A phoney story?
Support tech caught by 'Technician Aura': the bug that only hides when you're watching
ON CALL Life is filled with random events, but The Register tries to make readers’ lives just a little more predictable by always using Friday morning to bring you a new instalment of On Call – the reader-contributed column that shares your tech support stories. This week, meet a reader we’ll Regomize as “Benny” who in the 1990s …
COMMENTS
-
Friday 17th April 2026 07:48 GMT Anonymous Coward
I've seen this a few times - the old phones used to cause enough odd problems that they were banned from being in machine rooms. We had to leave them outside and rely on a landline.
I've also seen the opposite effect - everything seeming to be working perfectly until you go to demonstrate it to someone only for it to fail in a way you've never seen before.
-
-
Friday 17th April 2026 08:15 GMT Korev
My first job in IT was in QA. The second product I worked on was the project from hell, but sales kept on pushing the unfinished and very unstable software. As a tester I knew how to make it crash very well, so sales had me design a demo that would actually complete that they could show to customers...
-
-
Friday 17th April 2026 09:17 GMT Mast1
"....could show to customers...."
"....could show to customers...."
As a vacation student I was testing a complex semiconductor being driven by standard 74-series TTL. The resulting drive waveforms did not look very square due to the complex load on the electrode. I built a lumped component circuit as a dummy load out of resistors and capacitors for comparison (before SPICE etc was available). The waveforms looked a lot cleaner. My manager suggested that these latter traces should be saved for the data sheet....... I was not around long enough to know if that was ever the case.
-
Friday 17th April 2026 10:43 GMT Andy Taylor
This was famously how the first iPhone demo made sure the handset wouldn't crash. That and swapping between phones during the demo.
The iPhone was announced months before release, which was very unusual for Apple. Normally it's "here's our new product, available today/tomorrow/2weeks from now"*
* not counting the vapourware charging mat and Apple Intelligence for which we are still waiting.
-
Friday 17th April 2026 21:33 GMT DS999
Wasn't part of the reason for the delay between the announcement and the release of iPhone 1.0 due to the required FCC testing? Jobs knew there was no way that could happen without some sort of leak. He wanted to show the world when people really had no idea what Apple's phone would be, some were speculating it would have an iPod clickwheel on the front and other crazy schemes.
-
-
-
Sunday 19th April 2026 15:53 GMT 4mc
Cathcarts rule for demos
I did hundreds of trade shows and live "show and tells" at conferences in the 80's and 90's
This lead to the truisism - don't show anything you have not rehearsed.
Most often remembered at the end of a demo where everything went perfectly, you relaxed and make some joke about what you just done, clicked or did something at which point the UI or system would crash...
-
-
-
-
-
-
Wednesday 22nd April 2026 12:45 GMT Jamie Jones
My 5G phone drops down to *2G* for voice because the provider won't support the perfectly valid 4G or 5G voip profile my phone uses. 3G doesn't exist in the UK anymore.
It's annoying, but for £2.90 a month with free calls and texts, i've decided to stick with it. My mates iphone with the same provider works perfectly with 5g
-
-
-
Friday 17th April 2026 12:39 GMT Mishak
EMC
Generally the case when mobiles were new and existing electronics had often not thought about EMC.
I spent a lot of time around then helping people get their hardware to work when mobiles were near by (including fork lift trucks that would lose drive power).
Anyone remember the "dit, dit-dit-dit, dit-dit, burrrrr..." on the TV news when someone in the studio had a phone in their pocket that received a call? ('dits' were the handset and cell establishing the connection; the 'burrrrr...' was the call).
-
-
Friday 17th April 2026 08:35 GMT GlenP
We had a similar one way back on a small warehouse unit setup using ACT Apricot's networking. Everyone was convinced it was a power issue so Apricot provided a mains power monitor on loan. Inevitably for the month it was in place there were no network issues, once it had gone back they resumed. The problem was never actually fixed, it just settled down to a manageable level, but there was a slight suspicion that the power monitor had provided some degree of conditioning.
These days we'd slap a couple of UPSs in on the computers but that wasn't a viable option in the late 80's for a small customer.
-
-
Sunday 19th April 2026 07:25 GMT An_Old_Dog
Running the Numbers
When dealing with those sorts of issues, with those sorts of people, it's useful to calculate how much money the organisation is losing every hour which the relevant system(s) are down or inaccessible. How many hundreds, or thousands, or hundred-thousands, or millions of pounds is that?
Sell your proposal (UPSes, line conditioners, redundant servers, etc.) to management as relatively-cheap insurance.
Beancounters and boards of directors understand that sort of logic.
-
-
-
Friday 17th April 2026 09:07 GMT Lazlo Woodbine
I can confirm this is an issue.
I've told this story a couple of times, so I'll be brief.
Back in the late 80's / early 90's I worked for a large retailer, our computer system ran on a pair of mirrored Compaq servers.
Every so often one or other of the servers would have a momentary brain fart, causing the tills to glitch, the main problem being that any sale in progress would terminate, but the stock issued to that sale would disappear from the system into a pending file and be restored overnight when pending sales were cancelled. This was a big problem if the customer had ordered the last of an item, as we would then be unable to sell that item until it was cleared the next day.
The support guys came and did their stuff; running power analysers, network sniffers, all kinds of tests, but the problem never arose when they where on site.
Until one day the manager's analogue cordless phone rang as she passed the comms room.
Clearly the EMP as her phone connected to the base station was enough to unsettle the fragile servers.
Swapping to a DECT phone was enough to fix the issue.
-
-
Friday 17th April 2026 10:17 GMT GlenP
Nokia phones always seemed to be the worst for that, but that may just have been how ubiquitous they were!
These days it's a slightly different problem. We had a visiting minister at Church who brought his Bluetooth soundbar to play the hymn music from his phone (we don't have a keyboard player any more). He was just starting the service when a call came in, which of course rang on the soundbar (and yes, he should have turned the phone to silent). It's one reason I'm keeping our own setup, which needs renewing, mostly analogue.
-
Friday 17th April 2026 10:44 GMT mickaroo
I used to work in big pharma, doing installation/operational qualification of production equipment. One of the operational test was ‘RF interference’ from walkie-talkies (cell phones were verböten in Production).
The vast majority of the equipment would be just fine, but I’ve seen a few cases where the equipment would have a brain-fart and fall over. Then we’d have to test how close you could be, and flag the equipment with ‘no walkie-talkies’ within X-feet.
-
-
Friday 17th April 2026 14:21 GMT Andy A
I had to mend a phone which could go BANG!
One place I worked had a phone inside an explosives area. When it failed I had a right old time sorting it out. Decode the frame and disconnect line at the exchange end having checked I had the correct extension. Head into explosives area, carrying relevant (manual) screwdrivers. Disconnect phone from wall and line. Carry it outside. Repair it. Head in and reinstall. Back out and reconnect at exchange. Ring the extension. Nobody answers. Back in and find the fault also involved the red light on top of the little cabin. Repeat to figure out what type of bulb the red light needed....
Each trip to the user's area took about half an hour. Fixing the fault took a whole day.
-
Friday 17th April 2026 18:22 GMT The Oncoming Scorn
Ahhhh yes the no phones, MP3 players etc, not only in production area's but also while walking around the GSK Dartford site.
I used to leave my un-powered phone on my desk while I supported my building{s) & I worked away from home.
Unlike North American phones that report a un-powered device is not available at this time, this led to the following evening exchange with the Ex Mrs Scorn.
Hi how was your day?
Ohhhh decided you want to talk to me now have you?
Pardon?
Where were you?
I was at work
Don't lie I rang you & you didn't answer.
(This goes back n forth a few more times until)
Look & listen, I cant use my phone anywhere except at my desk, it was powered off while I was away from desk.
Don't lie it rang I heard it.
I said OK lets stop this shit now, your on your mobile yes?
Yes
Alright now Hang up, turn off your phone & ring it from the main house phone, then call me back with an apology.
It rang, you were right (Still annoyed because shes worked herself up into a right snit).
A brief conversation happened, it was around that time, I decided I was going to walk to my local bar (Icon) terminating that evenings daily chat.
-
-
Sunday 19th April 2026 06:40 GMT Richard 12
Walkie talkies will even trip most RCD/GFD and AFCI/AFDD circuit breakers if "clicked" within a few cm of them. In some cases they'll even trip MCBs.
They have an absolutely massive output, as that initial burst is basically unregulated.
Never use your radio near electronics.
It's one reason the emergency services have been changing over to "mobile phone" technology. The other being encryption, of course. "Police scanners" don't pick up LTE based trunk calls.
-
-
-
Saturday 18th April 2026 19:59 GMT dmesg
A few Sundays ago at our UU service the minister was leading a guided meditation during which he said something to the effect of "receiving a call." At the very moment someone's phone rang. But for it being a quiet moment everyone would have laughed, so the congregation (and minister) all had a good smile.
-
-
Friday 17th April 2026 16:23 GMT doublelayer
"It's one reason I'm keeping our own setup, which needs renewing, mostly analogue."
I don't think Bluetooth is the problem here. Even if everything was analog audio, if it's being driven from an active phone, it'll still ring for an incoming call. Most phones on silent will still play the ring sound through any audio, cabled or wireless, because the software assumes you're not within vibration range and want to be aware of the incoming call, but even if you disable that, they tend to interrupt the playing audio so the incoming call would still silence the music. The problem is what is used to play the audio and the interference of other software on that, so that's the part you should implement differently.
-
Monday 27th April 2026 11:24 GMT Robert Carnegie
I don't know if you can disable or divert incoming calls on the phone.
Wanting to use Bluetooth is a complication, but you could still use some kind of "Faraday cage" arrangement. When I needed a phone case for a 2G candy bar, I tried using a metal spectacles case. I didn't get any calls, and my battery ran down, so I think the case was blocking radio signals, and my phone and the network were going crazy trying to talk to each other.
Perhaps you could lay the Bluetooth phone either inside or under a metal biscuit tin or cooking pan - in such a way that the phone can't "see" the outside world of phone signals, but it can "see" the Bluetooth soundbar, either directly above it, or directly below.
-
-
-
-
-
Friday 17th April 2026 09:16 GMT Bebu sa Ware
Had similar
The landline (some cisco thing) back in late '90s early '00s would interfere with the adjacent Unix workstation's sound system. I never used the sound system but just before the phone actually rang the sound system would snap, crackle and pop.
This helped to enhance my tech aura when consulting with a user, I would interrupt saying "excuse me, I have to take this call" just before the phone rang.
Plenty of perplexed users. A BoFH that can actually see the future is a window far too open for most people.
No one tumbled to it before the phone system and handsets were upgraded to some boring voip thing.
-
Friday 17th April 2026 11:05 GMT MattieD
We call her Ruth
We've got a staff member who is magically capable of breaking anything she goes near. Doesn't matter what - software or hardware - as soon as she's nearby something goes titsup with it.
Ruth is the person for whom 'You should never see this error' was written.
It's quite a handy skill though. We include her in all the demos and proof of concepts that companies want us to try. If it can withstand Ruth then we're interested.
-
Friday 17th April 2026 12:02 GMT Andy A
It's not just GSM
We had a customer who regularly had a machine lose the contents of its drive - MFM in those days. They would extract the machine from their "computer room". We would do a low-level reformat and restore the image and all was well for a couple of weeks.
One day we were there for a different reason, and the machine failed.
We went into the computer room and realised that there was an old-fashioned landline phone on top of the plastic PC case. When that phone rang, its bell caused a huge magnetic field, and the platter underneath it got wiped.
A quick trip to the cheap shop up the road netted a modern phone with a squawk for a tenner. It saved a good few long drive callouts for us.
-
Friday 17th April 2026 13:45 GMT NITS
Re: It's not just GSM
Back in the day I was using my laptop to troubleshoot the network connection to a cash register in a department store. I placed my laptop on the counter and went to the switch closet. Upon my return, the machine had blue-screened. I had absent-mindedly set it atop the magnetic thingy that detaches the antitheft tags, and it had degaussed my drive while it was running.
-
-
Friday 17th April 2026 12:16 GMT Jedit
An oldie but a ... goodie?
This reminds me of a famous story. A man was condemned to death in the UK. They got him on the gallows, pulled the lever, but the trap stayed closed. They took him back to his cell, tested the trap and it worked perfectly. Execution take 2 - same as the first, the trap wouldn't open. They successfully tested the apparatus again and gave it one more go, but again the execution failed. Having tried three times, the condemned man's sentence was commuted to transportation for life.
Unfortunately for the next guy in line, they figured out what had happened: a board in the gallows frame had warped slightly, and if someone stood in a certain place it would bend and wedge the trap closed. During the test only the executioner was on the platform and so the trap worked, but when the full execution party was present the chaplain would be standing at the crucial spot.
-
Friday 17th April 2026 14:38 GMT Daedalus
Sometimes a little human engineering is the key
I mean, if all the hardware fixes aren't dealing with the problem, maybe asking people what goes on around the issue could be the way to diagnose it.
Of course, asking people to connect two things together, or even remember that they happened, can be a bit much. We hear a lot about "xxxxx doesn't work" when what really happened was something popped up on the screen and was immediately dismissed without being read.
And then there's ego:
"What were you doing when it happened?"
"Are you accusing me of causing it?"
Soft skills, which are often deficient in your average techie, sometimes save the day.
-
Sunday 19th April 2026 06:40 GMT Bob Royal
Re: Sometimes a little human engineering is the key
In the early days of computer controlled telephony, we had problems in private exchanges in India & Scotland.
The software would crash hard on a regular basis, which meant all calls cut, start again.
My mate was in Bombay (before they tried to change the name) & knew it was something to do with the Operators Console, but what?
I was in Edinburgh with the same problem, so off to observe the operators, hei-ho.
(NOT crumpet watching skiveing believe me. Not in Scotland!)
It happens! The m/c crashes.... What did you press? "Nothing."
Come now, which button? "Nothing."
Show me. Seems "nothing" is the Scottish word for zero?!!
-
-
Friday 17th April 2026 15:05 GMT Old Used Programmer
Another form of external interfernce
A company I worked at would push software updates to in-store servers and pull daily sales activity over night. At one store, we were having a terrible time doing this. After a lot of swapping in new components, someone noticed that the phone line used for data transfer connected to the building right next to the transformer for the neon sign...that was only turned on at night...
-
Friday 17th April 2026 16:09 GMT cob2018
"Technician Aura" takes many forms
Back in the '80s I was employed by a US military contractor, which of course had occasion to develop software to run on military-grade computers. The model in this incident had already been around for probably 15 - 20 years, and the unit in question was DEFINITELY one of the older ones ( serial number 18, IIRC ) and had been the target/victim of multiple vendor updates and modifications performed on-site.
As one of the few people who were actually trained on the maintenance of this beast, I got the call from the software development folk when issues arose. This time around, they could do partial builds of the operational software, but when the FULL build was attempted, it would crash just before completion. Several times I visited this machine, and even though I never got within 5 feet of the unit, it would perform PERFECTLY as long as I was there to watch. Go figure.
Eventually, the situation deteriorated enough that I was able to observe the failure for myself, and after a rather long day with diagnostic program listings, logic diagrams and schematics ( I said it was old ) I was finally able to pinpoint the problem and identify the faulty circuit card to replace. Problem solved, but it took a while for my blood pressure to recover.
-
Saturday 18th April 2026 02:26 GMT Chloe Cresswell
Re: "Technician Aura" takes many forms
I had a friend who worked on a machine with multiple cards.
To test one, you'd pull the card, put an extender card in place, so the card under test was outside the chassis so you could get to it for testing/diagnostics.
Worked fine till one time he met a fault card that turned out to have timing issues.
Didn't work properly in the machine at random.
Add the extender card, and it worked perfectly under test.
The extra traces of the extender "fixed" the timing issues, making it a bugger to work out the actual fault!
-
-
Friday 17th April 2026 16:20 GMT Patrick R
I've had interferences from the street light on my TV. During Euro 2024, halfway the evening match, the image would suffer interferences. Both neighbors were OK. It took me 3 days to figure it out. My coax TV cable runs 40 cm vertically before going in through the wall, and the street lamp is just a meter away. The lamp was turning on and flickering from ~22:30. It was fixed 2 weeks later. Side note: last week we got LEDs.
-
Friday 17th April 2026 16:55 GMT Paul Cooper
It's not just hardware!
I used to manage and act as in-house technical support for a complex commercial software package. Users often had to follow a chain of operations to get the required result. The number of times I got "Paul, it isn't working - what's wrong?" but it worked perfectly when I sat next to them and guided them through the procedure was astronomical! And funnily enough, it was usually the same people. I think that there are users who simply don't quite "get" that EVERY step in a procedure is important - that if you miss one, it won't work!
-
Friday 17th April 2026 21:02 GMT Scene it all
I was working in an IBM shop with a 370/158 that suddenly started crashing in random jobs. This eventually turned out to be a failure in the hardware that advanced the Program Counter, taking into account the 370's variable-length instructions. It would advance the PC by 4 sometimes, when it should have been by 2, thus skipping over some short instructions. But along the way to figuring that out, IBM sent along some factory technicians with a bizarre test device that looked like a cross between a hat rack and a bed spring. They set this thing up a few feet from the CPU cabinet and opened the doors. Apparently this thing generated all sort of RF and static electric interference fields that would expose any flaky circuits. So they must have had problems with this sort of thing to bother inventing such a contraption. This was before the days of cellphones. They did not find anything that way - the final problem was diagnosed by the senior IBM software support guy, going over crash dumps, looking for patterns.
-
Saturday 18th April 2026 20:52 GMT Vincent Manis
The University of BC had a System/360-67 running the Michigan Timesharing System (I still miss that combo sometimes). One Fortran program would only run or fail randomly. It turned out that there was a timing heisenbug affecting one of the arithmetic instructions—I forget which. After tracking this down, the customer engineer solved the problem by opening one of the cabinets and snipping a small bit of cable out. This change in propagation time made the bug go away, never to return.
Try that on a system-on-a-chip!
-
-
Friday 17th April 2026 22:47 GMT Vincent Manis
Back in the day, many friends were Amiga fans. I was always impressed by their specs, but every time someone offered to show me some Amiga thing or other, the machine would fail just as they were about to demonstrate it. To this day, I have never seen an Amiga actually running. I am convinced I emit an anti-Amiga aura.
Back in the GSM era, I routinely found that my phone would erase hotel room keys. I learned to keep my phone in a pocket with nothing else, a habit that persists to the present day.
-
-
Sunday 19th April 2026 04:19 GMT M.V. Lipvig
Re: Can't monitor something without changing it
This, exactly. A bane of my existence in the telecom world is the intermittent error condition. Customer complains, we test, no trouble found, problem's gone. Two weeks later, customer calls it out again. Same problem, comes clear while testing.
One thing that can cause this is wet cable - water gets in and causes a minor short where a little power bleeds out but not enough to take it down. This requires drying and sealing the cable, or replacing it. Another issue is DSX jack corrosion caused by passing a light current through dissimilar metals. Over time this causes corrosion, usually fixed by sliding a plug in and out a few times. If that doesn't get it, then the DSX is replaced. Finally, there is a defective cable. No idea how a cable inside a rack that is untouched for years suddenly goes bad, but it actually happens quite a lot. This requires either a new cable or a new head installed on the existing cable.
If your telecom provider gives you any of these RFOs, just press the I Believe button. We're not blowing smoke at you. The fixes I just detailed account for 40 percent of my circuit fixes. I will tell you this though - some techs will exercise the jacks every 2 weeks until the end of time if you let them. It only takes a few seconds and the problem is gone now and with any luck, they won't be on duty next time. If the last fix was exercised the jacks and it's been less than 6 months, demand that they replace the DSX jack.
Oh, and squirrels LOVE fiber cable, enough so that all the telecoms I've worked for had animal chew as an official RFO and one specifically had squirrel chew as a code.
-
-
-
Wednesday 22nd April 2026 15:43 GMT Rob Daglish
Re: Can't monitor something without changing it
I have the opposite problem - when it dries up, the ground dries out and shrinks, causing the junction box to slip and move, shorting the cables so my broadband goes off. Last OpenRetch engineer suggested I take a watering can to the bottom of the pole when we have a dry spell...
-
-
-
-
-
Monday 20th April 2026 06:05 GMT C R Mudgeon
A bang is worse than a crash
There's a quarry not too many miles from where I grew up. Back in the 70s there were signs on the road that ran past it, warning people to turn off any radio transmitters. It seems there was some risk of a CB radio or whatever detonating the explosives the quarry used, before the workers were a safe distance away!
(I've never understood how that might happen, TBH. Would they have been using radio-controlled detonators, perhaps?)
-
Tuesday 21st April 2026 11:30 GMT H in The Hague
Re: A bang is worse than a crash
"(I've never understood how that might happen, TBH. Would they have been using radio-controlled detonators, perhaps?)"
The electromagnetic field of a transmitter might just induce enough power in the detonator circuit to trigger it. Probably not relevant in most cases, but with explosives it's worth taking safety seriously. Similarly, oil rigs maintain radio silence during some operations with explosives, see e.g. https://glossary.slb.com/Terms/r/radio_silence.aspx
-
-
Monday 20th April 2026 12:17 GMT Guido Esperanto
They say opposites attract.
My wife has - what I refer to as her ESPN (Mean Girls!) whereby just tech in her vicinity will do weird and unpredictable things.
I as the resident nerd of the household will bask her in my glorious aura and those problems seems to vanish, making a complete liar of her.
-
Tuesday 21st April 2026 20:06 GMT J.G.Harston
******ing portable/mobile phones. If I sent a three-legged wooden horse galloping through nearby electronic equipment with my amateur radio, I'd have the licensing authorites down on top of me like a huge pile of building material stripping my license from me. Yet ******ing phone companies get away scott free.
-
Tuesday 21st April 2026 21:39 GMT J. Cook
The only goofy RF interfereance story I have is from the church I used to attend- the digital radios the local police use would sometimes emit frequencies that picked up on the church's wireless mics, which happened often enough that the people using the mics would turn them off when they were not in use.
the last time I was there, they had swapped over to using hardwired mics instead.
-
Wednesday 22nd April 2026 14:49 GMT molletts
I remember my Dad coming home from Christmas "Midnight Mass" (which has got steadily earlier as the average age of the congregation at his church has increased and was at about 5:30pm last year) on one occasion and telling Mum and me (a pair of heathens) that the nearby pub was having a special comedy night which had been interfering with the mics at the church!
I don't know what they did but I'd have been inclined to mute them at the mixer but leave them transmitting just for the satisfaction of knowing that they might also be interfering with the pub! Return the favour, so to speak ;) It might have made some of the jokes a bit surreal. "Y'know what I told the mother in law when she came home?" "Take this cup and drink from it. This is my blood..."
-