User was told three times 'Do Not Reboot This PC' – then unplugged it anyway In last week's edition of On Call, The Register's weekly column dedicated to readers' days being damaged by demands to deflect needless disaster, we wondered if the inevitabilities of death and taxes should be joined by meaningless managerial interventions. This week, we have a new candidate to join lists of iconic …
It is not quite the same but in the last few weeks of working for a company I had to go and do some maintenance work on the network for the only office that ran 24x7 (it was an out of hours claims line).
It had been arranged that I would do the work in sections starting at 2300 to avoid disruption.
I got there and prepped the work and then at 2230 all the machines powered up (if not already on) installed patches and shutdown again.!
But this is a 24 hour operation? I got told that they do this every night…. This is because the machines were set to install patches at the same time as the other call centres which shut down at 2200.
So then the question came haven’t you reported this to the service desk to which I was told no it gives us a break…..
On my return to the office I mentioned this to the person in charge of patching to be told that nobody in the 6 months of the site being running had bothered to raise this issue - for all I know it could be the same 5 years later….
Do Not Attempt to Re-boot This System – then rebooted the machine by pulling out its power cord.
To be fair it may not be clear to someone that pulling out power cord would be sort of equivalent to a re-boot.
This is actually a very poor message, especially for people on a spectrum.
From a dictionary: attempt - make an effort to achieve or complete (something difficult).
Is pulling a plug something difficult? does it require an effort worthy of calling it an attempt?
It also doesn't have any information why it shouldn't be rebooted, so the user may not know how important it is.
Final point - if the message looked like an advert, someone might genuinely not see it or think the message is just another scam.
I think this story is set before the mass proliferation of adverts and scams .
I agree its badly written though. "Reboot" is not a term you would count on your users understanding.
"switch off" would be more clear , and indeed is the term used by mswindow's current upgrade routine.
> "switch off" would be more clear
"But I'm not switching it off, don't worry, I'm just trying to get it back to working order 'cause I need to get work done."
"If pulling the plug doesn't work, I'll try kicking it repeatedly (works for my lawn mover), or pouring some cold water into it. "
You're right, the message wasn't explicit enough for IT-illiterate. I would suggest rather using "Do not touch this computer" with skulls and bones in ascii art.
Still not precise - although most people think the keyboard is a computer (especially those over 50), it should probably read something like:
"Do not touch any computer in the room, do not touch or press any buttons on the keyboard, do not touch any mouse until further notice). Do not power off the computer, do not disconnect any cables or devices."
It should also add "Computer is running an update process that cannot be interrupted by you and anyone else under any circumstances."
But ideally workers should schedule a meeting for the time of upgrade so they are nowhere near their computers or even better perform upgrade during the weekend when nobody is in.
"Do not touch any computer in the room, do not touch or press any buttons on the keyboard, do not touch any mouse until further notice). Do not power off the computer, do not disconnect any cables or devices. ... Computer is running an update process that cannot be interrupted by you and anyone else under any circumstances."
To which the response from some people will be the semantic equivalent of "tl;dr" :-(
I say this not to criticize your suggestion (which I suspect was meant in bitter jest anyway), but to point out that there's very likely no one warning that will "take" for everyone. People are just too different.
The best that can be done is to come up with a version that will work for most people, and be prepared in advance to clean up the inevitable damage (as opposed to being taken by surprise by it on the day).
This post has been deleted by its author
Many of us were using computers well before they were conceived.
I implemented a four hour minimum for on-site visits in (roughly) 1990, a couple years after I went solo. Double on weekends/holidays. A few clients balked at the new rate ... I simply told 'em "Don't call me unless you actually need me".
A new issue arose. Convincing 'em to pay 4 hours for a one minute visit. The old TV repairman's maxim applied, "I'm not charging you for thumping your telly with a screwdriver. I'm charging you for knowing where and how hard to thump your telly, and for showing up to do it". The explanation seems to have worked ... although about four years ago a child CEO wondered why I'd need to thump a telly with a screwdriver.
Back in the 1970's, when TVs still used vacuum valves and I was a student earning a few bob working in a TV shop on Saturday's, We (the minions/staff) clubbed together to buy the boss a small rubber mallet. We were fed up with customer's complaining to us that their TV cabinets were scratched on return from repair: the boss would routinely wander through the workshop and give sets a sharp tap with a screwdriver - the rubber mallet did less damage!!
Yeah, I'm 45 and probably the youngest of the non-computer generation.
I didn't touch computers at school really. DIdn't have a specific IT lesson. (Those arrived about 2-3 years later).
I was recently talking to the mother of a guy I went to school with, she was telling me how he'd got a new job, but he hated it as he had to use computers all day, and he'd never had to before. He's my age.
I'll be honest, I was a little shocked, as I assumed they were something everyone these days knew how to use.
But then I work in IT, so probably don't have the widest world view. :)
"... invented by people over fifty"
Yup, as confirmed by a recent visit to https://www.tnmoc.org/ in sunny Milton Keynes, UK where most of the volunteers are ... not young.
Some wonderful reconstructions of a Bombe (I was particularly impressed by the extensive lubrication system), EDSAS, Decatron, etc. And I spotted my slide rule in a display cabinet - guess that makes me a museum piece.
That seems contrived. If I left a sign that read 'do not attempt to switch off lights' it would be perfectly reasonable to expect that no one capable of working in an office, no matter how "on a spectrum" would smash the bulbs to achieve their desired darkness.
Pulling out the power was simply the result of not seeking further information when encountering a problem.
I used to run back office for a now defunct high street white goods retailer. Most of the updates and patched ran overnight, but sometimes we had to manually dial in, which required one of the shop staff to press a few buttons. On a few occasions I ran the update while the staff member rebooted the machine as required. The shop tills were just PCs with the EPOS software running on them, so when I asked the staff person what they could see on the screen towards the end of the patch, expecting something like a C prompt or a yes/no popup message, they said "nothing". I assumed it had al gone titsup and restarted the process (it took around 15 minutes each time), at the same point I got the same answer. I asked if the screen was on. "yes". Is there anything on the screen at all? "yes, there are some numbers and writing but not what I usually see"...
So basically, don't assume anyone to have the slightest clue what "reboot" means, or anything else that you take for granted.
I one had to explain by phone what a slash is...
Early in my career (just after the C-T boundary), I was in need of telling a user in some other part of Europe to type //HC into his terminal and hit Enter. After 15 frustrating minutes of trying to explain this arcane ritual, I sent them a fax.
Judging by the quality of much "journalism", you need to prove you are incapable of finding any punctuation on a keyboard before being allowed to work for any of the media. Look, just hold the shift key down and press 5. Shift key. Yes, the shift key. What's a shift key? Give me strength!
I had somebody try to talk me through a telnet session over the phone, without actually telling me he wanted me to start a telnet session, with an appalling accent, and seemed to only know TAB abbreviations. It came out something like "contabby pee tee twennif". If he'd just *****y well said: connect. to. port. 23. instead of what I suddenly realised what he was saying "con TAB (expands to connect) p TAB (expands to port) twentythree". I don't care what abbreviations *YOU* use in your own time on your own system, when passing on instructions you need to specify the, well, actual instructions.
Relying on command completion is a bad trap as well! Just a little update of the telnet client and "con" get "console" "convert" or something alphabetically before "connect". And "p" can change too.
Replying "it says convert now, is that correct?" and "it says ping now, it that correct?" could have fixed that :D.
The Cretaceous-Tertiary boundary (often spelled K/T boundary) is associated with one of the most investigated mass extinction events, i.e., the end of the dinosaurs. The age of the K/T boundary is currently estimated to be about 66 million years based on absolute dating methods.
That's quite condescending and assuming. It is reasonable to you, because you probably have not met people to whom this wouldn't be reasonable.
I worked with a number of people you would call neurodivergent where you had to put an extra thought into designing a content that they would have to consume and understand.
Ideally you would want to have their input during the process. What does "do not attempt to switch off lights" mean to you? Then iterate until consensus is reached.
Even if you see that no one has switched off lights in the office, you would learn that some of them would also never switched off lights at home and anywhere else etc.
They are perfectly capable to work in the office and they could do work of multiple people at once, but in a limited scope.
Just to balance the scales, I've dealt with a number of neuro-normies who had no idea about technology. To them the internet was the 'e' icon, they refused antivirus products because "it's a machine, it won't catch the flu" and attempts to explain went nowhere. So, I feel this has less to do with a person's internal wiring and more to do with their exposure to technology. After all, it took me fixing a start-up blue screen three sodding times (all praise Hirren's boot CD) before they understood that yanking the plug out of the wall was not an acceptable way to turn the thing off, it's not like a fancy typewriter...! But I did have to make a shortcut to shutdown.exe because they just were not getting the idea of clicking on the thing marked Start in order to Stop.
I feel this has less to do with a person's internal wiring and more to do with their exposure to technology
Many eons ago (subjectively - it was during the y2k kerfuffle) I was Evil Kontract Scum. And the contract I was working at was replacing all the old OS/2 desktops at a large Bristol-based insurance company. The idea was that all the logisitcs would be done beforehand (new machines delivered, still boxed up to site) than we would go in after-hours and unbox the machines, set them to re-image (newer version of OS/2 with all the business apps pre-installed) and then swap out the old units for the new units.
This would take however long it took - we were each paid for 10 hours per night[1], whether it took 5 or 15 hours. Usually, because we'd got it all pretty streamlined ("you two unbox and move the PCs to the right desks, I'll start the imaging, you swap them over post-image and you clean up and move the old ones to the disposal area) it would take us between 4 and 5 hours a night.
Such a shame we were pretty close to some great curry houses :-) I think I ate more curry on that contract than ever before or since. Sadly, after a couple of months they realised that they were going way over budget and started shedding contractors. Sadly, being the last one to join the team, I was the first to leave. Still, it paid off my 'company' motorbike!
One office we arrived. The site team leader signed us in and basically told us that she didn't know why the company was bothering - she'd never used a computer and never would. To which my response was that. if she maintained that attitude, her only future inside jobs would be stacking shelves in the supermarket because pretty much *every* job other than purely manual ones will require the use of computers. And, even with her job, sooner or later, she would also have to be using computers, even if it was only to read reports.
[1] I suspect that none of the permie team wanted to be spending late nights with a bunch of sweaty EKS's and decided that the cost of paying us a flat fee per night was preferrable to paying a permie or two double-time and then having to go through the hassle of signing off timesheets..
I don't think it's condescending to expect someone to interpret a sentence according to the rules of the language it's written in. The contortions we've had to use to explain this situation, arguing over what technically counts as rebooting or whether an attempt is an attempt if you know how to attempt it are ridiculous. Lots of messages are vague, and I don't expect every user will unerringly understand each one if it's terse or uses unfamiliar terms. I've seen enough examples of a clear error message being bypassed not because the user didn't understand it but because the user ignored it, so I'm not very willing to take the blame every time a user causes a problem by not reading by assuming that there was a magic phrasing of the message that would have perfect success if only I asked enough people what this meant to them.
That would be true if the sentence was written in a way that could be universally understood, which is not the case here.
There is a world of difference between "universally understood" and "I understand it." Sadly there are many, many people in tech who will never, ever get that distinction.
It's why so many web sites are so horrible to use, and why the "Help pages" or FAQs are so utterly useless. Unless you're able to put yourself into an end user's skin you can't create useful instructions.
"That would be true if the sentence was written in a way that could be universally understood, which is not the case here."
Isn't it? The person in the article who didn't follow the message didn't express that they thought it said something else, or that their actions weren't contradictory to its instructions. When asked about the message, they understood that they had made a mistake without needing the message clarified. Their problem was not a lack of understanding. It was a lack of following instructions. I've seen lots of vague messages that omitted important details or were phrased in a confusing way. This wasn't one of them.
But I don't know how to make something universally understood, so teach me. What part of this message wasn't universally understandable, what phrasing would be universally understood, and what is the method you can use to guarantee that when writing future messages? My guess is that your answer to the first part will be tenuous as with the arguments over "attempt" and "reboot", your second will be as easy to quibble with as this one, and your third will be unusably vague or difficult. I'm ready to be proved wrong, and I write messages in programs, so it's worth teaching me if you can.
So using your dictionary, if something isn't difficult then I cannot attempt it? What if it would be easy except something went wrong before I completed the action, for example I was going to press the reset button and someone blocked me from doing so? Can that be termed an attempt? I reject this definition. Attempt merely means to take actions you think will lead to a goal, which may be successful or unsuccessful.
"To be fair it may not be clear to someone that pulling out power cord would be sort of equivalent to a re-boot."
I had the same thought at first. But then that "oh!" told me that the user understood perfectly well what "reboot" means -- and also the connection between that concept and a hard power-off.
> It also doesn't have any information why it shouldn't be rebooted, so the user may not know how important it is.
The article did say
> replaced the desktop background with a big sign, white writing on a bright red screen, saying 'Software installation in progress. Do Not Attempt to Re-boot This System'."
So there was some information about the reason it shouldn't be rebooted.
Also doesn't sound like it would have looked like an advert.
The article felt pretty weak to me for other reasons though. Especially when it says "The few machines that hadn't upgraded weren't a worry: someone always turned off their PC by mistake or did something else that needed intervention" but then goes on to explore the details of one case of this happening. Why is it worth doing that, if this was an expected and normal part of the process? Is the leap from "accidentally turning off their machine" to "pulling out the plug to try to get it working again" really so vast that it elevates the story into something worth reading?
If it happens to one person it's no big deal. When you're dealing with even a small to medium sized organisation you'll get multiple people doing this. Anything larger and you need a procedure for dealing with them.
Yes, most people either don't read emails from IT or don't take them in, and a clear message on the screen isn't enough for some too.
you need a procedure for dealing with them
In our place, we send regular updates (which everyone ignores) then the senior management sends out the same updates as part of their regular updates. Then we send the team leads and managers the same updates with specific instructions to pass onto their teams. Then the senior management does the same thing.
We *still* get people who complain that they were not warned. Fortunately, it's now very rare that they are important enough for us to have to grovel to.
Last place I worked at we received a email complaining that the mandatory updates being pushed to the field service techs laptops could cause the very expensive bricking of a Combine Harvester or other large scale construction vehicle when they were connected to the VPN via their phone & also causing excessive data charges, while working on-site in the middle of a field*.
My handle (Icon) came into full force as my reply stated:
They have already deferred this update twice, now it's mandatory.
When the update is applied the laptop displays a countdown warning that they have\had 90 minutes to complete their task before reboot or reboot first.
If they bothered to connect their laptops to the network, while on company premises they wouldn't be receiving mandatory updates in the field using expensive mobile data in the first place.
Having given the above information, further advised that the only person responsible for bricking a combine harvester would be the technician.
*We had one ticket come in complaining he couldn't connect to the internet to download updates for the combine he was working on, with his phone as a access point when he was in a area of no connectivity for his phone. I responded perhaps he should try contacting Rogers & ask if they wouldn't mind improving their coverage to include the farmers field or perhaps try to download the patches before he left for site.
Those technicians are the reason for your paycheck.
Either find a way for them to get the updates on their schedule, or HR will be having a chat.
IT are there to support the company, not run it into the ground.
Field techs have to spend most of their time in fields, not the office.
That may be true, but depending on what the updates were for, that might be the reason for the technicians' paychecks. If those prevented the equipment from breaking or making expensive mistakes, then they're important maintenance and the technician needs to perform important maintenance.
The same arguments are used against IT all the time. Why do you want to spend a bunch of our money on backups and enforce security policies that get in the way? Don't you know we pay you? Yes, but if you get hit by ransomware or a building fire and your data isn't there, you won't be able to pay anybody. If you have technology, it's probably critical to the company making money. If you don't think it is, have a trial all computers and machines connected to them are broken week and see how things go.
Having had to deal with a number of car manufacturer's various OBD systems (Ford ETIS, Mazda whatever it was, Rover's T4, Vauxhall, Suzuki, Fiat, VW/Audi...) they are all pretty horrible, and given that they mostly have the same issues, I'd be surprised if software for $farm_equipment_manufacturer was much different.
I'd have loved it if you could download all new updates/patches to be applied to customer vehicles at the start of the week/day/when they are released, but the systems are designed in such a way that you can't generally do that - you need to have the vehicle connected, send the VIN back to manufacturer, manufacturer will then determine what software they want to push and then allow you to download it to the laptop to push into the vehicle, and then delete it once successfully installed.
I suspect it's a stupid way of making sure people don't install software on the wrong vehicle, but it's been designed by someone who's never been at the end of a dialup modem/GPRS connection instead of only a LAN port away from that update...
I remember working in Japan with my UK issued laptop - in the days of dial up, when you used to have to plug your hotel telephone cable into your lapt op; and use a local free phone number.
At midnight (UK time) (getting up time in Japan) it started to download the weekly system updates; and displayed a message saying "do not disconnect - or power off". I had to go to work, so I had to disconnect and go to work with the customer.
I sent an email to the support people who told me how to disable these updates for a week. I manually did the updates one weekend - and it took over 12 hours down the slow phone line.
The download package got smarter - it detected your connectivity type, and what time zone you were in, and if they were not good - it did not do the updates.
Timezone can go wrong!
When working in Actuarial Ops, I once wrote a package for analysing the runlogs of some fun multi-threaded, multi-server statistical simulation jobs - say you have seven product lines to analyse, so machine A does 3, and machine B does 4 of them.
I was getting some really good data on performance, stuff we could feedback to the model devs to improve things, as well as optimising the loads in our scheduler. Nerd stuff upon nerd stuff, the usual.
However! Some of my logs were coming out wrong. Sometimes they'd report wildly excessive runtimes, occasionally *negative* runtimes!
It turned out that one of the 40-odd servers doing the work was set to the wrong timezone - so depending on how the work was split between the machines, a given sim could start and end in different timezones...
(we obviously had the DC folks fix the locale, but I also realised the timestamps were fully ISO compliant so we could correct for the TZ using that!)
I was involved in a PC roll out.
This would involve changing the login screen which can be a bit confusing, so we did what any reasonable IT person would do...
Thought they were all idiots so we made it as easy as possible for them.
We emailed documents with instructions for the new login prior the the weekend roll out, we chatted with people where possible to informally let them know what was happening and what they needed to do.
We even printed out (with screenshots) a simple guide and placed it neatly on everyone's keyboard so it was impossible to miss. I think it even had something like "READ THIS FIRST " on it.
I guess about 50% of the people removed the annoying bit of paper covering the keyboard and threw it away.
They then saw the screen was different, panicked and called the helpdesk to ask what to do....
And these were allegedly intelligent people with degrees and stuff ( I don't want to say what in y'honour because of reasons)
You're not alone.
Around the turn of the century, I was involved in a system upgrade. Because of the nature of the environment, we had an air gapped network, so it had to be done on premises (though the bulk of the activity was from a local server, but some final tweaking on each machine was needed).
For 3 days before, we had messages on the screen backdrop; we posted notices up everywhere (and I mean everywhere - including above urinals and the back of cubicle doors in the toilets on the basis that everyone must go there sometime during the day); we even [in breach of H&S regs no doubt] had a poster sized message taped to a trolley in the entrance lobby which people had to see and walk around.
Come the day, we STILL had people turning off machines when they went home, thus stopping the upgrade. We also had one individual staying to well past 23:00 who refused to go and have a coffee while his machine was upgraded as "he had important stuff to do" -- in his case we got him to sign a statement that he was offered the upgrade and refused it.
My manager, who had an interesting attitude to authority, sent a message to HR and the directors asking about the recruitment processes as they clearly had employed people who could not read. In the case of the awkward guy, who raised a call saying "his machine would no longer work properly"; it was put at the bottom of the priority pile; his manager came in fuming (demanding an immediate fix and an investigation into the incompetence of IT support) and was then shown the signed statement, re-shown the notices, reminded of the importance of the upgrade and asked whether many hundreds of people should be prevented from working for this one 'special case' -- fuming entrance / sheepish exit. Again my manager had a sarcastic question to add**.
I suspect that there's no way to get round those who WILL not listen (BOFH style cattle-prods connected to chairs are apparently not permitted).
** If I recall; correctly it was along the lines of calendars being used in the days of ancient Sumeria and well established thousands of years ago by the Babylonians, so why is such a shock/surprise when the end of month/quarter comes round? Couldn't some form of planning be done to avoid the last minute rush and panic?
"Come the day, we STILL had people turning off machines when they went home, thus stopping the upgrade."
Don't underestimate the power of muscle memory. Even if someone has read the message, understood in exquisite detail exactly what it means and what the implications of not following it are, and has even reminded themselves just before heading home for the weekend that they really, REALLY, need to leave their PC running, it all pales into nothingness compared to just how strongly your muscle memory will take over control of your arm/hand and reach for the power button without you even realising it's happening. At least, not until a gnat's crotchet *beyond* the point of no return, which is almost always when you'll then have the "oh shit" moment and realise with crystal clarity just what you've done.
We've all done it, and no matter how much we like to kid ourselves that we're better than that, we'll all end up doing it again sooner or later, it's the unstoppable force of human nature at work.
All you can do is try to mitigate against it by trying to disrupt the muscle memory action from completing - e.g in the old days of physical power buttons, you might arrange for covers to be placed over them all so that muscle memory meets immovable obstacle allowing rational thought to kick in again and trigger the "oh shit" moment sooner rather than later, or in the days of software controlled power-downs you might quietly roll out a pre-update that disables the power down process, and then (just to cover the possibility that the user is also a complete muppet) you ALSO stick a cover over the plug so they can't try powering-down from the wall instead once their initial muscle memory action has been thwarted.
In this case, if muscle memory turned the computer off but the user remembered, you could always just turn it back on immediately for the updates to be applied. While there are some people who didn't have the realization that they did something wrong until after they left the building, there are so many examples of people ignoring messages that I wouldn't put it all down to muscle memory and leaky other memory.
If the messages were worded along the lines of "please do not switch off your PC when leaving for the weekend", there's a risk that users might think, having now switched their system off, that the damage has already been done, so what would be the point in switching it back on again.
OTOH, if the messages were more like "please leave your PC switched on when you leave", that'd be telling them in which state you actually wanted their PCs to be left, thus easier for someone to understand what they'd need to do if they had autopiloted their way to turning off the PC first.
More to teh point, it's not so much the cover. It's the difference . The automatic actions, muscle memory and unconscious patterning need to be interrupted. At the end of the day the desk jockey will have their own shutdown routine, just as the computer does. A sequence of events that are just performed one after another..
Put pencil in drawer, set phone to out of office, return mug to kitchen area and leave it next to the dishwasher for the dishwasher fairy to place inside,* go back, check phone is in pocket, have bag/wallet purse, turn off PC, put on coat, head for exit.
It's all one sequence: The end of the day ritual. And the DJs don't want or need to do anything differently. On the contrary, any variation will likely lead to them forgetting their phone or something.
It's a standard behaviour which can frequently be annoying or even dangerous, when someone just follows the usual pattern and forgets that today is different. Usually it's no worse than "Don't forget to collect little Jimmy from granny's on the way home". But it can be.
*Sorry cynicism got the better of me there.
On the contrary, any variation will likely lead to them forgetting their phone or something.
A couple of years ago, my car was being repaired, and I was working away from home, but near a friends, so he gave me a lift home.
Unfortunately, yes, he broke my routine. Hi, ok, I'm here, come on, let's go, I've got the car waiting, some on, hurry up!
So my routine, pack my bags, left pocket wallet, right pocket purse, car.. oh wait, no car, phone in shirt pocket, havHURRYUP, andCOMEON! withCOMEON! lock office.
Two hours later, arrive home. Pat pockets... hold on, house keys... y'know, THE ONES USUALLY ON MY CAR KEYCHAIN WHICH I DID NOT PICK UP BECAUSE I WAS BEING RUSHED OUT OF THE OFFICE AND DID NOT HAVE MY CAR SO DID NOT PICK UP THE KEYS WHICH OF COURSE I ALWAYS HAVE BECAUSE THE ONLY WAY I CAN GET HOME IS WITH THE CAR, SO OF COURSE THE HOUSE KEYS ARE ALWAYS ATTACHED TO THE CAR KEYS SO THAT I DON'T FORGET THE HOUSE KEYS COS IF I FORGOT THE HOUSE KEYS IT WOULD BE IMPOSSIBLE TO GET THE CAR TO MOVE.
Very very luckily, by sheer fluke, I'd just happened to have got a spare copy a few days before and it was still in the bottom of my rucksack as I hadn't got around to putting in my "bits" drawer in the kitchen, but I was staring at my office keys thinking WTF??????
If I am dressed, the few "must have" keys are in my pocket.
House, tack room, feed barn, tractor barn/shop (these three are keyed the same), daily drivers (one bike, one pickup, one car), four gates (keyed the same), PO Box.
7 keys in all. I could get rid of the key to the gates, they are solar and the pads always work ,,, but why. It'll be there if I need it.
"Couldn't some form of planning be done to avoid the last minute rush and panic?"
When I was younger, I used to like to sit in a little cafe at the top of a shopping centre on the 24th of December and watch all the little panicked ants scurrying around, as for some, it seemed that Christmas snuck up on them without warning.
I used to live opposite a supermarket. It was a fascinating watch on Christmas Day. It was always amazing the number of people who would drive into the empty car park, presumably marvel at finding a space, get out, walk to the store, not even notice the lights were off and security shutters down, and only realise something was wrong when they couldn't get through the door!
I swear people just do not look at anything when going through familiar routines.
(I'm not going to mention all the times I nearly pulled up at the office when I was supposed to be going directly to a customer site that morning... I generally realised as I was making the last turn, when I guess my mind also turned to thinking about what I was supposed to be doing that day.)
what about an upgrade that is sent to a list of pilot users, with a warning that "no, it can't be done over VPN", and then on the next day ALL the pilot users from the Network team (the ones that warned the upgrade wouldn't work over VPN) come back with their computers ready to be reinstalled because they tried to perform the upgrade over VPN???
Regarding "read me first"
At a previous job, I routinely ordered custon test fixtures from a supplier. The test hardware had a standardozed set of files that described 95% of what they needed to properly build the fixture. There were typically a few nuances that couldn't be expressed in the default files, or could be unclear or anbiguous, so I always included a "README.TXT" file in the dataset.
After a couple of instances where my instructions were ignored, I started putting the fixture files in a password protected .zip file. I'd upload the .zip file and the readme.txt to their web portal.
The first few times I took that approach, I got a call from the fixture engineer assigned to the project. The call would go like this:
Eng: "hey Anon, you accidentally password protected this .zip file, can you re-send the files?"
Me: "Oh, the XYZ fixture, hey, did you read the Readme file?"
Eng: "uh... yea I looked at it."
Me: "well, the password is in the readme file, so as long as I have you on the phone let's go through the readme together to be sure everything's clear."
I think I turned up the snark level on repeat offenses. I thought about, but never did, add a clause to the readme that the fixture needed to contain a bag of M&Ms with all the brown ones removed.
Still, you don't have to accept those "suggestions". You can create selective ublock rules. You have the Dropper/Pipette tool there, so you can select what you want to block and it will suggest a custom rule for this website-element only.
if you select wrong you can disable whole websites though!
easily passed over or otherwise ignored
The 'block element' feature in Brave does come in useful..
(Most often used to block those utterly irritating auto-play videos that US websites are so enamoured with! Or the irritating survery questions that El Reg has started infesting the bottoms of pages with)
Unfortunately, it depends on the OS. In Windows 3.1 on some very old hardware, "restart" might not clear the memory as it could remain powered on during the restart. Only a cold boot (shutdown, wait 5 seconds, boot) would really properly restart the machine. For several versions of Windows, reboot and shutdown-then-boot were essentially the same. Then came Windows 10, where "shutdown" was really a deep hibernation, and only "reboot" would properly restart the machine!
"NASA built and used a manned space capsule using well tested switches. Luckily all survived.' - Damn nearly didn't.
When Buzz Aldrin and Neil Armstrong Were Nearly Stranded on the Moon. Aldrin saw a broken-off circuit breaker switch lying on the floor of the lunar module and "gulped hard."
https://www.history.com/news/buzz-aldrin-moon-landing-accident#:~:text=Stories-,When%20Buzz%20Aldrin%20and%20Neil%20Armstrong%20Were%20Nearly%20Stranded%20on,module%20and%20%22gulped%20hard.%22&text=NASA-,Aldrin%20saw%20a%20broken%2Doff%20circuit%20breaker%20switch%20lying%20on,module%20and%20%22gulped%20hard.%22
"A common mistake that people make when trying to design something completely foolproof is to underestimate the ingenuity of complete fools."
Well, yes, foolproof doesn't exist. The Universe will beat you at it always, since it already produced a better idiot !
True story for this week:
- company had the bad idea of ensuring conditional access policies for all country's PCs to have access to ressources only if compliant. And yes, for all 300 users ! No rolling update :(
- 2 weeks before, every user on a non-compliant PC received an IMPORTANT email stating they need to issue a ticket to become compliant or be cut off company ressources
We had half a dozen of panicked users following the lock-up day. Yes, they all lied about not receiving any email, suckers ! Full re-install for some of them.
Novell GroupWise would show you the timestamp from when someone had opened an email you'd sent, which is why the highest tiers of our management pushed for (and got) our mail services converted to MS Exchange. Everyone below those tiers wanted to keep GroupWise. The highest tiers wanted freedom from being held accountable.
Rarely. In one job, I wrote software that was run exclusively by other software devs during compilation. This has got to be the group of people who are most used to reading error messages, at the point during their work where they're most likely to expect error messages, with sufficient technical knowledge that they should understand error messages, and with experience of users ignoring error messages. To be fair to them, we had a lot of users most of which never contacted us. Still, we would get emails from time to time asking for help with a specific error message. In general terms, that message read as follows:
[Tool name] has detected [number] errors in your code. Details of each error are printed above this line. Each reported error has an associated documentation link to provide more information about the error. If any of the errors are incorrect consult [link to wiki page about how to ignore, avoid, and/or report them].
Some users managed to read this, go to the wiki page, scroll through the wiki page until they found our contact address at the bottom, and send us an email to ask why our code crashed. We'd helpfully copy the error verbatim from the output and show them that they could either fix or ignore this. No matter how many ways we phrased this concept and how many wiki pages it went on, there was always someone who would send the email.
Yes. The instructions said that the paperwork had to be filled in in a very specific way, using the actual codes that would go into the database fields. User put in roughly what they wanted it to look like instead. I rejected it with a reminder to read the instructions. User re-submitted correctly. A couple weeks later, the same user submitted another one - correctly, first time, having just reread the instructions!
I think it's more the "the user always read his Friday afternoon emails after the incident" line which is ambiguous here, and yes, I also read that wondering about the sudden regendering of the user... "his Friday afternoon emails were always read by the user..." might have been clearer, or "the user always read their..." would have simply allowed it to be interpreted in both ways, equally valid.
Ever since painting the work trays with an arrow and putting a sign on them saying 'load this way round'... then having to retrieve said work tray because the numpty loaded it the wrong way round....
Anyway... on the subject of turning stuff off.... I needed to back up a machine's HDD so plug the network patch lead into the network port... start the d/l on the laptop and hang a sign saying "Do not power down this machine until I give the OK"
Said user did not power that machine down... good for him.... nope he pressed the main power breaker switch for that factory unit and powered everything down.....
I've seen a Centronics printer cable plugged in upside down
and any number of attempts to force a telephone connector into an ethernet port (or vice-versa - I was impressed how an RJ45 connector managed to do so much damage to the motherboard of the multi-function printer when forced into the fax connector).
Moral of the story - just because it looks like it should fit doesn't always make it able to!
(fnaar)
I'm so glad I just do deskside and network support for my close family now. I can tell them they're the issue in clear unambiguous terms if necessary. I had to be so diplomatic in an earlier life...and sometimes my inner Bitch Pudding strained to be released. Maybe that should be Conan The Librarian...
I used to work in a University computer lab. If we reinstalled a machine, the process was largely automated, but the machine would be out of action installing applications for about 3-5 hours (depending on network conditions). We had to be logged in for or old deployment system to work, so we'd leave the machine logged in, locked and a with message on screen explaining that due to software installation, they needed to use a different machine. Yes, I know an out of order sign would have been easier, but, our boss didn't like them, and, TBH, more often than not, the students would rip the notice off, reboot the machine then complain..
You could bet that if you re-installed a machine, even if the rest of the lab was empty, at least one student would come in, reboot the machine, then try and use the machine, often complaining to us that the software they needed wasn't installed, or wasn't working properly (likely due to an install missing bits).
in a lab I worked at for 20years we had a few networked Xerox Phaser hot wax printers that users could print from when they need anything printed in colour. Now the phaser hot wax range were a funny old thing, they produced nice bright colour prints and were quite cheap on the consumable front, you just dropped new colour wax sticks in the top and changed the maintenance kit every now and again. BUT they REALLY didn't like to be powered off, as when they started again it did a lengthy cycle of cleaning etc which used A LOT of wax, and to shut them down properly you have to go in to the shutdown menu on the printers UI and do it from there. So we had a sign on each one saying DO NOT POWER OFF. We also had an annoying boffin who liked to tell you how you should be doing your job and what ways you could do a particular task better, he also had the misfortune of looking like Ned Flannders, didly, didly.
Anyway sat at my desk when I got an snmp alert to say one of the xerox phasers had shutdown, ffs thinks I. It was in a room just across the corridor from my office so I popped over and found "Ned" stood over the printer.
"Have you powered it off?"
"Yes"
"did you not read the sign?"
"Yes but I thought that didn't apply to me"
TW@T
I loved working on those printers. Because the other side of not switching them off ever was that there was always a supply of melted wax inside them, and Xerox training was to power the machine off when you arrived onsite, then wait 40-50 minutes for it to cool and solidify so the wax didn't splash around and ruin the gearbox when you moved the printer from whatever recess it was usually installed in so you could actually fix it...
Given that the four wax blocks were different shapes and sizes, and were numbered 1-4 as were the slots on the printer, it was always surprising how many waxes ended up in the wrong slots!
> it was always surprising how many waxes ended up in the wrong slots!
I'm always puzzled that people are surprised by the things end users do, mostly due to a lack of understanding.
For example many, many years ago when networking was still a novelty to most companies and wide area networks were incredibly slow if they existed at all local file and print servers were the norm. When I first worked in IT it wasn't abnormal for these servers just to be stuck in the corner of an office somewhere. So one thing I got used to seeing was a plug labelled something like "SERVER - DO NOT UNPLUG".
One of my first calls in a new job was from a site complaining that none of the staff in the office could access their files. I tried to access the server over the staggeringly fast WAN connection (64K IIRC) but there was no response. I could see other devices so it looked like a dead server. Since the site concerned was just the other side of the dual carriageway from the office I risked life and limb to walk over there. On arrival I located the back room that housed the server. Sure enough I could see the problem a soon as I entered that office. There were only three desks and some filing cabinets in the office. Two desks were occupied the third, against the back wall, was home to a bulky yellowing laser printer (this was in the days when smoking in offices was still the norm) and under the desk sat a chunky Novell Netware server. A chunky Novell Network without the usual LEDs on the front panel. Next to the printer was a double power socket - both sockets were occupied. One plug was labelled "PRINTER - DO NOT UNPLUG", the other plug was unlabelled and lead to a kettle on top of the adjacent filing cabinet. On top of the desk lay a power cable which led below the desk, on inspection the plug of this cable read "SERVER - DO NOT UNPLUG".
I asked the occupants of the desks if either of them had unplugged the cable to plug in the kettle. No, came the answer, but that kettle is usually in the kitchen. It was there when we came back from lunch. I plugged the server back in and went off to find out who was responsible for the kettle while the server was going through it's file system checks and long boot procedure. It wasn't hard to find the owner of the kettle. There was a door labelled kitchen so I entered. In there was a man with a central heating boiler in pieces. As soon as I entered he said "sorry you can't come in here while I'm working on the boiler" I apologised but asked if he'd moved the kettle. He answered no, that was the receptionist. So on to reception area where I found the woman who had called me in the first place. Had she, I asked, unplugged the server in order to plug in the kettle? Of course she had. When the engineer had arrived to work on the boiler he'd told her nobody would be allowed in the kitchen for an hour, so she'd taken in upon herself to move the kettle in case anybody needed a brew. And rather than just whipping the kettle out of the room she'd also taken it upon herself to find a location and a plug socket for it. When I asked why she had chosen to unplug the server rather than anything else in the building or maybe find a vacant socket he only answer was to say that she couldn't unplug the printer as somebody might need it. When I explained that the printer wouldn't work without the server she was surprised. Fair enough I thought, maybe she doesn't know the printer is connected to the server. But then I asked if people had started complaining about their files not being accessible straight after she unplugged the server I saw a lightbulb come on behind her eyes.
That was an early lesson to me in the fact that end users don't necessarily understand what IT staff understand.
When I got back to the office my boss was kicking off about the receptionist being an idiot. So I asked who's idea it was to put the server in an open office when it would have been better to tuck it away somewhere. Followed by why the power socket had been in such an accessible position. And why there was no UPS. And whether many other sites had the same setup. And probably a few other questions along those lines.
People in IT have a history of overestimating end users' understanding of technology/
Someone in the User Admin department called me to check out a warning on her PC from our corporate anti-virus package. I went to her office, noted the message about an infected file, and placed a big sticky note in the middle of the screen reading, "DO NOT USE THIS COMPUTER". I then went back to my office to remotely access her PC, to copy over the infected file for futher analysis, before the AV package auto-deleted it.
At my PC, I connected remotely and found ... the file was gone. How?
Waste-of-air I.T. person in another sub-department had come in to her office, taken off the sticky note, rebooted her computer, and was sitting behind it, using it when I returned.
Or, if you're writing the malware and I'm not sure why I'm giving you advice, you make your program start at boot and stay silent so someone can use the machine without making it obvious you're infecting it and your attack survives a reboot. Like they've been doing for decades. You cannot sell ignoring a message as a security precaution.
Oh foolish person.
we sellotaped a4 size "upgrade" notes over ALL the keyboards and mice from over 50 PCs once whilst a major upgrade was in progress. on a weekend. We taped the power plugs into the sockets.
Some bright spark found the emergency breaker (on a Saturday when no-one was scheduled in) and turned the power off to the entire building
This post has been deleted by its author
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
