On Feb 23, 2011, at 8:35 AM, Dave McGuire wrote: OK, remember that this *is* a trick question! You're over thinking this *a lot*! TTFN - Guy

On Feb 23, 2011, at 9:39 AM, Jules Richardson wrote: Ian's post on this was on the right track. TTFN - Guy

On Wed, Feb 23, 2011 at 02:09:28PM -0800, Rich Alderson wrote: Also, I've been told that you could "compact" the tape by breaking the reel with your finger while formatting (writing timing and mark tracks). I'm guessing that electromechanical differences between drives would influence the actual speed of the reel and therefore the usable size of the tape. Regards, Pontus

On 2/24/11 2:58 AM, Philipp Hachtmann wrote: It shouldn't matter over the range that the G888A can decode it. DECtape uses Manchester encoding, so it is self-clocking. Head skew relative to the timing tracks becomes more of an issue, but even that was mitigated by putting redundant tracks on different parts of the tape and or-ing the heads together. The trick with holding your thumb on the reel became necessary when DEC started shipping less tape on the reel than certain PDP-8 operating systems expected. G888's are odd beasts. They expect the head amplifier to oscillate when the tape isn't moving. That is needed for the up to speed circuit to work in the TC-11 (and probably in the TC-08) A while back, I added some background information on the Wikipedia entry for DECtape, going back to the Raytheon tape drives used on Whirlwind. The direct ancestor to LINCtape was the TX-2 high speed tape system designed by Tom Stockebrand. DEC patented the encoding used on DECtape, so the earlier LINC Tape was copied by a couple of other companies, in particular Computer Operations, Inc. in Beltsville, MD and sold interfaces for DEC, HP, and DG. Sept, 69 Datamation: minicomputer tape deck The C0600 LINC Tape System, a minitape system for minicomputers, was named in deference to the MIT computer project called LINC which first defined a need for such a peripheral. The C0600 consists of two tape drives, each with 150 feet of ~4 inch tape, which hooks up with a mini-cpu. Each reel of tape is capable of storing over 100,000 16-bit words in blocks of 256. The data transfer rate for the device is roughly 4.2KC, and the time required for a pass of a full tape is listed as 27 sec. The system approximates a random access storage device. It is not used as a standard tape deck, but more as a disc. The 256-word blocks are given permanent, pre-recorded numbers. The numbers are used to retrieve the 256-word data elements as though they were recorded on addressable disc tracks. Each block is given a checksum, and tape searching may be done moving the tape forward or backward. Two models are available, an "A" model which uses an I/O channel of a Varian 620/i and costs $9,950, and a "B" model which uses 620/i direct memory access, transfers through a double buffer on a cycle stealing interrupt basis, and costs $10,500. Software issued includes utilities, a loader, assembler, FORTRAN with math and I/O routines, and bootstrap. Although the Varian-compatible version is ready, models for the Honeywell 516 and 316 and the General Automation SPC 12 are expected. COMPUTER OPERATONS, INC., Silver Spring, Md. The photo in the announcement looks identical to the drives on the LINC There is a later single transport version which is shorter and the tape path is "S" shaped.

On Feb 23, 2011, at 9:40 AM, schoedel at kw.igs.net wrote: YES! Finally someone got it! :-) DEC never shipped millions of DEC Tapes. So the contents of all of the DEC Tapes ever manufactured would fit on a 1TB HDD (it's not even close so all of the issues that were brought up about file system, block size, etc are pretty much irrelevant). TTFN - Guy

I have gone through 2 boxes of TU-58 tapes with about 200 tapes in each box plus other odd assorted tapes for a total of about 500 tapes in all. I know of a number of other sites which have used several thousand tapes. At one time, the PDP-11/44 and the VAX 780 used the TU58 media as distribution tapes. How many installations existed? If you are referring, in particular, to the TU-58 tape media, then it seems a bit unlikely that if I can personally be aware of between 3000 and 5000 TU-58 tape media, that my experience will not have been duplicated a few hundred times for a total of over a million TU-58 tapes. If you have exact shipping information on how many TU-58 tapes DEC ever sold and shipped, then I will agree to that limited number. Otherwise, for a quick estimate, I can probably stack 30 to 50 TU-58 tapes (in their plastic containers for better stability) on a 1 TB HDD. With a capacity of 512 blocks (of 512 bytes per block - or a normal HDD block size) for each TU-58 tape and approximately 1,000,000,000,000 divided by 512 bytes (or about 1,953,125,000 blocks) on a 1 TB HDD, my estimate would be close to 3,814,697 TU-58 tape images could fit on a raw 1 TB HDD when no additional information such as a file structure is used. RT-11 would be ideal in that case if used to access the information. If 465 files, each with exactly 2,147,483,648 bytes each holding 8192 TU-58 tape images each were allocated on a 1 TB HDD with additional room for a FAT32 file structure, that would hold 3,809,280 TU-58 tape images. RT-11 under Ersatz-11 could easily access each of the TU-58 tape images in that case without a separate file structure being required for each of the 2 GB files. This would be an actual solution that will work, aside from the practical problem if some sort of index system to keep track of the contents of the 3,809,280 TU-58 images. But then you did not say that it was necessary to keep the index system for the TU-58 tape images on the same 1 TB HDD. I believe that is a bit more than the 3,086,419 TU-58 tape images previously estimated. Jerome Fine

As my contribution to the discussion, I would suggest that not all of the real-world tapes were full of data. - John

I did not realize the DECtape/Minitape format was used by Data General. Might you have model/part numbers? I'd like to know more about that. Thanks -- Ian ________________________________________ From: cctalk-bounces at classiccmp.org [cctalk-bounces at classiccmp.org] On Behalf Of Rich Alderson [RichA at vulcan.com] Sent: Wednesday, February 23, 2011 2:19 PM To: 'General Discussion: On-Topic and Off-Topic Posts' Subject: RE: Bit of DEC Trivia From: Mr Ian Primus Sent: Wednesday, February 23, 2011 11:16 AM DECtapes (originally called "Microtapes") were invented for the PDP-4, and were used on the PDP-4, PDP-5, PDP-6, PDP-7, PDP-8 family, PDP-9, PDP-10, PDP-11 family, and PDP-15. They were also used on Data General systems, and the alternate format (wrapped the opposite direction around the reel) called LINCtape was used on LINC, LINC-8, and PDP-12 systems. There were three drives, the 555, the TU-55, and the TU-56; the latter two were used on 12-, 16-, 18- and 36-bit systems. Until the advent of the floppy disk, DECtape/LINCtape was common on a lot of minicomputers, and even on mainframes. Rich Alderson Vintage Computing Sr. Server Engineer Vulcan, Inc. 505 5th Avenue S, Suite 900 Seattle, WA 98104 mailto:RichA at vulcan.com mailto:RichA at LivingComputerMuseum.org http://www.LivingComputerMuseum.org/

Rich Alderson wrote: Really? Do you have a citation? All of the information I've seen is that LINCtape was invented for the LINC, and modified into Microtape/DECtape by DEC for their machines, so I wouldn't think one could really say that it was invented for the PDP-4. I suppose one could say that the relatively minor differences between LINCtape and Microtape/DECtape (mostly to support reverse transfers) were invented for the PDP-4, if that was in fact the first DEC machine to use it, and I'm not entirely convinced of that. Eric

Yes - you are correct. I just looked again at one of the old tapes that no longer functions. That "II" is INDEED there.

I'd love to laugh :) Can you explain the joke? :)