News: 1619767806

It was not unknown to get a service call to a desk in the morning, "My computer doesn't work". Upon inspection I would find RAM missing, the CPU missing, the hard drive missing. Changing out the contract cleaning staff who, ahem, "serviced" the office spaces in the evening cured the problem, mostly. Fitting security screws to the cases reduced the failure rate a bit too.

Indeed. It's not unusual to find cleaning staff consider their pay as their second source of income

That happened where I worked. Security cameras were install. Next month a brand new, and at the time, very expensive 60" TV for the video conference suite went missing. And surprise, the security tapes for than night were all found to be blank.

Exit one complete security team.

I spent some time at a customer site where they were building a huge wing next door to where we were working, and it was great watching all of the activity as plumbers took in furniture for the toilets, and the carpet fitters took in the carpets.

One day there was no activity at all. Someone had done a spot check, and found "officially (from the amount of stuff delivered) " there was at least one wash basin per person, and the amount of carpet that had been delivered would have had the floor 1 ft deep in carpet!

The stuff was being taken in during the morning, and out again into vans in the evenings.

All this was discovered because a policeman was taking his dog for a walk one evening and spotted the activity.

Had the same thing at a previous role. IT desks were in an open-plan area and although the main kit was locked to the desks with Kensington locks and bulk of our spares were in a secure room, odd cables and accessories did disappear from time to time from our desks. In the end I set up a couple of webcams on my system (with blu-tack inside to shield the LED), and put "I-Spy" on my machine to record whenever movement was detected.

Few weeks later, my toolkit disappeared. Looked back over the footage and found one of the company window-cleaners (who came around once a month to clean the inside of the windows) discretely slipping a towel over the toolkit and walking off with it.

Showed the footage to security, and the window cleaning company was promptly kicked off-site.

Printers are intrinsically evil, but that is especially cruel.

In the late 80s and early 90s it was not unusual for offices to have all the memory cards removed from desktop PCs by burglars, but that is definitely enemy action.

Shame about the name of the new analysis system. Our attempt to develop the Program Error Notification and Information System survived two design meetings before the name was changed.

We got away with the "Automated Reporting Systems Environment" for a good 6 months before anyone important noticed it.

We also got "T(company name) Information Technology System" past three very important people, before they showed it to someone as childish as us :-(

Strangely, they were all ok with "T(company name) Weekend Activities Team" as the name of the social club for over 10 years.

Which brings to mind a story of another system. This was to track oil tankers used in commodity trading. The name was "Tanker Information Tracking System".

Which was actually used, until one young female analyst found a new trading route, and promptly announced it with glee, only for someone to yell across the desk, "Then get it on your T*TS".

Management heard, and the name was changed that afternoon!

True story.

When I worked in a Bank the Mainframe Logins were of the form Firstname + First Letter of Surname.

We had an English Contractor named Chris Tompkins who was very very knowledgeable.

If there was ever a question about the system we would happily reply "Christ knows"

The tech blaming the cabling obviously didn't ship the shoddy card, but I think he can have a bit of the crow pie anyway...

Blame the cable was always good opening gambit. I have a very vague recollection that there existed two versions of D-types; normal and crossed, and Sod's law decreed that the only spare that could be found in the lab was the wrong kind.

See also the fun with SCSI terminators and if you needed a single-ended or differential.

I bought a dodgy monitor, weird random blankness. took it back, the ppl at the shop said no it's fine take it away son. cue grinding teeth. a week later after faffing, I found out I had a dodgy VGA cable. so there you go...

Way, way more than two. Does it use H/W control & if so which lines? (Memory says there were a couple of alternative conventions.) Then there are D25s & D9s. With D9s whoever laid out the board headers hadn't been told that the pin numbering on a header isn't the same as that on a D connector so a plain ribbon cable won't help*. And if all else fails a wrong gender connector can be thrown into the mix.

* Got caught out by that one in reverse back in the early days. Spotted that the pin numbers on a 26 way header didn't correspond to the numbers on the D connector & started to wire up a custom cable. Oh, look, if you just use a ribbon...

Back in the 80's, the company I was working for (one of the oil majors) had a serial network around their exploration HQ. Our team had Dec Rainbow PCs running Lotus Symphony to manage our part in the procurement process. All high tech as we could communicate direct with vendors via telex, straight from our PC. Printer communication was also serial. It wasn't uncommon for printer settings to get changed and, without knowing what they were, it was next to impossible to print anything. IT support was available, but they were overworked. Enter my Psion II PDA: it had a program that scanned through the full permutation of serial commands - hook it to the recalcitrant printer, run the program and whichever settings worked would be printed out. Simples!

Our offices had more serial sockets than we needed at any one time, and unused lines were often re-patched for use elsewhere (office relocation had a full-time team). But we'd regularly find the ones we were using had been reallocated over the weekend. Enter my Psion II again: plug it into the various ports to see which were communicating and the office was back up and running.

Must dig the ol' Psion II out from whichever loft box it's in for a bit of nostalgia...

Blame the cable still is a good opening gambit.

Just yesterday, I received a logic analyzer (one of these cheap Saleae knock-offs).

Plugged it in with the supplied mini-USB cable:

$ dmesg

[20830.498682] usb 1-6: New USB device strings: Mfr=1, Product=2, SerialNumber=0

[20830.498684] usb 1-6: Product: Logic

[20830.498685] usb 1-6: Manufacturer: Saleae LLC

[20958.751963] usb 1-6: USB disconnect, device number 44

[20991.903015] usb usb1-port4: Cannot enable. Maybe the USB cable is bad?

[20992.030997] usb 1-4: new high-speed USB device number 47 using xhci_hcd

[20993.011142] usb usb1-port4: Cannot enable. Maybe the USB cable is bad?

[20993.011180] usb usb1-port4: attempt power cycle

[20994.231147] usb usb1-port4: Cannot enable. Maybe the USB cable is bad?

Plugged in another cable I had laying around and it works.

Good thing it is my habit to check dmesg when I plug in new hardware.

And thank kernel developers for these useful messages.

Often saves a lot of head scratching.

Although I wasn't responsible for it I got involved in the first Cat-5 cabling that went in at a former employer (we'd previously used thinwire Ethernet with the make-before-break plugs).

The cabling people insisted it should all work but one part of the building would not connect whatever they tried.

I took one look and replaced the straight through cable from the master hub to the one servicing those connections with a cross over cable - this was before auto-sensing ports.

The so-called professional installers had come along with a mixture of identical looking straight and x-over patch cables then wondered why they had problems. They also didn't have enough sense to look at the port lights on the hubs and see hey weren't lit.

Not surprised the supervisor installed it himself. Back in the day cabling cost real money and was a black art.

Once upon a time, a contractor mate in Novell picked up a job doing the setup for a very tight-fisted bunch. He was bemoaning the fact that a) he'd been screwed on his price and b) they wouldn't pay for decent cabling. He showed me the cabling quotes and said they were going for the cheapest (thinwire ethernet - yes, that long ago).

I saw the big number at the bottom and swore. I asked if they'd like the place floodwired in Cat5 for 16Mb token-ring, with a proper cabinet and patch panel next to the server, for less money. He thought they might, but how? I pointed out that I knew what I was doing and had all the tools required.

We did it over a weekend and pocketed well over half the big number at the bottom[1] between us. No wonder there was no shortage of choice in cabling contractors, the profit margins were astronomical. One of the more profitable items was the patch cables, which I made out of left over cable and crimp-on ends (10 quid for a bag of 500). It seemed cheesy to charge per cable, but I undercut the opposition by 50%, at 10 quid a pop, to salve my conscience.

[1] Even with the higher spec kit, cards and cable than the etherstring lads. They must have been really coining it.

Working in computing labs where students can come in pretty much when they want, we've always had a bit of a problem with theft. Even to the point where we have to secure the cables to the desks.

This usually works. On day, I got a call to say a student had stolen a mouse. Not sure why I was told, as we only pay about a fiver for the mice, so we usually just replace them.

Then, I saw the computer. The student hadn't done the sensible thing and cut the cable ties. No, No. He or she had cut through the mouse cable, leaving the USB plugged into the computer and walked off with the mouse. This was almost worse because he or she *could* have shorted the USB socket.

I was tech support manager for an organisation with a large in house dev team. After several years of debacles where the devs produced applications which could not run on end user PC's I restricted the memory size on the dev PC's when they were replaced, they still had plenty of ram for the time and it was more that the vendor recommenced. These apps were being rolled out to hundreds or thousands of machines and memory upgrades were cripplingly expensive at the time. When the next app was deployed sure enough the app failed to load due to insufficient memory. I called for a desktop engineer to bring and install more Ram to get the first team operating and when he opened up the PC half the memory as missing. I smelt a rat immediately and stormed off to the Dev office. Sure enough the developers machines had 4 times the memory they were installed with. They had actually stolen the memory from the user PC's because 'users don't need that much' I was absolutely furious as buying higher spec user pc's to run the app had cost me hundreds of thousands of pounds. Sure enough even with the additional memory installed the app wouldn't work. Not being willing to find another £500,000 to upgrade the entire estate the roll out was delayed by 2 weeks while the lazy b*stards actually did some tuning. Done even get me started on the quality of their server side coding.