Microtech Dart wrote:
Actually, I've already done something that works well for a lot of what you mention. [With] my Universal QIC tape reader.... I've read a number of different head/track configurations with this, by finding the tracks of the 9 that overlap with the 4, and choosing the correct direction on my manual switch control.
http://microtechm1.blogspot.com/2015/09/ kennedy-6450-tape-drive-data-format.html
I have a question, but first some set-up for it. I have a point in spelling this out. Over the years, a number of people have read off binary data streams from either cartridge tape drives or very old hard drives; they store the binary samples as huge files (relative to the original data). And there are devices like the Catweasel series of cards, which operate floppy disk drives and sample their binary data streams. "Sampling" means there's many bits saved, to detect each "bit" of real data on the medium. Sometimes, these methods aren't sampling, they include the drive's "decoder" to produce actual binary data from the media. So a recovered "bit" is a real "data bit". But not all of those who produce these methods, take the next step of decoding the samples into blocks of original-as-recorded data; and then decode the blocks back into the original files which were written-to the storage devices in the first place. result: there's data "recovery" but not file recovery. I've explained what I mean by that. Now, I've looked at the microtechm1 blog. Since like most blogs it's written as a sequence of events, there's no obvious-to-me summary of current progress, beyond the "most recent" entry. I don't have time to read all prior entries. I can see there's progress on decoding the recovered data; and some hints that data at the "block" level may have been recovered. Some posts about "here's a binary extraction". This is good work, this is hard work. But I'm not clear about how far it has gone; and I made a good-faith effort to look. So here's my questions. Have you 1) reproduced the actual block-by-block data from the tapes you've processed ? 2) do you have programs to decode the blocks to interpret checksums, block numbers and so on? and 3) can you therefore process the successful blocks further to recover actual, individual files? I'm not trying to minimize all the good, and necessary, technical work done. I'm asking if files have been extracted, and if so where they may be, and where the supporting programs may be. Why am I interested? I "get" your point, these methods may be useful to me, in recovering other data from other tapes or devices; as was suggested. Herb Johnson -- Herbert R. Johnson, New Jersey USA http://www.retrotechnology.com OR .net
Herb, thank you for taking the time to read my material, and ask these very well though out questions. Hopefully my answers are worthy of the effort you put in to this: 1) Have you reproduced the actual block-by-block data from the tapes you've processed ? Yes, I have. 2) do you have programs to decode the blocks to interpret checksums, block numbers and so on? Yes, I wrote them all myself, and it has been, as you say, a monumental task. The checksums are working on some, and I'm so far unable to crack the code on others. I can add more specificity if you like, but I'll sum it up to say this. QIC-11 and QIC-24, I've decoded, and writen into my programs, the CRC validation check on ever single block, but I have not yet succeeded at this with the Kennedy 6450 block format. It is proving to be very enigmatic. The majority of the tapes for the Microtech that I have were written in this Kennedy format, but the rest, a much easier QIC-11 and 3) can you therefore process the successful blocks further to recover actual, individual files? Yes, and I have done so. As I am able to work it in, I want to create more videos demonstrating how I have done this. But, since this is all a hobby for me, it is something I am working at in available time, so this could mean a week, or it could mean a couple of months. I'm certain what I've done here is not revolutionary, as many have done this with other formats, but I've never found where anyone has done this with these QIC-tape formats, with the ability to at least analyze, if not manually or programetically correct, block errors, which is what I wanted to accomplish for myself. This is where I am in my development stage of this, where I can repair some block errors, but not all of them yet, however, at least I understand where the problem is with all blocks with errors. On, say, QIC-24 systems, even though I have equipment to read that format, when a block would error out, I would have no way of knowing what the exact error was, or how I might fix it. Now I can, at least for some. "if files have been extracted, and if so where they may be" I have an example of a simple one at: http://microtechm1.blogspot.com/2015/09/minicom-disk-to-tape-copy-utility.ht... This one was a file from a single 8096-byte block on the Kennedy format tapes, but just this week, Bruce Ray of http://NovasAreForever.com has been able to execute it on some of his emulation systems for the DG Nova/Point 4, of which this vintage of Microtech these tapes are from *should* be at least somewhat compatible. But, this at least proves to me, anyway, that my method of decoding this data is actually working successfully. I'm writing another blog page on this alone, but there is so much detail and minutiae, and I'm working at getting it accurate, and at least somewhat organized and comprehensive. That may not be useful to you, however, unless you were wanting to re-create one of these similar machines. "and where the supporting programs may be". Currently, I have written them all in Lotus Script (yes, from the old IBM/Lotus Notes). They are only on my local computers and backups right now, and I've been hoping to get them well refined and commented enough to post them on my blog, so others could write their own versions of them, if they wanted. But I haven't done this yet, because they are quite messy, and I don't think this would yet add the value to the community that I hope to at least offer. Thanks for taking an interest in this, and asking these great questions! I really enjoy the discussion. -AJ On Fri, Nov 20, 2015 at 11:22 AM, Herb Johnson via vcf-midatlantic < vcf-midatlantic@lists.vintagecomputerfederation.org> wrote:
Microtech Dart wrote:
Actually, I've already done something that works well for a lot of what you mention. [With] my Universal QIC tape reader.... I've read a number of different head/track configurations with this, by finding the tracks of the 9 that overlap with the 4, and choosing the correct direction on my manual switch control.
http://microtechm1.blogspot.com/2015/09/ kennedy-6450-tape-drive-data-format.html
I have a question, but first some set-up for it. I have a point in spelling this out.
Over the years, a number of people have read off binary data streams from either cartridge tape drives or very old hard drives; they store the binary samples as huge files (relative to the original data). And there are devices like the Catweasel series of cards, which operate floppy disk drives and sample their binary data streams. "Sampling" means there's many bits saved, to detect each "bit" of real data on the medium.
Sometimes, these methods aren't sampling, they include the drive's "decoder" to produce actual binary data from the media. So a recovered "bit" is a real "data bit".
But not all of those who produce these methods, take the next step of decoding the samples into blocks of original-as-recorded data; and then decode the blocks back into the original files which were written-to the storage devices in the first place.
result: there's data "recovery" but not file recovery. I've explained what I mean by that.
Now, I've looked at the microtechm1 blog. Since like most blogs it's written as a sequence of events, there's no obvious-to-me summary of current progress, beyond the "most recent" entry. I don't have time to read all prior entries. I can see there's progress on decoding the recovered data; and some hints that data at the "block" level may have been recovered. Some posts about "here's a binary extraction". This is good work, this is hard work. But I'm not clear about how far it has gone; and I made a good-faith effort to look.
So here's my questions. Have you 1) reproduced the actual block-by-block data from the tapes you've processed ? 2) do you have programs to decode the blocks to interpret checksums, block numbers and so on? and 3) can you therefore process the successful blocks further to recover actual, individual files?
I'm not trying to minimize all the good, and necessary, technical work done. I'm asking if files have been extracted, and if so where they may be, and where the supporting programs may be. Why am I interested? I "get" your point, these methods may be useful to me, in recovering other data from other tapes or devices; as was suggested.
Herb Johnson -- Herbert R. Johnson, New Jersey USA http://www.retrotechnology.com OR .net
-- Thanks, -AJ http://MicrotechM1.blogspot.com
On 11/21/2015 1:44 AM, Microtech Dart wrote:
Herb, thank you for taking the time to read my material, and ask these very well though out questions.
Hopefully my answers are worthy of the effort you put in to this:
I"ve responded to AJ's answers in private email, as it's a level of detail and comments that aren't that useful in a larger and public discussion. But the gist of the exchange, is that AJ is providing progress as he goes in his blog, and will come up with more concise results and descriptions of methods in the future. My remarks amount to ways of description, and of course encouragements about the work being done and described. While there's a history of reports about recovering data from magnetic media without the original computers and devices used to record them; not many reports provide details, how-I-did, and partial results. So this work is good to see online. Herb -- Herbert R. Johnson, New Jersey USA http://www.retrotechnology.com OR .net
Herb, I just thought of another example of my successful construction of a full, "large" (by early 80s standards) file from a QIC-11 format tape: http://microtechm1.blogspot.com/2015/11/microtech-m134-iris-r7-tape-read.htm... The file I produce on this page was reconstructed by my programs, from about 21,000 individual 512-byte blocks of data written to the tape. Each block of data was decoded from the GCR 4/5 data pulses that I captured with the logic analyzer, and then decoded straight to hexadecimal characters, which I stitched together in a hex editor. I think if we were to continue this conversation much further, it may greatly help me organize what I've done into a format that would be ready to share more quickly, so I appreciate this discussion very much! Before talking with you about this, I've mostly been working in a vacuum on this project. Just discussing this with you is helping my organization of it. On Fri, Nov 20, 2015 at 11:22 AM, Herb Johnson via vcf-midatlantic < vcf-midatlantic@lists.vintagecomputerfederation.org> wrote:
Microtech Dart wrote:
Actually, I've already done something that works well for a lot of what you mention. [With] my Universal QIC tape reader.... I've read a number of different head/track configurations with this, by finding the tracks of the 9 that overlap with the 4, and choosing the correct direction on my manual switch control.
http://microtechm1.blogspot.com/2015/09/ kennedy-6450-tape-drive-data-format.html
I have a question, but first some set-up for it. I have a point in spelling this out.
Over the years, a number of people have read off binary data streams from either cartridge tape drives or very old hard drives; they store the binary samples as huge files (relative to the original data). And there are devices like the Catweasel series of cards, which operate floppy disk drives and sample their binary data streams. "Sampling" means there's many bits saved, to detect each "bit" of real data on the medium.
Sometimes, these methods aren't sampling, they include the drive's "decoder" to produce actual binary data from the media. So a recovered "bit" is a real "data bit".
But not all of those who produce these methods, take the next step of decoding the samples into blocks of original-as-recorded data; and then decode the blocks back into the original files which were written-to the storage devices in the first place.
result: there's data "recovery" but not file recovery. I've explained what I mean by that.
Now, I've looked at the microtechm1 blog. Since like most blogs it's written as a sequence of events, there's no obvious-to-me summary of current progress, beyond the "most recent" entry. I don't have time to read all prior entries. I can see there's progress on decoding the recovered data; and some hints that data at the "block" level may have been recovered. Some posts about "here's a binary extraction". This is good work, this is hard work. But I'm not clear about how far it has gone; and I made a good-faith effort to look.
So here's my questions. Have you 1) reproduced the actual block-by-block data from the tapes you've processed ? 2) do you have programs to decode the blocks to interpret checksums, block numbers and so on? and 3) can you therefore process the successful blocks further to recover actual, individual files?
I'm not trying to minimize all the good, and necessary, technical work done. I'm asking if files have been extracted, and if so where they may be, and where the supporting programs may be. Why am I interested? I "get" your point, these methods may be useful to me, in recovering other data from other tapes or devices; as was suggested.
Herb Johnson -- Herbert R. Johnson, New Jersey USA http://www.retrotechnology.com OR .net
-- Thanks, -AJ http://MicrotechM1.blogspot.com
participants (2)
-
Herb Johnson -
Microtech Dart