mariu5
AltomareSome neat info about 70s hard drive contaminants.
From the DEC RK05 disk drive maintenance manual (DEC-00-HRK05-C-D)
(🖬🖫). ₍^. .^₎⟆@altomare oh to be a disk lint speeding by at 58 miles per hour. 
ltning
Ozzelot 
myrmepropagandistThis is directly related to the chapter of "the book I can't talk about it until it is done" that I was writing today!
I'm getting so close to being done it's making me nervous.
Abie@futurebird @altomare great news!
I will read anything and everything you write. Even without any ants in it!
I will read anything and everything you write. Even without any ants in it!
@CAWguy has moved@altomare 13 inch diameter!
PhilSalkie@CAWguy @altomare
How about _removable_ CDC Hawk 5MB(!) 14-inch disc platters in two-piece plastic cases like giant cake Tupperwares. We used them at ABC in the mid '80s for the TV Graphics artists (on Dubner CBG and Chyron machines) to store their images on. An 8" floppy could hold a single full-screen graphic (if it wasn't too complicated) in 1.44MB. 5MB on a Hawk (plus another 5MB internal platter) was luxury - but it was so so so so delicate, and if you even set the platter case down too hard on the table, you could crash the heads when you put it in the drive. Oh, and the drive plus controller cost about $4K (like a small car), the platters were around $100 a set. For reference, as a junior engineer at the time I made $18K a year.
I regularly lose 500,000 times that much storage in the bottom of my briefcase, because a micro SD fell out of its holder.
(Edited with proper pricing and disk sizing, thanks!)
@PhilSalkie Thanks for your backstory on how you used this tech - very interesting! You’re right about the loss of storage nowadays: other than the data it may contain, no one is really too concerned when they misplace a usb drive somewhere.
Jim Rea@PhilSalkie @CAWguy @altomare Hawk packs were 5 Mb, not 10. There was also a 5 Mb internal platter, so 10 Mb total. As for cost, the drive was about $3k, platters were about $100. Expensive, especially in 1980 dollars, but nowhere near the price of a house.
The coolest thing about those drives was the sound they made when spinning up to speed, which took 60 seconds. I wish I had a recording of it, it sounded straight out of science fiction. The packs weren’t sealed, so brushes came out to clean the disk as it spun up. If the heads did crash it made a horrible screeching noise, your data was gone and you had a very expensive repair bill. But you could use the ruined platter to make a very expensive custom clock.
@provuejim @CAWguy @altomare
In 1986, I bought a 40MB 5" hard drive and controller card for $4K, my annual salary as an engineer was $16K. I'm thinking $3K in 1983 for a box the size of a washing machine sounds low.
The 5M removable / 5MB fixed is completely correct, I'd forgotten that.
We looked at moving some of those into the engineering department once we switched the graphics gear over to 8" iOmega drives, but our offices didn't have enough power to spin one of them up.
@PhilSalkie @CAWguy @altomare I had a Hawk drive in my 1 bedroom apartment in 1981. It was plugged into an ordinary outlet, no special arrangements. I seem to recall that it took four of us to get it up the stairs though.
I wouldn’t blame you if you were skeptical that a 23 year old had one of these drives. I had been working as a programmer at a minicomputer company called Alpha Micro, and I quit and started my own company in my apartment - I wrote and sold a word processor. By the way, the software was available on floppy or Hawk pack, with a $125 extra media charge if on a pack, which basically was what i paid for the packs. I wound up sending a lot of packs first out of my apartment and then my house. I haven’t had the minicomputer or the drive for a long time, but I still have the 19” rack with a glass door in my garage.
@provuejim @PhilSalkie @CAWguy Was this the same Alpha Micro as the one that made WD-16 S-100 boards? I've always been interested in them for the sheer oddity
@altomare @PhilSalkie @CAWguy Yes, that’s the same Alpha Micro, and my Hawk drive was connected to an S-100 system with the WD-16 (the CPU board was actually a 2 board set that plugged into adjacent S-100 slots, with a ribbon connector in between).
For about a decade my life pretty much revolved around Alpha Micro, first as an employee and then as a third party software vendor. I was even on the board of the Alpha Micro Users Society for several years, and helped to organize an annual conference. I still have friends that I keep up with from those days - it was a great community. And as you say, the computer and operating system were quite unusual and interesting.
@provuejim @PhilSalkie @CAWguy Oh wow, if I ever get my hands on these boards I now know who to pester with questions!
I didn't know they had such a long lifespan. The little info I saw led me to think they made an odd one-off board that wasn't a thing for very long. Now I have some reading to do.
So far I managed to get a WD-16 board, but it's a Pascal Microengine one with different microcode.
@altomare @PhilSalkie @CAWguy If you got your hands one one of those board sets I’d be unlikely to be of much help. You would really want to talk to my friend Paul, who used to be the head technician at Alpha Micro and built a complete working Alpha Micro system out of broken discarded parts that were going to the dumpster. Sadly, however, he passed away a couple of years ago.
I’ve heard of the Pascal Microengine, of course, but I’ve never seen one. I think at most they only sold a few hundred, if even that. I think Alpha Micro sold in the high tens of thousands, and they were multi user machines, so there were quite a lot of users. At one point I estimated that there were about 50,000 users of my word processing software. (1/2)
However, the WD-16 was obsolete even by 1980, so in 1981 they transitioned to the 68k - they had to rewrite everything because it was all in assembler. They continued with the 68k architecture for about 15 years, then switched to Intel by writing a 68k interpreter. Then they switched to just selling software, kind of like NeXT. I think they are actually out of business now, but I just checked and there is still a web site someone is paying for. But it says it hasn’t been updated since 2012. (2/2)
@provuejim @CAWguy @altomare
Awesome! Looks like the Hawk drive required a grand or two worth of controller cards to make it play, which I recall was part of the cost of my 40MB drive. Took an adapter card in the drive, then an S-100 card in the Zenith Z100.
Even at $4K-ish, that Hawk had to be a serious savings over something like an IBM 3040 - though I don't find a quoted price for those things at a quick glance.
Also, as someone who owned a (very useful) chain printer at the age of 23, I'm not surprised you owned a Hawk - what's more amazing is that you happened to see my misinformed post about them!
Petra
pedro 🍉@PetraOleum @altomare a clear concession to the metric system. It should have been 12^-3 inch.
F4GRX Sébastien@PetraOleum @altomare next logical step below the thou!
the Hearth 
@altomare of note is that those aren't microns in the dimensions there--people outside of very particular fields might not be familiar with the microinch or μ" (yes the symbol is a travesty), but those are indeed given in μ", not μm. So even smaller than you might think on first read!
-F
-F
Matt McIrvin@altomare I 100% remember seeing this diagram somewhere Back In The Day. I don't think I was looking in that DEC manual, I think someone showed it to me.
@altomare (Best guess--it was my dad, and he was showing me some piece of internal General Electric training material that had plagiarized it.)
just call me jack. 🐒@altomare good vibes
Russ Sharek
zl2todA copy of that chart was pinned to my office wall *mumble* decades ago.
The RK05 was a 2.5MB single platter disk cartridge about 18" across, and yeah, they crashed a lot.
ARGVMI~1.PIF
Ian
Tracy Thomas@altomare How did it ever work?!?
With great caution and a high failure rate, I'm sure.
Everybody migrated to the modern disk-sealed-in-drive style pretty much as soon as it became commercially feasible. Removable hard disks were pretty much extinct by the 1990s.
SyQuest and Iomega did make a few more removable HDDs in the 1990s and 2000s, but they were unreliable and generally flopped in the market.
Removable HDDs only truly succeeded when the entire drive was made removable (i.e. USB HDDs).
@argv_minus_one @altomare I was thinking maybe really good error correction was key. But I wonder if that would have worked if the error was something physically blocking the head movement.
Yeah, error correction isn't going to help much when a hair wedges itself under the head. That'll cause a head crash.
sungo@altomare @RussSharek they mean "micro inch", don't they? my dreams will be haunted by diagrams with micro inches...
StarkRG
Elric Erkose@altomare When I was at HP, I heard lore from back in the days when you could smoke in the office. Some guy notorious for smoking cigars walked up to one of the early Winchester drives and exhaled smoke from his cigar into the cabinet. The drive failed.
Captain Bara@zoowar @altomare @soviut
In the computer lab back in 1980, I remember one of the techs there who would always smoke in the lab. Her supervisor told her it would damage the hard drives. She insisted that the drive's air filter would remove the smoke particles. Somehow she was allowed to continue smoking. I don't know if they had any drive failures.
In the computer lab back in 1980, I remember one of the techs there who would always smoke in the lab. Her supervisor told her it would damage the hard drives. She insisted that the drive's air filter would remove the smoke particles. Somehow she was allowed to continue smoking. I don't know if they had any drive failures.
Collette Lynner@soviut at the time nobody really believed that cigarettes were dangerous.
@collette I think that was mostly willful ignorance; smokers were addicted so they turned a blind eye to the negative health effects.
The term "smokers cough" has existed as long as cigarettes have. While the cancer link came later, it was still pretty obvious that smoking had detrimental health effects.
Western Infidels@altomare The outer rim of a modern 3.5" 7200RPM disk travels at about 75 MPH relative to the heads. A fancy 10,000 RPM data-center drive would be about 104 MPH.
Ted Faber
RS, Author, Novelist, Prosaist@altomare
Note that the important thing here is the smoke particles. If you are a sysop and you're in charge of the computer room, and especially if you are changing the disk packs, DO NOT SMOKE. If you smell of smoke, it b real bad, man!
Note that the important thing here is the smoke particles. If you are a sysop and you're in charge of the computer room, and especially if you are changing the disk packs, DO NOT SMOKE. If you smell of smoke, it b real bad, man!
Christopher Snowhill@altomare Meanwhile Americans use anything but the metric system. Micro inches?
Karl Auerbach@altomare We had some RK05s on some of our Unix machines at SDC in the mid 1970s. We never had a failure. But they were not used very much - we mostly used RP04 (and later, RP05/6) drives.
My wife sold RK05s (and PDP-8s) when she worked at DEC.
Jinna
Atkelar
Trammell Hudson@altomare based on that diagram would you say this RK05 DECpack platter is out of spec?
Abram Kedge🏴🇨🇦 wryly@altomare on modern sealed drives, the firmware still has to deal with imperfectly flat disks. It uses head temperature to build a map of microscopically small bumps and distortions, and adjusts the fly-height dynamically to float above them.
Doug Webb@AbramKedge do you have links where I can read up more about this? Sounds like a very interesting subject
@altomare I'm sorry, I don't. I worked for Western Digital for ten years, teaching people how to get the most out of ARM processors and optimizing mission-critical functions. I went from group to group, so I picked up little snippets from everywhere.
There are some fiendishly clever people throughout the company who play tricks with fundamental physics, and then their work gets sold at commodity prices 🤷♂️
The thing that blew my mind was that as of ~2006, each magnetic 'bit' on the disk was no more than 50 atoms, yet that read-write head could locate it and read its polarity in the brief time as it passed underneath. Mind you, there was some heroic error detection and correction going on!
David Smith@altomare I wonder what the numbers for modern disks are like.
〽️ɪɢᴜᴇʟ 
ClearSkyImaging@altomare For many years I was an on-site engineer at a manufacturing business for my sins.
I still remember the Monday morning I turned up for work to find 3 CDC 300 MB drives down and an unbootable system.
The operator had found it unbootable when they arrived so had taken the disk pack out of the primary drive and tried it in 2 others after switching the drive IDs.
Result was 3 crashed disk drives which to eventually repair exhausted the company's Europe wide stock of disk drive heads.😮
I still remember the Monday morning I turned up for work to find 3 CDC 300 MB drives down and an unbootable system.
The operator had found it unbootable when they arrived so had taken the disk pack out of the primary drive and tried it in 2 others after switching the drive IDs.
Result was 3 crashed disk drives which to eventually repair exhausted the company's Europe wide stock of disk drive heads.😮
🇳🇿 
💉*9 Roger
Kernel Bob 
@altomare And yet... I once hand-aligned an RK-05 platter. I loosened the six screws at the hub, slid the platter a little bit, retightened them, and iterated until the read errors stopped. No clean room, no mad skillz, just good luck.