they only had one head. To make use of the second side, side 1, of the disk, users usually punched another write enable hole into the jacket and then flipped the disk over.

Many selfmade disks for the C64 are flippy, and many games that need more than one side make use of the flippy technique as well.

User avatar
mr.vince
Posts: 1873
Joined: Tue Oct 05, 2010 5:48 pm

Re: KryoFlux FAQ

Post by mr.vince » Tue Oct 04, 2011 5:12 pm

Q: What's the problem with "flippy disks"?

A: Important! We are using the correct term "cylinder" here, which means the physical location of data on the platter. The word "track" is very often used instead of "cylinder", but usually means the lower or the upper side of a cylinder. Cylinder 0 has two sides, 0 and 1. Speaking of tracks this would be track 0 and 1. Cylinder 1 has two sides, 0 and 1, with the tracks 2 and 3. Therefore a disk with 80 cylinders (0-79) would have a total of 160 tracks (0-159).

But back to the question. There isn't. At least not as long as a flippy disk is read in a floppy drive it was made for. Newer drives, which refers to all standard PC drives made after 1985, are dual headed. To avoid mechanical problems (too much pressure applied to the platter from two sides), there is an offset between these two heads. The relative distance is 8 cylinders. This distance is irrelevant when using such a drive with a disk written with or for such a drive. When accessing a cylinder, e.g. cylinder 10, side 0 for this cylinder will appear at head 0, side 1 for this cylinder will appear at head 1.

Now let's try this with a flippy disk. Let's try and read cylinder 10 again. Side 0 will read back correctly and will read data meant for cylinder 10. Side 1 will have a problem. The disk was written in a drive with one head only. Therefore track 10 is on the exact same position on both sides of the platter. Because of the offset (which is +8 for head 1 to be precise), head 1 will read data meant for track 18, not 10.

This problem could be taken care of in software. If you know the offset, data can be shifted to appear at the correct position. But let's try to read cylinder 0. Side 0 will read back correctly again, side 1 will return data meant for cylinder 8.

We would have to step back another 8 cylinders to access data for cylinder 0... but we can't! The drive will stop stepping when reaching cylinder 0.

There is a simple solution to this problem. Drives with one head only had their disks flipped to read or write the second side. Just flip the disk, and side 1 will appear at head 0 at the correct position.

BUT: You might notice the disk is not spinning. Chances are high it really does not. It might, but only if the disk has two index holes punched into the jacket. The reason is that modern drives use the index hole to detect drive speed and if the disk is spinning at all. Using a jacket with one index hole only will make this hole appear on the wrong side which the drive can not see.

To read a flippy disk with a modern drive, there are three major options:

1. Cut the jacket open and place the platter into a jacket that does have two index holes. You can punch in another hole into the original jacket as well. Now tell this a collector and make sure you can run fast enough...

2. Add a fake index to your floppy drive. Place a small magnet on the underside of the motor spindle and attach a small hall effect sensor (it will measure the magnet going by) to the drive's electronics. The drawback is that the index you generate is not connected with the original one. So while this does work, it's not usefull for preservation.

3. Modify a drive to make it step to -8.

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest