Nico de Jong wrote: The error recovery on serpentine recorded tapes uses a tape block number to do data recovery. The QIC tapes did not backspace to do error recovery in general. If a read after write occurred, it would throw out more blocks of the error data, and at read time the order would have to be sorted out by the read firmware. For whatever brilliant reason, the QIC designers and implementors never designed in any useful error recovery and published it. So as a result, the fact that the QIC tape firmware throws it's hands up and dies leaves you with your problem. I would think that if you can read the track in binary, and pick out the block numbers and decode them, you should be able to do the recovery. The key is finding a drive that will read the track, and having total access to it. Sounds like Al is up to that, so you might be able to pull this off. Several "data recovery" guys I dealt with back in the day were ex Tandberg tech guys who had gotten hold of secret or confidential (to hear them tell it) diagnostic software, and some of the tape drives and they would be able to read over defective parts of the recorded tape. I dont know of any other QIC drive that had such utilities available. I believe that they used the 1gb or 2gb model tandberg's which were 5 1/4" form factor to do this. If anyone has or knows of such, maybe they can comment on whether these programs are now available. The time frame I am talking about is to 90 or so. Jim

Re: I know that a few data recovery companies have specialized DDS drives that can read past the (new) end of such an overwritten DDS tape, but they tend to be very expensive. The basic problem is that at the end of any "session" of writing to a DDS, a "logical EOT (end-of-tape)" marker is written to the tape, and the drive firmware won't let you read or seek past it. Here's a technique I've used successfully in the past ... be aware that it overwrites (loses) a bit more data :( Put the DDS tape in drive, write-enabled. Write a enough data to it to positition the tape just beyond the EOT that was written by the accidental overwrite. (Thus, new data has overwritten the EOT marker.) (You could first determine the number of kilobytes of data on the tape prior to the new EOT, and write that many plus a few more KB.) Now, unplug the power to the DDS drive, and plug it back in. The drive should power up as normal, rewinding the tape to the start. This leaves a "gray" area on the tape, at the end of your last record and the start/middle of the first remaining old data record. Attempting to read over this area may or may not result in an error! (That's important to remember!) Now... Method 1 Simply try reading the data you wrote, and then the rest of the data. If you get it with no errors, great, you're done! Otherwise... The rest depends upon the firmware of your drive ... most drives I've tried won't let you continue reading past an error (sigh). Method 2 (smart/nice firmware) Read the data you wrote, until you encounter an error. Try a few times to read, ignoring any error that comes in. When you successfully read a record, it's old user data...save it and continue with reading/saving until you hit EOT (end-of-tape). Method 3 (dumb firmware...every DDS drive I've tried) Do a "seek to EOF" to find the first end-of-file marker on the tape. This should skip over the gray area with no problem. (I assume you didn't write any EOFs in your new data.) Now, read and save the data until EOT. The drawback with method 3 is that some tapes are written with software that's brain-dead with respect to performance and data recovery. E.g., tar ... which writes no EOFs during its work. (A backup program that writes an EOF after every file will produce a DDS tape that's MUCH more recoverable than a standard tar tape.) Said differently, if the original tape was a tar tape, method 3 won't work for you. You could try skipping records instead of EOFs, but I have never had success in this. Good luck! Stan sieler at allegro.com -- Stan Sieler sieler at allegro.com www.allegro.com/sieler/wanted/index.html