PEBCAK problem transformed young techie into grizzled cynical sysadmin With Friday once more upon us, another week has descended from the energetic optimism of a Monday to the deflated downtime of Friday. To ease that slump The Register as always presents a new instalment of On Call, our weekly reader-contributed column that shares tells of tech support triumphs and terror. This week, meet a …
Never assume anything, not even that you and your interlocutor are talking about the same app.
My mum called the bank immediately after she responded to a fishing email and then realised her mistake. The bank were very understanding and asked her for her account details to lock everything down and change numbers passwords etc. She told them what she’d typed and the woman at the other end apparently sounded puzzled. She asked mum for her details again and then said “not to worry” they weren’t her account details.
Mum now reckons that there are advantages to getting old and slightly losing your marbles after all.
There is a bridge website that hosts individual games for free but charges for entry in competitions. Some bridge players were getting a very poor exchange rate. The solution is to put the correct URL into bookmarks, show users how to get to their bridge site from a bookmark and explain that if they ever use a search engine to find a similar looking site they will pay for their mistake.
I once got a very worried phone call during the day from someone telling me that the computer was asking for a password, so what should she do?
Thinking that something was amiss, I went round to see the user to try to work out what it was. After all, this was the Finance department, so...
I was confronted with the standard Windows 98 network logon box. When I asked the user what had happened beforehand, she mentioned that the computer ran updates and did a restart, but she wasn't sure if she should put her password in. I stayed with her while she put it in and confirmed that, yes, it was just standard updates.
So, don't always assume.
I had a similar one once. A user in the next office over came in and said that she had a strange error on her machine and could I take a look at it. I figured it was only in the next office and would be easier all-round if I just went to take a look.
I can't remember what the exact "error" message was. It wasn't quite as bad as "Do you want to save your changes? YES / NO", but not by much. I took great care in not letting my expression change, then asked the woman the exact question that the dialogue box asked. She said Yes. "Then click Yes." I replied. I nodded, walked out, and facepalmed as soon as I was back at my desk.
For all that her job involved constantly working with a PC, she was terrified of doing anything with it for fear of breaking something.
It is tempting to think that she was stupid because she didn't read the prompt, but plenty of commonly used apps ask thar question if you simply open a file to look at it, and then close it again without making any changes. The program itself might have made changes invisible to the user (e.g. to metadata) but the user isn't to know that.
In those situations it's the designer(s) who are stupid, not the user.
Excel springs to mind...
quote: she was terrified of doing anything with it for fear of breaking something.
You would be amazed at how common that is. Members of my family are terrified that they will lose the contents of their bank account or have their ID pinched if they click on the wrong thing, even in completely benign environments. The news media has a lot to answer for. With the govt. censorship bill coming up, the BBC are running half a dozen scare stories about social media at a time.
"...Members of my family are terrified that they will lose the contents of their bank account or have their ID pinched if they click on the wrong thing,.."
Congratulations, you have done a perfect job as Technical Support (Unpaid Branch). Your reward is getting less "but I only pressed...." conversations.
I had numerous calls/texts from my elderly mother that her bank / pharmacy / email / WiFi were broken.
Troubleshooting was always a bit tricky...
Early May, I had an epiphany and built her a "traffic light" type thingy - a few WS2812 adressable LEDs driven by an ESP chip, built into an Ikea picture frame.
Now, my first question is "which lights are green and which are red?"
No. 1 is WiFi access works, #2 is the router responds, #3 is Google (as a handy stand-in for "the internet"), #4 is our e-mail server, and #5 is the control box for the PV panels on her roof.
Best mother's day present (for me :-) ever!
... that anything they do will break something or delete important data: they simply have no understanding of computer systems.
It's all witchcraft and magic to them, and they have no solid basis from which to logically decide whether the next thing they do with the computer will result in "goodness" or will result in "badness."
For all that her job involved constantly working with a PC, she was terrified of doing anything with it for fear of breaking something.
I spent hree decades of working with/managing highly skilled, specialist teachers and supporting our IT as a side role.And that worry was a common thing pretty much from the 80's until teh 2000s.
I don't think staff got less worried, just theystopped caring.
Better than the calls I got occasionally... "I had an error message, yesterday." "What was it?" "I don't know, and it's all working fine since."
That's about the same as getting a call to the helpdesk saying that their application doesn't work.
- Did you get an error message?
- Yes
- What was it?
- No idea, I just clicked to make it go away. It had a long number in it, does that help?
-- [Helpdesk person rips apart the third stress-relief toy that week]
There was a company that tried a relatively inspired way of dealing with that. Instead of error numbers, they used a list of adjectives and nouns, so the error message would come up with:
- Please report a Fluffy Green Parrot error to the helpdesk.
This had a marginally higher success rate than a hexadecimal string, but the advantages have obviously not been enough for it to become an industry standard.
This post has been deleted by its author
Once in the 90's I found a post-it on my desk: "There was a box with an error message on my computer". I went over.
"What did it say?"
"I can't remember, I just clicked 'ok' and it went away"
"What program where you using, what where you doing at the time?"
"Working on a document, not sure which one."
"Windows XP for Embedded Systems license key and 4GB of RAM (new models usually get 32GB)"
This is a bit suspect to me. I don't think I ever saw any version of XP with 32GB of RAM. 32Gb=4GB maybe, but even that was exotic on the desktop.
Other than that, I think we all saw the problem. If user ID is rejected, despite confirming it, it's probably being checked against a different database.
I don't think XP Embedded was ever released in anything other than a 32bit version, so anything over 4GB of RAM would be pointless.
Besides, why would an ATM require more than 4GB of RAM? I'm old enough to remember numerous ATMs running OS2 quite happily with a fraction of that amount of RAM. I'm assuming a bit of confusion here. Still a decent On Call, this aside.
32MB - that was a large hard disk for a PC, never mind memory or file size!
My first home PC (80286, 1MB memory) came with a 20MB drive but I "acquired" an RLL* controller which pushed it up to around 32MB.
*Run Length Limited - the bit density on the track was by length rather than angle as in MFM drives. ACT had achieved similar with 5 1/4" floppies by varying the rotation speed as the head moved in and out, hence their musical disk drives.
The heaviest thing I've ever lifted was a 10mb hard drive from one junk room to another - that was in '95 , god knows how old the hd was . It was about the size of a 70's home stereo system and the weight of a Morris Minor engine!
(this was a in a college graphics dept , where students waited overnight for things to render)
Twas the Victor 9000 / ACT Sirius 1 which had "singing" 5.25" floppies.
They squeezed 1.2MB onto the same media as "standard" 360K drives by having a lot more sectors on the outer tracks than the inner ones.
This was done by changing the rotation speed by using a stepper motor rather than one with fixed speed. Just use more steps per second as the head moves outwards. The encoding of the sectors remains the same.
RLL changes the way that bits are recorded onto the media; it is the changes in magnetisation which are important, rather than the direction.
Run Length Limited has nothing to do with "angle". It is simply an encoding method to make it take much longer for the magnetic spots to migrate toward each other, and thus increase the usable bit density of the drive. Seagate made the ST-225 which was a 21 MB drive. They also made the ST-238R, which was "certified" for RLL 2,7 use, which gave it a capacity of 32.7 MB.
You could use an RLL controller with an uncertified-for-RLL-use ST-225 drive, or other company's similar drive, but the reliability of such use would be ... variable.
Couldn't remember off the top of my head, but a quick check confirms my suspicion that while the minimum RAM for regular XP was implausibly low but not quite *that* low- 64MB apparently- I'd have expected the embedded version to work with less.
And indeed, the minimum RAM for Windows XP Embedded *is* apparently 32MB.
(Side note; though I notice "Windows XP Embedded" apparently isn't the same thing as "Windows XP for Embedded Systems" or any of the other embedded versions of XP?
Then again, MS wouldn't be MS if their marketing department ever stopped giving confusingly similar names to different things. Or, for that matter, 27 confusingly different names for the same thing.)
And doubly especially so if you were around when kilobytes was the way in which system memory was measured, and megabytes were just a theoretically obvious next step in the naming scheme, but not something you really expected to be seeing any time soon sat on your desktop...
As far as 32MB seeming like a lot goes, I always remember the feeling of complete and utter awe I had whilst doing some postgrad work at uni (where all the lab PCs were specced with 8MB, and at the time my home PC was really rocking it with 16MB) when one day I happened to be talking to one of the other postgrads as his PC started up, idly watching the POST messages appearing, and feeling my jaw drop as I realised the memory total being displayed wasn't just the usual "8192KB OK" (or thereabouts) I was used to seeing on the other machines, that my eyes weren't deceiving me, and it really was saying "81920KB OK" instead...
These days I still find myself feeling both slightly awed and bemused in equal measure at the thoughts that my moderately well-specced laptop a) has more dedicated memory for its discrete GPU than most PCs would have had in total up until fairly recently, and b) has enough CPU cache to make even a typical Win98-era PC jealous. And yet with all this power at our fingertips, we STILL have to sit and wait for crappily written software like File Explorer to do stuff that really ought to be achievable in the blink of an eye. Sigh.
"has enough CPU cache to make even a typical Win98-era PC jealous."
I remember being awe of CPUs having 16KB of on-chip cache because that was the entire RAM size of my first computer. And the days you could buy secondary SIMM-like plug-in secondary cache modules, often a whopping great 128KB or even 256KB if you were rich!
"This is a common mistake in remembering/anonymising old system specs, I think."
Or quoting current ones. Just the other day I offered a friend "a spare 2 gigabyte drive" I have available. I meant 2 TB of course, but "gigabyte" is what came out of my mouth.
"Especially if you were around when 32 _megabytes_ was a lot of RAM!"
Just so.
My first computer had 2 *Kilo*bytes.
Designed from schematics in a Zilog Z80 data book and hand-threaded and soldered. Surprisingly, it worked. Programmed in hand-assembled machine language (Z80 was easy, very orthogonal instruction set).
As a pupil in the late seventies, I wanted a computer but had more time than money. Learned a lot from building that thing.
I was thinking to myself "4Gb is overkill for an ATM anyway, all they need to do is display a GUI and contact the bank's database, it's not like they need to run a web browser".
Then I realised, I bet a lot of ATMs do have to run a web browser, to display adverts etc.
I'm sure that's fine, with no possible security implications whatsoever...
Back in the 1978 the Open University had two computer sites, one in Milton Keynes and one in NE England (Sunderland?). These two sites did not share all the data as I discovered when sending in an assignment consisting of a handwritten bit of code that relied upon data saved by the previous assignment. On several occasions my assignment failed because the data was stored at the other OU site.
Reason for handwritten bit of code? I was living in Brora, Sutherland and nearest site with access to the OU system was over an hour's drive away (and access, via a 75 baud T100 teleprinter and a telephone-based modem, was decidedly flakey at the best of times)
What was the course? My first ever OU course, PM978 "Computers and Computing"
Did I pass? Yes, grade 2
PEBKAC
Clunkiest stupidest acronym ever , no wonder the author got it wrong in the article title .
Especially when there's that other one that means exactly the same thing but is far more fluid and rolls off the tongue easier
It escapes me right now ...
{edit}
Seems PEBCAK is also valid , but equally clunky imho when there is there much more palatable PICNIC - "problem in chair not in computer"
Customer: My internet is broken!
Me: Mine is working fine.
A colleague was in the room and told me after the call "Maybe you should not be saying this". Seconds later he says "Heck, just keep doing it." I did, but only with a very small number of high maintenance customers that were particularly annoying.
"I stuck my neck out to help this guy, and then that happened. If that won't turn me into the grizzled, gruff, and uncharitable sysadmin of yore, I don't know what will."
Embracing the PICNIC-first mindset has meant that I am that grizzled, gruff, uncharitable person who deals with people's problems. So much so now I'm told I'm too miserable, it's something I should change. Be the "friendly face of IT".
Change it I would, but my PICNIC brethren need to stop being so fucking stupid first before I become happy. While everyone has the capacity for change, there are some aspects of a personality that can't be. And for as long as being the PICNIC person in a company, I can't be happy.
> Be the "friendly face of IT".
Trust me, this _is_ the friendly face of IT.
Like young Wednesday, you don't want to see me when I'm smiling, as I enforce the company policy on your server directory.
You may even hear me give a friendly chuckle - although I am under strict instruction to hold that in until I'm alone in the server room, something to do with the PFY and PTSD.
This is where you need to practice your "Hello, I'm not a serial killer, really!" manic grin.
Trust me a few days of that, and your boss will be begging for your grizzled normal face again...
(occasional manic laughter, muttering, and random stabbing motions, can also help... :P)
Rule one: never believe the client, always verify that what they think is true is actually true.
Years ago, when broadband was a newfangled thing, I was summoned to help set up an NTL cable connection that just wouldn't play. The guy had been trying for days, but nothing.
I went round, redid the setup, nothing. Confirmed with him that he was using a crossover cable, he assured me that he was. Called NTL, spent ages on the phone with them who were equally mystified.
Eventually decided to check the cable myself, despite his assurances. It wasn't a crossover. A quick cable swap later, and all was good. If I hadn't believed the bugger and had checked it myself, I could have gone home hours earlier.
I had exactly the same when the company I worked for moved over to Cat-5 cabling instead of thin-net. I wasn't directly IT by then as I'd moved onto a project but they couldn't get a segment to communicate. They'd checked and rechecked the wiring, still nothing, so I offered to have a look.
Quick glance at the hubs (yes, hubs not switches, it was a long time ago), no lights on the connecting ports of the central hub and the segment hub. Pull the connecting cable, get out my trusty cable tester, and yes, it's a straight through cable not the necessary, in those days, crossover. It turned out the network provider had shipped a box full of identical cables with no indication of which were straight through and which weren't. I found the right bit of string, duly marked each end with an X and plugged it in.
I knew the writing was on the wall at that company as I was then criticised by senior management for getting involved!
We've had a fun one as the company had a phone exchange as well as network cable sockets. All were RJ45, not all were marked accordingly. The joy of badly documented legacies..
For some reason, some phone links actually worked as network connections when patched in while others didn't, I suspect the pairs were connected in a different way. Only after we got a decent tester that could actually assess possible throughput could we identify the sockets absolutely not suitable and mark them accordingly (in this case we used a Fluke CableIQ, but there are a few on the market).
I'm sure we'll eventually get round to checking what's up with those sockets, but presently the problem is solved enough that we can leave it on the backburner until we acquire a spare PFY - someone on the team loses a bet :).
Assuming the PBX isn't doing something funky with ports to switch them between phone and network depending on what it detects being connected, and given your comment about legacy stuff which suggests this might be an older setup you've taken over, then it may be as simple as the PBX tapping off the pair connected to pins 4-5 on the sockets, and using that as a "free" way to also run basic analogue phone lines out around the building at the expense of rendering those network cables incapable of handling gigabit ethernet (which, for a sufficiently old setup wouldn't have been an issue). Or even if there is a seperate phone line running out from the PBX to each socket, you could still then wire the socket to give pins 4-5 to the phone line and the other pins to the network cable, which would have the same end result.
In this context I find the trend that some have removed link lights from RJ45 sockets stupid, irritating and shortsighted beyond belief.
The 0.03 in any currency that that saves on the price of a network connection is recovered the first time someone glances at them to see if there is even a network connection - no lights means no link, which cuts down on the diagnostics to get things to work again.
And yes, we still use a lot of cable, because when it works it works reasonable well. With WiFi APs you have to make sure they're MU-MOMI compliant (and it should be enabled) and then you still have to keep an eye on AP saturation when users congregate in a large meeting room who all want their own f*cking Teams session because they were sipping martinis when the brains were handed out.
When they started complaining we dealt with that by generating 'hardware problems" at the AP (read: we clamped the router port so it limited traffic to 10MB), and then gave a few cables from the floor ports, casually observing that they could always hook up one laptop to the big screen up front with the expensive camera (that was bought for that explicit purpose and was acquired at great expense after they whinged about not having it).
Over the years it has become evident to me that some users should be very lucky that their bodies have an autonomous nerve system so at least it keeps functioning independent from the presence of any intelligent thought processes. Personally I think that design was a mistake as it kept a few floating in the gene pool that ought to have sunk like a stone..
Anyway, beer. I wish you all a recuperative weekend.
If any users go missing, that has nothing to do with my interest in cleaning and repairing the large paper shredders on Monday, nothing at all.
they were sipping martinis when the brains were handed out.
Over the years it has become evident to me that some users should be very lucky that their bodies have an autonomous nerve system so at least it keeps functioning independent from the presence of any intelligent thought processes. Personally I think that design was a mistake as it kept a few floating in the gene pool that ought to have sunk like a stone..
Literally today:
"we can't look up a new product on *these* machines, but it works on *this one*"
I go and check, and sure enough, I can't find the product in the data file, then I think to check the 'working' one and get really confused, it's not showing up there either. Turns out the report of "it works on this one" was erroneous, and whoever was supposed to add the product had just fucked it up universally.
Simple fix for me: "go find who was supposed to add this, and tell them to do their fucking job right"
You know what, I honestly believe that part of the problem is that help desk companies and others do tend to employee people based on their ‘technical skills’ (well sometimes just the ability to cram enough information into their short term memory to scrape through a Microsoft / Cisco exam, which as we all know, equips you with maybe 30% of the skills and knowledge you need to provide proper support). But I digress!
What is missed is ‘people skills’, the ability to put yourself into the end users shoes, try to imagine what they are seeing, and think ‘hang on here, old Mrs Applegate is claiming xyz, but that doesn’t make sense, so what else might be going on, however outlandish it might seem’?
And that is why I refuse point blank to do telephone support, because although I understand the concept, I appreciate that I’m not good at people skills, as a failed marriage would testify. But stick me in front of a machine or recalcitrant server / router / switch and let me see for myself what it is doing and it’s going to get fixed!
"And you may look weird talking to yourself.."
Back in the day, when you saw someone talking to themselves walking down the street, it was the local nutter[*] and you crossed the road or otherwise avoided eye contact at all costs!
These days, half the people in the street are chatting away to their invisible friends and the ear buds aren't always obvious.
[*] Hey it was a different world back then, different words, different attitudes.
Heres a quick fix for you :
If you get a call from the helpdesk , who are not waiting for you to read the ticket , or from the PA of someone high up .....
Its really urgent!!! The top brass are having a big pow wow in the executive meeting room and everything just went dead , suddenly , in the middle of the meeting.... immediate assistance needed!
.
.
.
.
then just calmly tell them to flick the switch on the power socket that the laptop is plugged into.
every. single. time . smh.
That's why we don't have them in Europe anymore.
Also has the benefit that it occasionally lowers user count, but the carbonised remains sometimes take a while to remove.
Joking aside, the thing I like best about UK plugs is that they have fuses rated to the cable attached. Well, theoretically anyway :).
I have been struggling to get my solar installation sorted for over 5 months now (Senergy Direct).
At the last visit by their experts, I insisted that we started from basics and asked them to double-check that the load-balancers that they were struggling to connect to were actually fitted.
Once you knew what you were looking for, you could see from ground level that they had not been fitted!
My original take was also "XP at the time of mobile apps???". Turns out Win XP SP3 (which was the only just-about-OK version) is from about 2008, iPhone from 2007, Android end of 2008. It's absolutely not far-fetched to think of an XP ATM in service for a few years, by which time I'm sure there would have been some banking apps around.
Not sure where you got the 3GB figure
3GB was the functional limit of what XP could address. Yeah, there were some long forgotten tweaks to make it use the full 4GB.
Of course this back in the day when a 512MB workstation was for power users. A full 4GB system would fill all the RAM slots, and that was IF the motherboard supported expensive 1GB sticks of RAM.
I do miss the days of roll-your-own computing. Built-in CPU, RAM, storage, and SOC have pretty much taken most of that away. It was fun 'reclaiming' components out of unused hardware to max out the capabilities of my production workstation. I knew exactly what top end CPUs & RAM sticks my motherboard supported, and kept a stealthy eye out for them to appear in the workplace.
There was no sense giving high-end hardware to a pointy haired boss who didn't understand the correlation between the roller-ball mouse and the cursor on the screen.
The 32bit version of XP supported 4GB.
An individual application would only get access to 3GB of that - unless compiled with a Large Address Aware linker switch to indicate that they could cope with the full address space.
That limitation and switch still exists in Windows 11, as it's part of the 32bit Win32 ABI.
There was also a 64bit version of XP, though I never saw that in the wild.
> An individual application would only get access to 3GB of that - unless compiled with a Large Address Aware linker switch to indicate that they could cope with the full address space.
Correction: 2GB unless compiled with the large address aware flag, which got you as much as the OS would provide to a 32bit process. Which was 3GB on 32but OSs. When running on a 64bit OS an executable compiled with the large address aware flag will get 4GB.
The difference is because the top half (normally) or top 1GB of the process address space is overlaid with the kernel's address space on 32bit versions of Windows. 64bit Windows (which became generally available with Server 2003: XP 64bit being a variation of 2003 and causing confusion with service pack numbers following 2003 rather than XP 32bit).
Ie. to maximise the available address space for 32bit processes you need a 64bit OS, rather than more RAM.
Addt6ional (that I remembered later: it has been a while since this was operationally significant).
On 32bit Windows the large address aware flag was not sufficient to get the additional user mode address space (it was sufficient on 64bit windows). 32bit Windows also required a boot.ini flag to tell the kernel to use only 1GB of address space.
This restriction on the kernel could cause problems itself (including performance) so should be thoroughly tested first.
If you think that something is so mindblowingly stupid to you.Go check it.
I've found terminators unplugged (SCSI and 10base2), keyboard and mouse swapped (PS/2), the wrong monitor cable plugged in (meaning the other side of the cubicle got their screen instead of their own).
Cross-over serial/ethernet being used by someone attempting their own diagnostics.
IEC13 to IEC14 plugged in a loop was a surprise, but the user was just "tidying up" and requested we get an electrician to check first.
Fan heater plugged into a UPS socket.
Just when you think you're a God and everyone else are idiots, check your own assumptions. The good old "Of course it's turned off" at the consumer unit/panel. That the computer is really turned off and unplugged. That the terminal you've got open is the one connected to the server your shutting down/rebooting. The drive your formatting is the one you want to format.
Ingenuity is very short on supply. Ignorance/Idiocy is boundless.
In this context I'm actually quite grateful for USB-C creeping in.
The cable self protects in that it will stop any idiot hauling more power through it than it can handle, and it ends the bewildering mix of connectors that have always provided evidence that some users didn't pay attention when they were very small kids and had to put shapes in the right holes.
I'm certain that some will still be able to mix up the totally different HDMI, USB-C and RJ45 jacks, but at least insertion in the wrong socket will now require so much force that it gives me the ability to take away the device from the user in question and deplete their department's budget with a fat repair bill or have said user sponsor our beer fund so they're not found out..
"but at least insertion in the wrong socket will now require so much force"
Not until USB-C takes over entirely from USB-A. My current laptop has its thunderbolt socket placed right next to a USB-A socket, and I can tell you from frequent first-hand experience that it is TRIVIALLY easy to insert a thunderbolt/USB-C plug into a USB-A socket without needing to apply any greater level of force than you're expecting to be applying to put it in the thunderbolt socket next door to it...
Fortunately the thunderbolt plug on my docking station has an embedded LED that lets me know when power is flowing, which gives me an instant visual feedback that I've managed yet again to plug it in the wrong hole the first time. Doesn't help that a USB-C/thunderbolt socket is so small that it's a) hard to see unless you're working in good lighting and are looking pretty much directly at the socket aperture (as opposed to at the oblique angles typical when viewing sockets along either side of a laptop), and b) is also rather more difficult to find just by feeling your way along the side of the laptop for openings.
I put lots of white labels to write on next to the keyboard on most of the laptops I've owned, to show which port is where around the sides. But there is a bit of a problem of perspective. The label is vertically above the port, but you probably aren't looking at it from directly above it.
"the bewildering mix of connectors that have always provided evidence that some users didn't pay attention when they were very small kids and had to put shapes in the right holes."
Oh, what a brilliant observation and association! I'll have to remember that one the next time some user wants something plugged in for them "because we've not been trained to do that" :-)
It's not always the clueless users that are hard work. Dealing with people that do genuinely know their shit, just not necessarily the stuff we have the problem with right now can be tough. Mainly 'cos they tend to interpret what's going on and their responses are coloured by that. Genuinely trying to be helpful, most of the time, not just dick-waving "I am the supreme being in the universe" showing off. Something I always try to remember when escalating calls to vendors, if not always successfully :-)
Most of the competent people I deal with/ have dealt with are more than happy if I say something along the lines of "I know you know your shit[0], but just take your brain out and put it somewhere safe for a bit, I just need you to be my ears, eyes and hands so I can figure exactly what's going on".
RTS
[0] Grammar; the difference between knowing your shit and knowing you're shit.
I'll just keep it short. VP of sales calls saying that his emails are not sending out. I ask for him to check at the bottom of the Outlook screen if it shows Offline or Online. He responds stating that is says DELL.
Finally discovered that he was sending out multiple emails with 32MB attachments in each email. That's the reason it wasn't going out, too many large emails.
The new company logo looks a bit like a few tetris blocks.
Email etiquette says to include the logo at the bottom of your external emails adding a few kb to the size, but most people add them to ALL emails... then someone noticed there was an animated version (a few 100kb) and used that... now someone in the logistics dept has added another animation to that with a bike, car, lorry and ship wizzing along the bottom of the email!
(I can't find a copy to see just how big his emails turn out)
"Horace Goes Skiing" had the animated vehicles in a very few kilobytes. It was a computer game which included a "Frogger" bit to cross a busy road to get to the skiing area, then a skiing game. Then to cross the road again, I think.
I suppose these days you can get it in Java and attach that to e-mails for people to play the game when they get it.
Being the technical one in the family I was often tapped to provide free ad-hoc tech support.
One time I get a call from a family member because the printer isn't working. I go through all the usual troubleshooting steps like making sure they're selecting the right printer in the print dialog box, updating the print driver, rebooting the computer... nothing is working. Grasping at straws I finally ask if the printer is turned on at all. It wasn't. Several minutes of troubleshooting later, it turns out that for... reasons, the person had turned off the power switch it was connected to instead of just turning off the printer.
Ever since then I've been fond of reminding people that there's a reason "is it plugged in/turned on" are still staples of the troubleshooting process.
I also had my own "I'm an idiot" experience not too long ago. Setting up a WFH environment with a new laptop and the VPN won't connect. Turned out I forgot to connect to my home wifi. I did, however, manage to figure it out on my own after a couple minutes. Not before sending a message via my phone to a coworker, but before they had a chance to respond.
Last week my wife complained that she had no internet after rebooting her home laptop. I checked, the laptop had no connection. I checked the server from my hardwired desktop, and could see no DHCP requests from the laptop.
After much wasted time checking laptop settings I logged into the router, and found that the WiFi was disabled. Very odd. I enabled it, and all was well.
Further checking in the router log showed that the WiFi network had been disabled as a result of someone pushing the WiFi button an hour previously. The existence of this button was news to me, but after some checking I discovered that the WiFi LED on the router was also a button, helpfully placed on the front of the router. Just where I'd bumped it when putting the vacuum cleaner away an hour before...
Argh. I got called out to my mother-in-law just this afternoon. "I've got no sound! "
She uses headphones on a laptop, almost exclusively did watching Netflix. I can't tell you how many ways she's broken things over the years. Anyway, I unplug the headphones so I can hear what's going on.. Sounds burst forth from the laptop speakers.. So I try the headphones on my phone. No sound. Then I look at them...
"Have you got the new ones you bought because these ones didn't work the last time you called me round?"
(she did, and surprisingly, using those solved the problem..)
The customer may always be right, but they rarely ask the right questions. Amongst others this week...
"Have you got any spare keyboards? Mine's not working" transpired to be "I've left the dongle for my wireless keyboard at home"
"Can you update a graphic - we've got a newsletter going out today with a typo" was really "the person who made the graphic has gone home, and the source files are on their laptop"
And - as this one - external customer chasing up an order for another company.
Always respond with a question. You might come across as a dick, but you'll hopefully get to what they're not telling you, and hopefully solve the problem rather than patching it.
Also, "energetic optimism of a Monday to the deflated downtime of Friday".
Is that a thing?
My company bought a truckload of the one-above-cheapest cheapo Logitech mouse/keyboard sets, but they have one massive problem: you can't reprogram the dongle like you can with the unifying ones. If we ever develop FTL travel, it will be based on the physics of the speed by which these dongles, mice and keyboards get separated from the trio they were together coded for.
However, we have developed a solution. All the sets we have in stock have now been marked with a dot or line pattern that only repeats every 10 units or so. We used UV markers because we have found that some users appear to have a near uncontrollable urge to mess with identifiers - what they don't see they can't mess with. Oh, and the dot/line idea is because there's very little writing space on a dongle, especially since the only decent UV markers are the thick Edding 8280 ones.. Anything that comes in as a return because of people leaving or moving gets tested if it's still a set, and then gets marked retrospectively as well.
So far, so good..
I once had a call with a client who was unable to log into their Citrix gateway. After much going round in circles, I eventually figured out that they had gone to sub.domain.com instead of sub.domain.com.au.
The American company using this domain also had a Citrix login on the same subdomain.
Kind of in reverse. Some years back I went to a site BigCompanyName.com
And was surprised to see that it was USA only. Not even links to other country's equivalent sites. I queried this with them. "Why does your .com go to the USA only site? .com isn't a USA only domain"
They said "Yes it is"!
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
