I assume that you also have the ability to perform a Low Level Format (LLF) of the diskettes which you want to use as "flippies". While the original side may have arrived with the correct LLF, even if the media was originally used as a DSDD media, when flipped, I doubt very much that there would be a usable LLF as a SSSD surface on the other side. Since you specifically mention a Shugart SA800, I assume that the controller and other hardware and software on your system has the ability to perform an LLF. For the PDP-11 systems which DEC manufactured, DEC explicitly omitted any possibility of performing an LLF on both the RX01 (SSSD) and RX02 (SSDD) drives. DEC always attempted to sell users DEC pre-formatted diskettes. Other manufactures of 8" floppy devices (some probably with Shugart drives) also included firmware on the controller which allowed the user to LLF any media as desired. In particular (the one drive which I used very extensively for many years about 20 years ago), the DSD 880/30, has a single RX01 / RX02 / RX03 drive which supports all four types of LLF on the media: SSSD, SSDD, DSSD and DSDD. NOTE that in all cases, the index hole for the first two examples of single sided media was close to the center line of the media. In the case of the double sided media, the index hole was about 1/2" further from the center line. In addition, while DEC software and hardware on the PDP-11 did have a command for the RX02 (SSDD) drive which used the characters "FORMAT" as the command name, the actual function of that command was definitely NOT an LLF. In order to be successful, the media HAD to already have an LLF. The "FORMAT" command was then able to change the media to SSSD from SSDD and visa versa. Media which were UnFormatted could NOT be used without first having an LLF via some other non-DEC hardware. In my experience, there is a bit of a problem with using both sides of a single sided media. While I agree that it was "done all of the time", I doubt that the percentage of the total was very high, perhaps 1% or even 2% of the time. As for the problem. it would only become a problem over time as the media was used AND accumulated dust in the jacket. When the media was flipped and the other side used, the rotation would be in the opposite direction and the dust would tend to be disturbed on the other side as it was nudged out of any locations where it had accumulated. This aspect would NOT be a problem with distribution media used as a "flippies" since the files would rarely be read, so the dust would not accumulate to the extent of becoming a problem. I remember that the manufacturer of the DSD 880/30 used a flippy (rather then two SSDD diskettes) to provide the user with the installation files. So YES!, but I recommend that only limited use be made of the media if both sides are going to be used. As for using SSDD media as DSDD media (and visa versa) and punching the extra pair of holes, there is an alternative!!!!!!!!! On my DSD 880/30, I installed a DPDT switch to flip the use of the index holes (each detector has its own electrical circuit). In the normal position of the DPDT switch, everything worked as expected. In the opposite position, all single sided media with the index hole close to the center line were (falsely) detected as double sided media. As a result, the single sided media could have an LLF done as DSDD and thereafter be used as double sided media. BUT, the extra index holes as DSDD were not required. Obviously, the reverse would have been true, but it seemed a waste to pay extra for double sided media with the index holes in a different position and use only a single side. By the way, after an LLF as DSDD media, the floppy could still be used as a SSDD media since, obviously, the LLF on the first side was identical to what the SSDD media received as its LLF. Also, I seem to remember that my attempts to LLF media which were originally SSDD into DSDD were successful over 99% of the time with all non-DEC 8" floppy media. (Surprisingly, when I attempted to perform an LLF on out of date DEC distribution 8" floppy media, a fair number could not support an LLF to the second side - although reverting back to an LLF as SSDD media was again almost always successful.) By the way, what system are you using the Shugart drive on? If it is a PDP-11, I might be able to offer additional information. If you have any other questions or I have not explained some items very well, please ask. I used the DSD 880/30 as my primary media with the RX03 as the backup for about five years until I could afford RD51 and RD52 drives. So I have a bit of experience. Jerome Fine

I also agree with your reply. It was just that the name of the "FORMAT" command was so misleading. On the other hand, as long as the SSSD sector headers were present, it was probably possible to fix the data areas using that "FORMAT" command when there were problems with the media. So maybe "FORMAT" was justified to some extent. And finally, when the RX33 drive became available when used with the RQDX3 controller, I almost had a conniption (whatever that might be) one day when I placed a PC non-LLF HD 5 1/4" floppy into the drive and a minute later removed an LLF media after using the RT-11 "FORMAT" command - THE FIRST TIME EVER DEC SUPPORTED AN LLF on floppy media. Oh well, those were the days - DEC hardware and software was so good that we were able to complain about the small stupid stuff. Now that DEC is gone, we realize what a loss it was. Or in your RX02 as well. The RT-11 "FORMAT" command, as you pointed out in your first part of your reply, was able to take a SSSD 8" media and convert it to a SSDD 8" media - using an RX02 drive, of course. As I mentioned, the DSD 880/30 disk subsystem emulated THREE RL02 drives and ONE RX03 drive: RX01 = SSSD at 496 blocks RX02 = SSDD at 988 blocks RX03 = DSDD at 1976 blocks After I added a DPDT switch into the detector circuit for the index hole difference from single sided media to double sided media, almost all the single sided media would accept an LLF as DSDD 8" media without the requirement for adding the extra index holes for double sided media. When floppy media are used as the primary backup, having 1976 blocks on each floppy was a great advantage. After I had purchased a few DSDD media at one vendor, they became hard to find again. So I spent an afternoon looking at the detection circuit and added the DPDT switch. What is quite interesting (well to me) is that another manufacturer produced a system using a DSDD 8" floppy that the controller converted to MSCP emulation, also with 1976 blocks. BUT, if I remember correctly, each sector on one side of the media was followed by the equivalent sector on the other side of the media, i.e. the same cylinder. So a DSDD media media for the SSDD portion on the first side. I did not use that system very much since I had only one and I never depended on any hardware without at least one backup system that was identical or the equivalent after I upgraded from my first PDP-11/23 that was based in a VT103 and used a DSD 880/8. That was a long time ago!!!!!!!!!!!!!!! Jerome Fine

... which were extremely expensive, if I remember correctly. Here in Canada and the US, I seem to remember $ 50 for a box of 10 DEC RX02 media. I think that unformatted 8" media were about $ 10 for a box of 10 media and even pre-formatted RX01 8" media were very little more. Some companies were alert enough to sell pre-formatted RX02 8" media at a cost MUCH below the DEC price. And YES!, I also did the same with a DSD floppy drive in a company I worked for before I acquired my own DSD floppy drive. The DEC pricing model for supplies like floppy media and tapes was one of the irritating aspects of using DEC hardware. Hobby users usually found a work around. But I guess when DEC sold only thousands of systems as opposed to millions of Intel PC systems, the marketing model was a bit different. I think I remember a show on the History Channel which described the early efforts of hardware manufacturers. The successful ones lowered their selling price below cost in the anticipation that increased quantities would lower production costs. Their increased production did lower the costs and continue to lower the selling price. The companies which attempted to initially sell at a profit were never able to increase production sufficiently to match the lower costs of the companies whose top management took the long view of total supply and demand. Jerome Fine

I have not bothered to retell my joke about DEC for a while now since I recognize how much their software is missed, especially the software!!!!!!!!!!!! "What was DEC's marketing policy?" - ADD NOTHING!! "What was DEC's hardware policy?" - ADD NOTHING!! Example: Buy a Micropolis 1325 for $ 500, ADD NOTHING to the price (add a zero to the right side) and presto, the price is: $ 5000!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Example: Buy a Micropolis 1325 and ADD NOTHING (well probably OEM it from Micropolis and change the model number to 1335 and add a zero ohm resistor to act as a jumper for the controller to detect as an RD53 - R7 was the designation) and presto a DEC RD53 is born!!!!!!!!!!!!!!!!!!!!!! At some point I purchased a number of Micropolis 1325 drives, added two posts to the circuit interface board (since I am not very good with a soldering iron and one post at a time in the hole where the R7 is placed was easier), then wire wrapped to form a jumper. PRESTO, the RQDX2 recognized the drive as an RD53! DEC was still selling the RD53 drives for at least a few thousand. Even the RD51 was just an ST412 with just DS3 instead of DS1. And don't get me to complain about the standard 4 button control panel on the BA23 which supported only ONE RD53 and an RX50 as opposed to the 6 button control panel which supported TWO RD53 drives and the RX50 sitting on the bench with long cables to the inside of the BA23 box. Why am I still so happy to use RT-11 after all these years if DEC was such a bad vendor? Jerome Fine

I agree with your recommendation. These days, I doubt if anyone uses a 1 MB DSDD 8" floppy to archive files, so the exercise is most likely to demonstrate a system's ability to function. BUT, 30 years ago, that 1 MB DSDD 8" floppy was valuable and a user might have had sufficient reasons to use just the first side of the media on an RX02 and both sides at other times in an RX03. So both index holes would have been useful. Better yet, as I suggested, place a DPDT switch into the detection circuit and any single sided media becomes double sided with a flip of the switch - no additional index holes to mess up the jacket!! In addition, if used frequently, my experience suggests there will likely be problems with dust in the jacket. When the media is flipped and the spin is in the other direction, dust will often be released from the jacket and clog the heads. For distribution (or just as an infrequent demonstration on either side of the media) purposes , a flippy has double the capacity. But keep the number of times you read the media to a minimum - at least that was my experience. Jerome Fine

Since the only system I ever used an 8" DSDD floppy was the PDP-11 using a DSD 880/30 with the (emulated) RX03 drive, I don't have sufficient knowledge of what any other manufacturers did. But, at least the floppy drive in the DSD 880/30 definitely had two sensors with two detection circuits, as mentioned by Tony. Jerome Fine

[continued after minor crisis interruption, sorry about that] I'm certainly not disputing that certain ones, presumably almost all, are SS/DS. But, that DOES cost a trivial amount extra. They have to add an "estra" sensor. There are at least three obvious analogous situations: 1) My AT 1.2M drives (a sloppy, but not very ambiguous name for it) can also read and write 360K disks (a very sloppy but not very ambiguous name for it). BUT, not all can! One of the very first "1.2M" drives that I got (an EARLY Mitsubishi 4854 with FIFTY pin interface connector!) could not reliably do "360K" The speed was fixed at 360RPM (which is easily compensated for by a 300Kbits per second data transfer rate), and there was no provision for alternate write current. LATER 4854 drives could do 360K just fine, and switched to a 34 pin connector. 2) It is generally possible to do "720K" (a very sloppy, but relatively unambiguous name for it) with a "1.2M" drive. BUT, not all. The above drive, for example, could not. Teac made 55F (720K), 55G (1.2M), and a 55FG (720K/1.2M) drives. Only the 55FG was explicitly intended to be able to do both. And the differenceS betwqeen the 55G and 55FG?? 3) "1.4M" (a sloppy name, but it's clear what is meant, and not as blatently WRONG as "1.44M"(multiply 80x2x18x512!)) can "normally" handle 3.5" "720K" (a sloppy name, but you know what is meant). BUT, that is not NECESSARILY so. NOT ALL DO. 4) Most of thsi sort of FDC can handle 4 drives. But many are hardwired for 2 drives, and some are hardwired for ONE drive! It MAKES SENSE for a newer design to be compatible with the previous one, IFF it can be easily achieved. Is "common sense" completely an oxymoron? Some manufacturers are so frigging stingy that they would skip that easy fundamental backwards compatability, just to save a few cents. How could a DS 8" drive manufacturer be unwilling to spend a few cents to be able to handle SS AND DS? I do not have an example. and that was what I was asking. How could a "1.4M" drive manufacturer be unwilling to spend a few cents to be able to also handle 720K?? But they exist! How could a "1.4M" drive manufacturer be unwilling to spend a few cents to be able to automatically detect which disk was in the drive?? IBM in their PS/2 series! How could a "1.2M" drive manufacturer be unwilling to spend a few cents to be able to handle "low density" ("360K"/"720K")?? But they exist! How could a "1.2M" drive manufacturer be unwilling to spend a few cents to have both 300RPM and 360RPM for EASIER use of both "360K" and "1.2M"?? Many early "1.2M" drives had only 360RPM; and required adding an additional 300Kbits per second data transfer rate to the existing 250K and 500K of the FDC, just to be able to use "300RPM" disks in the 360RPM drive! So, YES, there are precedents for manufacturers being so stingy as to cripple their product, when a trivial expense would provide full compatibility. -- Grumpy Ol' Fred cisin at xenosoft.com

On Sun, 15 Sep 2013, Chuck Guzis wrote: Another thing that tends to bite people with NEC (and clone) PC98 machines' floppy drives is that they almost all use a true Shugart interface (even for 3.5" drives!) and not the bastardized "simplified" IBM-PC compatible interface that became so common due to the IBM PC/AT. They may have the same number of pins on the connectors, but they are not /exactly/ pin-compatible. You can't simply plug a common IBM-PC compatible drive into an NEC PC98 machine and expect it to work (it won't). When dealing with these, the best "fix" is to clean and align the existing drive, but the alternative is to find one of the somewhat rare drives that can be set to function as true a Shugart compatible drive (changing just one or two of the signals doesn't cut it with NEC PC98 machines, it has to be able to emulate a true Shugart interface which means a /ton/ of jumpers, not just one or two). Even today people in the CNC world try to take PC98 machines to "computer repair" shops and get frustrated when the IBM-PC compatible floppy drives they buy won't work in their PC98 computers.

And then DEC took the opposite approach. DEC added "a trivial amount extra", then often sold the product for TEN times the price of the rest of the computer market. I include the example I just replied with to Tony: I have not bothered to retell my joke about DEC for a while now since I recognize how much their software is missed, especially the software!!!!!!!!!!!! "What was DEC's marketing policy?" - ADD NOTHING!! "What was DEC's hardware policy?" - ADD NOTHING!! Example: Buy a Micropolis 1325 for $ 500, ADD NOTHING to the price (add a zero to the right side) and presto, the price is: $ 5000!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Example: Buy a Micropolis 1325 and ADD NOTHING (well probably OEM it from Micropolis and change the model number to 1335 and add a zero ohm resistor to act as a jumper for the controller to detect as an RD53 - R7 was the designation) and presto a DEC RD53 is born!!!!!!!!!!!!!!!!!!!!!! At some point I purchased a number of Micropolis 1325 drives, added two posts to the circuit interface board (since I am not very good with a soldering iron and one post at a time in the hole where the R7 is placed was easier), then wire wrapped to form a jumper. PRESTO, the RQDX2 recognized the drive as an RD53! DEC was still selling the RD53 drives for at least a few thousand. Even the RD51 was just an ST412 with just DS3 instead of DS1. And don't get me to complain about the standard 4 button control panel on the BA23 which supported only ONE RD53 and an RX50 as opposed to the 6 button control panel which supported TWO RD53 drives and the RX50 sitting on the bench with long cables to the inside of the BA23 box. The cost of pre-formatted RX02 and RX50 media were additional examples which were often around $ 50 for a box of ten. About the only media with a reasonable cost was the RX33 which is identical to the PC HD 5 1/4" 1.2 MB floppy, so DEC even added the ability to the RQDX3 to do an LLF to an unformatted media. So DEC seemed to take the opposite approach of adding "a trivial amount extra", then overcharging by a ridiculous percentage, often by 1000%. I agree that prices for hardware have been falling, but mostly because the volume is so large. If cars were made that way, the cost would be less than $ 100. Jerome Fine

On Sun, 15 Sep 2013, Jerome H. Fine wrote: Thanks. As you've no doubt noticed, I am not familiar with DEC offerings, and often get them wrong. Absolutely! My first retail product was a jig for marking and punching index holes for making 5.25" disks into flippies. ("Berkeley Microcomputer Flip-Jig") And then, briefly, an 8" version for marking both sets of holes. Likewise, all of the DS 8" drives that I ever owned. I used them with [modified] TRS80 model 1, and IBM PC (with appropriate aftermarket FDCs). The answer to that specific query will await anybody who has a specific example. -- Grumpy Ol' Fred cisin at xenosoft.com

That depends on the machine. I certainly use them on PERQs and PDP's This works if you want to use a DS disk as SS in a DS drive. I believe the OP has a single-head drive with just one index sensor, so you can't swich the sensors round electrically. -tony

I beleive ESDI interfces existed for the Unibus, but they are not common. I have never seen one for the Omnibus (I did say 'PDP's' not PDP11's :-)). And if you can design an ESDI interface for the PERQ you knwo a lot more about that machine than I do... IIRC the read/write board in the RX02 (DEC) does have the connectors and switching diodes for 2 heads per drive. What I cna't rememebr is whether the standard RX02 firmware supports double-sided operation. Something tells me it does, but I would have to check my hand-commented listing to be sure. -tony