I still have my portable microfiche reader. You just peer through a lens and hold it up to the light.
Blue flashes and blue language await in this week's Who, Me? where a Reg reader dispenses an unexpected education to a tour group of schoolchildren. "Mike" is the teller of today's tale, which is set 30 years ago at the site of a major UK aerospace manufacturer. He was one of three engineers tasked with maintaining the banks …
a mighty 67MB of storage
RM05s were 256MB formatted, in a washing-mchine sized cabinet. I remember, because I had to back one up every Monday morning. Direct disk-to-disk using two drives, and careful checking that you'd pushed the write-protect on the source before launching the copy...
They were also somewhat temperature sensitive, and sometimes after a cold weekend they would fail their self-tests until they warmed up.
I would think, if those RM03s and RM05s were getting all that finicky, that they'd be phased out for newer equivalents. I can't find the info now, but I'm fairly sure by then vendors sold equipment for PDPs and the like that let them replace these drives with regular SCSI hard disks (possibly even IDE, although I rather doubt it.) Although I've certainly seen shops like that, if it works don't touch a thing other than the required upkeep.
My personal experience was that aerospace companies tended to keep what they had running for as long as possible because their product developments times were so long, they liked to have the same IT equipment throughout the development term (At least in the 70's, 80's and 90's). I recall PDP-84 having SCSI disks so this was definitely an option, but it would probably be like VAX's replacing RA82, RA90's with SCSI/DSSI disks it was definitely an option but if you had spent a small fortune on RA90's for instance people tended to want to get their money worth not to mention that the computer company would want to keep the nice high margin maintenance contract going
In the UK, records are supposed to be kept for something like [(useful life of airframe)+50 years] in case of any questions arising. Which is fine, you can keep that room full of 2400ft mag tapes from your ICL systems for quite a while. Of course, finding a tape drive and a computer that can still read them might be a bit more problematic...
(not to mention problems like bleed-through, heat/moisture damage, the effects of an air traffic control radar just over the road, and all those million-and-one "minor" issues the beancounters conveniently ignored...)
Several years back, when companies started asking for documentation on DVD, there was a rule that each disk had to be copied to a new one at intervals of no more than 10 years (preferably 5) as DVD writable disks were only guaranteed to be readable for 10 years. This was in oil&gas where infrastructure had operating lives exceeding 50 years - and documentation would be needed for maintenance and repair. I recall writing off kit costing >£500k because the documentation was lost and a minor weld repair was needed. “Minor” if you knew what material you were welding, a potential in-service failure if you guessed. The operator I was working for wouldn’t take the risk (a potential sub-sea leak of well fluid).
Despite their “dirty” reputation, and the sometimes “misguided decisions of middle-management”, most of the major oil companies (in the UK and Europe, at least) have strict policies on safety and environmental protection. The procedures that are a consequence of this are often mocked by frontline staff as excessively OTT. I’m sure I’ll get downvoted for that…
One place I worked had everything stored really carefully.
Software on VM's that were checked on a regular basis to see that the compiled output had the same checksums as originally, etc. (VM's copied and the copies checked and stored off-site regularly as well.)
The hardware (so much of electronic design years back was on slightly odd Pc's, cards, dongles and OS) was carefully stored, some of it on carefully segregated shelving and some of it in a nitrogen filled store room (yes, room!)
Then we were taken over by a venture capital firm who bought in new management. Over one long bank holiday weekend they had the storage areas cleared of the old equipment to make space for more business analysts and finance droids.
If a plane crashes in the next 30 years and the FAA/CAA/etc. come looking for us to replicate the original boards and software, we engineers have the personal LinkedIn and email contacts for those managers! At least 2 HW Engineers have sworn affidavits that they lodged with top law firms at the time (we may have been careful with the kit, but even more so with our arses!)
(Different AC here) I only worked with Part 91 hardware (general aviation, not airliners), but IIRC, those managers can't hide behind the corporate veil, they can find themselves personally liable under criminal and civil law. There's a reason my test records were scanned in, backed up, and still filed as paper copies.
When I (and my complete Tech Pubs department) were made redundant, our workload was handed over to Sales Department, and the Sales Engineers were tasked with writing the Instruction Books for any new contracts they sold. Manglement also junked all of our carefully preserved and catalogued manuals going back 10 years for normal contracts and 50 years for nuclear kit. Five years later I was head hunted back to set up a new Technical Publications Department because the Sales Droids had failed to keep up to date, and the company was facing several court battles for failure to complete the contracts. We never did manage to recreate all those scrapped manuals, but at least we had plenty of room to file the new ones.
(posting anonymously because of the implications).
A few years ago Big Company decided to change their backup solution from their standard issue to a different one. For legal and financial reasons, they had to keep the previously backed up records for 10 years. Since they had contracted the services of Steel Hill (wink, wink) to rotate and store the tapes, they thought they were safe. They still had to have a way to restore those records, so a recovery rig was commissioned half a country away. The plan was simple: The main backup systems' databases would be backed up to tape, then the tapes would be shipped to the restore rig's location, and the systems would be rebuilt there on newer hardware.
It was a wonderful plan.
Until someone working on the decommission of the old site mangled up the Steel Hill shipping manifests and ended up marking the database backups for deletion.
It didn't end well.
I was never a PDP-11 specialist, although working with a number of models over the years with RSX, RSTS and UNIX so my memory and the following recollections may not be 100% accurate.
I don't believe that Digital ever produced a SCSI interface for UNIBUS and hence for the PDP-11. There were certainly third-party products (EMULEX?) and I can well believe that an 11/84 (which was a relatively late model) could very well have been supplied with such through an integrator (e.g. SYSTIME in the UK but far from unique). Any systems with an RM03/05 and TU77 would probably date from the mid- or late- 1970s contemporary with PDP-10s and 11-7xx series VAXen, and would likely have been "larger" models (11/70 quite possibly) given the size of the MASSBUS interface.
MASSBUS was a pretty horrible interface anyway, and at my first job the US-parent had invested in third-party equipment to connect strings of IBM 3330 equivalent drives to their PDP-10s (KI and KL models). In many ways quite glad to see the heavy and inflexible cables superceeded by the quite effective UDA and HSC connected RA/TA series disks and tapes with much thinner and very flexible cables.
"they liked to have the same IT equipment throughout the development term (At least in the 70's, 80's and 90's)."
Space agencies still do. It's a bad habit with contemporary hardware and one I'm trying to break my employers of. The data works just fine on new hardware
In aerospace and medical applications it's also often a matter of licensing when it comes to hardware. If these are systems being used for simulators or as hardware analogs for design/software verification on other gubbins then the hardware to do that has to be proven to be analogous to the real thing. Sure you could replace the drives, but then you have to do a mountain of paperwork to certify the new hardware works in exactly the same way as the old hardware (including known bugs, timing, errors and all because they were present when the original was certified so who know if they are important or not. If they're not, you better prove that in your paperwork!)
Licensing. Ye gods, licensing.
I'm reminded of one place I worked where we had a dedicated server and a single terminal to run a timeclock application. It was command-line based; the terminal was not in fact a VT220 but was emulating one to talk to the server; and the pair of them lived on their own network, unconnected to anything else. I was a wee helpdesk grunt at the time and never got to lay finger on the servers, so I don't know what exactly was under the hood.
Nobody ever used that timeclock application.
I could not understand this lunacy and asked about it. My manager at the time told me that the owner had negotiated a steep discount for the subscription (?!) for this custom timeclock app and that it was cheaper to keep it running than to pay the penalty for stopping service.
I'm guessing it was some variety of circuit breaker or other protection, and when he pushed it in and held it for a moment (before panic ensued), it completed the circuit again, providing power to whatever failed component had tripped it in the first place, which component then provided the obligatory light-and-noise show.
Either that, or it was the magic smoke release button.
Even though me working on electrical infrastructure is an increasingly rare occurrence, to this day I still completely remove the breaker from the panel that is related to the circuit I am working on and keep it on my person.
Why? You really cannot predict what
the fuckwitts some people will do if they feel inconvenienced.
Thanks to a happening years ago which involved someone over-riding the 'lockout' I had left in place (Actually, cutting off the red 'lockout tag' padlock I had used to ensure no interference) and nearly barbecuing yours truly, all because the light in an unused stairwell wasn't working.
I still have the scar on my thumb to prove it ('Twas a 100V line public address system that bit me, if anyone is interested!)
It's taken very seriously at our facility. Had one tech suspended for 3 days without pay for removing someone else's lock - after confirming the other guy was away on vacation and that the equipment could be safely re-energized. Had he NOT confirmed these things first, it likely would have been a termination-level event. A repeat (even under the same conditions) would likely be as well.
There is a specific, established, very strict procedure for removal of a LOTO log by someone other than the person who put them on. Requires multiple people high-up at the site to sign off on it. The procedure's there for a reason.
"Someone who did that in my early days of apprenticeship was charged with attempted murder. It was taken that seriously by the authorities in Kiwiland"
And rightly so.
If I ever found out someone cut off my LOTO lock/tag there will be consequences. Police would probably get involved too. Either because I've managed to remain calm enough to file a report to them or because I haven't been able to keep calm enough about it. LOTO procedures are another one of those things where every rule has been written in blood.
"Either that, or it was the magic smoke release button."
Ah yes, we all know that all those electronic contraptions work by smoke. Along comes a newbie, and asks how do we know that, and the answer is have you ever seen one work once the smoke gets let out? Which is no.
Overcurrent reset ‐ motor windings could overheat, the insulation melted and shorted the copper winding wire together.
The trip would go, the wire would còol holding the copper tightly in contact - press the button in and BANG! - magic smoke escapes. :-)
The customer simulation department (think hydraulic-powered flight simulator platforms) presented a particular challenge.For a fleeting moment as I was reading that, I was imagining the simulation being run on one of those hydraulic computers you see as demos of the principle, usually. What a shame, especially since I used to work on aircraft hydraulics; would've been a nice cross-over.
My first read was that it was the customer that was being simulated!!!!
I figured they must be using fuzzy logic to the handle the dithering and constant changes to requirements. But then realised that a couple of PDPs would be capable of simulating the perceived IQ of the average customer quite comfortably.
On second reading I thought it would have been more fun as the customer stimulation department!
This department would dispense hookers, coke and booze right up to the moment the customer has signed the (binding) contract.
And of course the C-suite inmates would ensure that they were the ones to do the performance reviews for this department!
------------> HC&B? Who else!
One of the most impressionable moments I had when going through the Careers Advice system in my last year of school was visiting a Marine Research Lab in Hayes. I found this one of the most boring boring boring visits I went on... until our guide picked up a lump of transparent plastic with various holes wending their way through it. "This is a Fluidic switch" he announced. It was demonstrated by showing air being pumped through a similar device in a rig that had a fork in its path. Air consistently flowed from the source to one of the outlets. A puff of air was applied to a junction just prior to the fork, and flow passed to the other leg of the fork. A puff of air at an opposite junction flipped flow to the other fork. I was familiar with flip-flops but doing this using air, wow. (Yes, everyone at school thought I was a nerd).
The other memorable moment from these visits was being shown round the Hoover factory in Perivale, when it was the Hoover factory (lovely architecture, now Tesco), with a view to embarking on a career as a toolmaker. The guy that showed us around called everyone Fred, regardless of man, woman, ethnicity... Fast forward a few years to when I was doing my ONC: there was this bloke on the course who called everyone Fred: the African's, the Indian's, etc. I asked him "You don't work for Hoover do you?" "WTF How'd you know that?" was his reply IIRC. His name was John fwiw.
You can build an entire computer out of air-driven "gates" -- and the USSR did...
We did some design work for an Israeli company back in the 90s and they had a Russian guy on the project -- yes he was called Igor. Before he left the USSR he'd been working on the control computer for one of their nuclear plants (might have been military?) which was controlled by a computer powered by compressed air, he said it made about the same amount of noise as a jet engine.
To avoid reliability problems, every "gate" was built using triple-voting-redundancy, as were modules -- he said the best thing about it was that if a fault occurred (the modules indicated this) you could unplug the faulty module and plug in a replacement *while the system was running*. Try that nowadays...
Why do all this? Well, if you have some kind of nuclear accident it's quite helpful if the control system is totally immune to radiation and carries on running even during a meltdown. Which gives you some idea of how big an accident they were planning for, Chernobyl wasn't in the same league :-(
I actually worked for a year (placement year for my degree) at one of the major flight simulator manufacturers (Link-Miles). I do remember a couple of significant incidents that happened when I was there. The first one involved the catastrophic failure of a hydraulic jack when it was being tested that resulted in a 10 ton (!!!) slab of metal being thrown through an adjacent wall, fortunately with no injuries to anyone. Did I mention that the "adjacent" wall was over 40 yards away!
The second incident related to the company's IBM 360 mainframe used to support all of the design work as well as handle project management data. The operator normally ran a routine backup every morning. One morning, having gone into the next door tape store to pull out the set of tapes that would be used for this morning's backup, he heard a sudden BANG followed my a lot of tinkling sounds. When he went back into the mainframe room he discovered one of the disk pack cabinets had a hole in its side with the steel bent outwards, and the mainframe was busy falling over. It turned out that the bearings supporting the disk pack's spindle had failed, and the spindle had promptly torn loose and headed out the side of the cabinet, at which point the disk pack turned itself into some very high-speed shrapnel. Took them a couple of days to wheel in a new disk unit and restore the system.
Not Rediffusion - Singer Link-Miles (yes, Singer as in sewing machines, never figured out how they ended up doing flight simulators as well).
Rediffusion was mostly commercial systems, Link-Miles focused mostly on military stuff.
"Singer as in sewing machines, never figured out how they ended up doing flight simulators as well"
I think sewing machines were an early example of precision engineering so they would have been well-placed to serve other industries. I think these days they make missiles.
Basically Singer went on a diversification/acquisition spree, then got into trouble and dediversified. It goes something like this:
General Precision Equipment Corporation bought Link Aviation Devices. Singer bought GPE, then the Link division bought Miles Electronics. Singer reincorporated their Link division as Link Military Simulation Corporation and that became part of L3Harris Technologies. In March this year, CAE Inc (previously Canadian Aviation Electronics) announced they were buying L3Harris's military training businesses, including Link Simulation & Training.
But Singer - by then called Bicoastal Corporation - separated Link-Miles from that Link division, and they sold it to Thomson-CSF, which later became Thales.
What I noticed in this photograph was the big red button on the panel... and the proximity that Mr Bean was to it
What I noticed in this photograph was the big red button on the panel... and the proximity that Mr Bean was to it
Fear not; from the photo it appears as though his hands have been tied behind his back to discourage their being placed anywhere they shouldn't be.
And I mean anywhere...
The photo was taken within the confines of Portsmouth Harbour. The horizon to the left would be part of Gosport, and the horizon to the right part of Portchester.
WFH I'm currently sat in my 'office' and looking out of the window I have a view over where that photo was taken.
The plank and the command island are covering up the view of what would be the site of the old Harry Pounds scrap yard at Tipner.
There is a photograph of the yard featuring a view of the submarines in Koo Stark's book of photographs published in the '80s.
Ms Stark has a tenuous link to this photograph of a RN carrier - having dated Prince Andrew, who was a RN Helicopter Pilot.
I think they have masks for POTUS with the seal of that office - so, Me Too Boris has to have a Union Jack on his mask.
He's gotten himself a Boris Force One/put out of use a RAF Voyager - next he'll be wanting to acquire a Merlin or Chinook for his exclusive use - something bigger than the one that JCB loans to him.
This post has been deleted by its author
I hate people like that, everyone and their dog calls them bulbs.
Worse are people who refer to plugs as "plug tops" no its a plug not just the top...they are usually the smug know it all kind as well, who are parroting some other know it all....
Ditto fuseboxes, yes technically they are a consumer unit / distribution board, but most customers call them fuseboxes....
I remember, as a trainee, going on site with a more senior engineer to do some maintenance on a venerable system which had filament indicator lamps. My colleague decided to use some long nose pliers to remove a failed lamp, rather than the recommended extractor. The glass popped and the pliers shorted the supply resulting in a rather unceremonious 'shutdown'.
I always used the extractor...
I saw a whole different story unfolding... Little Jimmy saying what's this red one do? Little Johnny not to be left out, says watch this! And proceeds to push one of the red ones with smug confidence. The button depresses and pops to the out position and silence in that area. All that can be heard is the pitter patter of little feet scramming into the distance!
I was a little kid barely knee high to a grasshopper & was part of a tour of a computer lab. The Big Scary Man in the puffy trousers (what the grey bearded old goat looked like to me) asked me if I wanted to learn to type my name on the keyboard. I said yes, stepped up, & carefully typed in "m y n a m e". He laughed, told me I was a fun kid, & told me to hit the enter key. So I pressed enter.
At which point the PSU caught fire & shot flames out the back like a flamethrower.
Best. Tour. Ever.
We were ushered out as quickly as possible while TBSM ran for a fire extinguisher. I've been in love with computers ever since. =-D
Had a dec vax 8600 with Xerox and apple fax modems wired up to it.
Also had PDP 11/34 and a PDP 11/84 with I think TU10 tape drives with custom software and boot loaders on tape (not using the RSX installed on the disk packs), those had custom hardware with fibreoptics to draw graphs onto roll fed Kodak film (had a dark room as well to process the film).
On a microvax, they had an Ibis disk drive which had 4 sets of read/write heads, the thing used to rock back and forth when in use. if the lid was up for maintenace, it had something like a 6 inch wide band connecting the motor to the fixed disk - not something you'd want to get a tie caught in when it was running unless you were into instant beheading.
On and off for years, been trying to find a picture of the Ibis without luck.
We used them for backups of our two vaxen in my first job. The problem came when someone was demanding a restore of a particular file urgently and there were several disk packs to search through. We hit the problem of the drive not liking multiple disk changes and it popped the red button referred to in the story.
We let a DEC engineer pop it back in!
Don't think I ever read the service guides, I just had the job of updating the entire manual sets for the VAX, VMS and a PDP 11/45 that was never turned on but couldn't be disposed of. An entire bookcase of loose leaf manuals with pages to be removed, replaced and added every month.
I recall the service guide advising that the magnets used in the device (Motor)
The permanent magnet in the positioner certainly was, but you'd usually not have your wristwatch being subjected to that field in normal operation swapping a pack. However, servicing the unit could definitely bring your wrist much closer to the magnet, and your watch would certainly prefer to not be on your wrist at such moments. The spindle motor usually was a simple AC motor, single or three phase, which wouldn't have much of a magnetic field left when powered off.
Removable packs allow having more storage available (though not online) with limited floorspace, a reasonable trade-off in the era of washing machine sized drives; downside is that packs can pick up dirt and dust leading to head crashes. And of course there would be the night shift operator who tries that pack that failed to mount in the first drive, in the second, third and fourth drive before his more experienced fellow op took him offline. Such a night would put a serious dent in the disk vendor's stock of spare read/write heads.
Fixed disk drives really gained in popularity once they shrunk to a size and weight where you could stack them three or four high without repurposing box girder bridges to support the raised floor. Most of those had sealed head-disk assemblies, allowing tighter tolerances and lower head flight height, which again allowed greater data densities. With these the positioner was inside the HDA as well, with maybe a few exceptions. One was the RP07, actually a Sperry Univac beast of 500MB with a Massbus interface tacked on. Swapping the HDA required having a D cell battery at hand, to make the positioner retract to its lockable position, after which you could unbolt the head/disk pack from it. So, basically the worst of both technologies.
He also learned "that when required I can move pretty quickly whilst also providing a running profanity-loaded commentary."
It's worth practicing for such exigencies by developing these skills separately beforehand. I'd like to say that personally I've worked very hard on improving my reaction times and developing my fast-twitch muscle fibres through many years of fencing and playing squash, but in reality I've not got much further than turning the air blue at every opportunity.
Have several (though different things).
- Someone (who shall remain nameless) threw something at a wall and it just so happened to hit the big red power off button on the wall. Cue the clunk as all the power went off. Totally unintended (but an impressive shot none-the-less).
- An engineer who was working on a removable disk drive (ICL, platters almost the size of Morris Minor's wheel hub) wondered why the drive motor had tripped. Pushed the button, the rather large capacitor on the side of the drive blew up. Very impressive mini mushroom cloud too - right under the smoke detector!
- Funniest one was a manager hitting the big red button on the aircon unit (this thing was BIG, at the back of the building with one or two massive fans and it blew air into the machine room). Then he couldn't work out how to re-engage it (there was a lock ring that had to be twisted to release the button).
I miss those days....
Exactly. Just yesterday I managed to remove a stripped internal hex (allen head/key) head screw with the liberal application of foul language and judicious use of tooling outside of it's intended operating mode. How are we going to teach kids anything if we have to start at the very basics like the various uses of the word Belgium? Or the joys of percussive maintenance?
Had the halon alert go off once when in a DC. As I left the room I noticed 2 guys sitting on the floor in that area with laptops and uniforms for that DC still there unconcerned about their impending death.
Gave it a few minutes then went back in to query with them (they were alive). It was a test. Well at least I know I can get out of that room in time now so successful test I suppose?
I used to work in a herbarium (I managed their Sun systems DB server and the Pcs that were used for accessing it) that had 3 storage areas complete with Halon dispensers. Strangely enough this was in the days when scientists would keep working until eventually they became part of the collection. One of the senior scientists (think a gentleman in his early 80s) was working back late when the system had a misfire and set off the halon. He was lucky to get out before suffocating.
Reminds me of going on a university open day in ummmm must have been 1986. I was on the Computing visit, at one point they showed us some sideboard-sized thing and proudly declared that it was their 10MB hard drive, I gob-smacked. We'd just installed a 20MB hard drive on our school server a year earlier, and it was the size of a shoe box - and that was just the case holding the drive inside.
Ah the Good Old Days.
Lighning & thunderstorm crashes = long and unshielded interconnects.
Long would mean dozens to hundreds of feet.
I remember the palaver teleloading a remote VT server in the spares department could be on an old ICL system. One operator at the console, one in the spares department several hundred feet away, phone to ear, one hand on the signal gain knob.
"Nope. Try again!"
"looks like rain..."
Biting the hand that feeds IT © 1998–2022