Cabling survived dungeons and fish factories, until a lazy user took the network down

Welcome once again to On Call, The Register's weekly reader-contributed column in which we share your stories of delivering excellent tech support amid your colleagues' ambivalence, anger, and unjust admonitions. This week, meet a reader we'll Regomize as "George" who told us about his first job in IT. "Fresh out of …

[lglethal]

What is it with managers and training costs?

In a previous job, I had a manager tell me he wanted me to move to another project. No problem, I said.

But I would first have to take a training course in order to work on that project (due to the special PLM software they used). No problem, I said.

And I would have to pay for the training course. Not happening under any circumstance known to Man or Diety, I said.

Cue a blazing row, with him shouting at me in a full office, which I simply sat through, before informing him I would be raising a HR complaint about his conduct, and informing him that under labour law he didnt have a leg to stand on (on either the training costs or his conduct).

Unbeknowst to me, the HR boss had walked into the office to find out what the shouting was about, and clearly heard my response. She hooked the manager, dragged him back to her office.

The next day, I received an emailed apology, confirmation that I would not be paying for the training course, and news that i would be transferring to work under a different manager.

The asshole manager never spoke to me again, something which I considered a blessing, and all went well.

Still what is it with Managers and training costs? They always seem to take it so personally? I guess for a lot of them it's tied to their bonuses, so of course it is personal for them...

[Korev]

Re: What is it with managers and training costs?

Where I used to work the powers that be kept reminding us we had a training budget and then berating us for not using it.

The conversation went something like

BOFH: "There's a course here that I'd like to do"

PHB: "Sorry we don't have any money in the travel budget to send you"

Fast forward a month or so:

PHB: "Don't forget you have money in the training budget"

[GlenP]

Re: What is it with managers and training costs?

When I worked at the local college (now a Uni) there were two different travel rates, one for attending training which was basically just about enough to cover the fuel and one for other journeys which was the statutory rate including wear and tear, etc. It was amazing how often the person at the council responsible for paying the expenses would redefine a "meeting" as "training". He'd also get out his road atlas and work out the shortest route between two places then only authorise the expenses for that instead of for the sensible route you'd actually taken.

I knew him socially as well, and he was just as tight with his own money - he'd happily accept drinks in the pub (from the few who ever bought him them) but would never buy any back.

[Sam not the Viking]

Re: What is it with managers and training costs?

When our accounts people started slashing expense claims, I pointed out that they were mere clerks who didn't have the authority to alter an expense after it had been approved by my manager. Unbelievably, they pursued the matter up their own chain of command where we had the disgraceful scene of senior managers having a row over whether engineers could be trusted. When it was pointed out who was actually bringing in the business that was paying everyone's salary....

For some reason, the attitude still riles me...... But we won the battle ---->

[Chris 15]

Re: What is it with managers and training costs?

Well maybe try and find a course that isn't being given in Tahiti or Hawaii? ;-)

[Mishak]

Not training related, but...

I once had a manager who came up to me at my desk and, in an angry voice, said "You and I have a serious communication problem!".

This was news to me, so I replied with "You must be right, as I have no idea what you're on about".

He then stormed off and I stood up and started walking into the engineering lab - at which point I felt a hand on my shoulder and was physically dragged back into the office "to continue the conversation".

That incident was later cited as "a verbal warming" when he decided to formalise another similar incident.

My friend, who was a part time police officer, said they would be more than happy to support my assertion that this was assault!

Edited to fix a typo.

[Prst. V.Jeltz]

Re: Not training related, but...

"Edited to fix a typo."

Very thorough of you to clarify that, I'd have just kept quiet about it unless the typo somehow completely reverses the meaning of the message , which *has* happened to me a couple of times

[Aladdin Sane]

Re: Not training related, but...

Was one of those times when you edited out a typo but left in "warming" instead of "warning"? Unless the manager was full of hot air...

[Mishak]

Re: Not training related, but...

Nah - that's just to remind me that I need new glasses...

[JWLong]

Re: Not training related, but...

A verbal warming

I typed that in to see what auto-screwup would do and it maligned it all by itself.

Auto-screwup does some weird shit sometimes!

[Joe W]

Re: What is it with managers and training costs?

My boss is the opposite. "Look, I know you are busy" (yes) "but I think you should broaden your skills with" (yeah, it is tempting and interesting and I actually want to learn that) "plus there is this project management thing we talked about" (ooh, shite, forgot about that one) "which you should sign up for - just pick some offer that sounds good to you, and not too far away, I'll sign it".

I have now two trainings signed up, with another one I would love to take - but I think one of the team should do it instead. I am a team lead, and that... well... while tempting I think somebody else should do that (though damn, I would love to!). My boss might send both of us, though. You know, no "single point of failure" et c.

[BBRush]

Re: What is it with managers and training costs?

Had something similar to that once before when I handed my notice in at a job.

Boss (massive bully) called me in and told me that he was putting me on gardening leave for the rest of my notice period (yay!) and taking some of my accrued holiday for my attendance at a mandatory course a few months previous (WTF??). A course that had been free, on-site and which he had insisted that we (the team) must come in early and stay late on the training days so that we could clear off the work we would be missing because of the course.

HR was in the meeting and looked embarassed. I found out he got sacked a few months after I left.

[Doctor Syntax]

Re: What is it with managers and training costs?

"HR was in the meeting and looked embarrassed."

They should have done more than that.

[NoneSuch]

Re: What is it with managers and training costs?

Never forget, HR represents the company, not you.

Too many people don't get that until they are walking out the door with all of their their possessions in a Bankers Box.

[bemusedHorseman]

Re: What is it with managers and training costs?

The secret is to railroad HR into perceiving that doing your bidding is "protecting the company"! Liberal use of the phrase "sure would be a shame if the public found out about..." is highly encouraged.

[Doctor Syntax]

Re: What is it with managers and training costs?

"Never forget, HR represents the company, not you."

And one of the things they need to do on behalf of the company is keep it out of tribunals.

[DS999]

They want employees with more skills

But more skilled employees can demand higher salaries so when you leave to get those higher salaries they refuse to pay you, they feel like they got screwed out of the training costs.

[Headley_Grange]

Re: They want employees with more skills

I can't remember who said it but it's always valid.

Boss 1 - "What if we train them and they leave?"

Boss 2 - "What if we don't train them and they stay?"

[Roland6]

Re: They want employees with more skills

I worked for one (european) company who did adopt the approach of training people, and thus encouraging those it didn't want, to use the recruitment agency partner to gain work elsewhere.

They put the entie cost down to marketing, in the belief of having helped people into jobs they wanted, they would be more predisposed to recommending solutions from their old employer to their future employers. Don't know how much business they did get, but the cost of the programme was significantly less than the redundancy pay they would otherwise have had to pay out by several €100M.

[Number6]

Re: They want employees with more skills

Richard Branson: "Train people well enough so they can leave, treat them well enough so they don't want to".

[Doctor Syntax]

Re: What is it with managers and training costs?

One of the advantages of being freelance: you invest in your skills and expect to profit (doesn't always work out that way) while being able to make decisions very quickly. I've been in a position where I was getting to the end of the contract and my client got a new contract calling for XML (which was new at the time), a skill which nobody in the company had, no more did I. "OK, if you give me a contract for that I'll take myself on a training course." They got more contracts with the same requirement and naturally so did I.

[BartyFartsLast]

Re: What is it with managers and training costs?

Ah yeah, I worked under a manager who refused to sign off or send on my annual appraisal because I asked for more training.

I kicked up a fuss and he eventually agreed to send it on.

Then two years later, I found out that he'd actually altered my appraisal to remove the training request and forged my signature on it.

HR couldn't understand why because he had a training budget to use for all of us but only sent two people out of a team of around a dozen.

[anothercynic]

Re: What is it with managers and training costs?

I hope you reported him for that in the end?

[Roland6]

Re: What is it with managers and training costs?

Back in the era this story is from, there were companies, like EDS (Electronic Data Systems Corp) that made repayment of training if you left within a certain number of years a contract condition. Obviously got negative press about it, so don't know to what extent they actually enforced it.

[IanRS]

Re: What is it with managers and training costs?

They did. I used to work there, and knew people it happened to. Some courses had payback periods as long as three years, although the amount did taper off.

[Doctor Syntax]

Re: What is it with managers and training costs?

AFAICR EDS also forbade discussion of pay rates. I suppose the way to leave without repaying training was to discus your pay and get fired.

[Daedalus]

Re: What is it with managers and training costs?

Well, EDS was owned by a billionaire, and he didn't get where he was by being loose with money.

Thank God he didn't get elected President.

[Roland6]

Re: What is it with managers and training costs?

Given the current incumbent, you might want to reassess that statement.

Yes he was a Republican hawk, but I think he actually did believe in democracy and the US constitution and would have consistently upheld it.

[DecyrptedGeek]

Re: What is it with managers and training costs?

At that time my boss was in the same Naval Academy class as that gentleman. He hated him with a passion. Seems he didn't like all the cursing in the Navy so resigned his commission. My boss predicted he would drop out of the race. I hated him because he earned all his money on government contracts and then claimed he was an entrepreneur. I've worked with so many businessman that made the same claim while all their business was government contracts. The whole lot of them are grifters.

[chivo243]

Re: What is it with managers and training costs?

I turned down a contract with the payback all training costs should I leave the company, then some stupid non-compete clause.

[Anonymous Coward]

Re: What is it with managers and training costs?

I asked for more info about the non-compete clause and the IP-belongs-to-us clause when offered a job at one company. The HR guy doing the hiring asked Legal, who didn't respond, and didn't respond, then the HR guy went on vacation... then they withdrew the offer.

By that point, I had gotten more info about both. The IP-belongs-to-us clause really only applies to IP that's related to the business (no matter what it says; them grabbing more doesn't hold up in court), so the two albums my wife and I recorded would have stayed mine. The "can't compete with us for a year after you leave, worldwide" was really "can't help produce competing products", which is not only unlikely I'd end up working on competing products, but even if a future employer had some competing products, I'd probably just work on other products for the first year.

[J.G.Harston]

Re: What is it with managers and training costs?

I've been on some contracts where the contract implies they want to memory-wipe you when you leave so that you no longer have the experience you gained while working for them.

[Daedalus]

Re: What is it with managers and training costs?

Now that's a different kind of "Severance".

[Evil Auditor]

Re: What is it with managers and training costs?

Guess someone there got ideas from the documentary "Paycheck".

[doublelayer]

Re: What is it with managers and training costs?

I think the training thing is, as with many others, starting from a sane place, missing the end of that and running well into unacceptable territory. At one point, and probably still now but less so, some people used to get really serious training as employees. A course that lasted a year or substantial part of one (or, in limited cases, something even longer) that students would usually have to pay to attend would not only be covered by the employer, but the student would continue to get paid their normal rate during it and would have transferable skills afterward. It made some sense to put conditions on that because, if people left immediately after the paid education, the employers would likely stop paying for anyone's*. A similar thing could be true when it's something being done for the primary benefit of the employee. Most employer-provided training that I've seen people take is either something the employee doesn't care in the slightest about or something from which the employer is going to benefit more. But you usually can't say that because they're suggesting the training isn't useless.

* The typical response to the paid training response is that, if employers were paying for training, then they could always hire trained people from each other. This is not really true, since there were always some companies that didn't pay for training and only hired people who already had it, whether they got it themselves or were given it by a different company. We mostly live in the world where this died because most employers now expect specialized skills to have been learned at various types of schools, colleges, and universities rather than at the start of employment. This means that almost all of the training that is still paid for by employers is the kind that shouldn't be paid back because they're the one requiring it and it doesn't help the employees anywhere else.

[J.G.Harston]

Re: What is it with managers and training costs?

I've noticed that more and more recruitment websites are now asking you the *APPLICANT* to pay to see job vacancies. If I could afford to pay to look for work, I wouldn't be looking for work.

[Anonymous Coward]

Re: What is it with managers and training costs?

Worked at one place where if you left you had to pay back some / all (depending on time since course) of courses they paid for.

I let them know I did not accept that & crossed those paragraphs out of my contract.

The training courses they paid for were not ones that employees suggested & wanted to do, but were courses the employer wanted you to take (so, you may have gained something, but it was mainly for employers benefit e.g. back in the day it helped a company keep Microsoft accredited status, become a "Gold partner" etc. if lots of their employees had MCSE & other Microsoft qualifications).

[Evil Auditor]

Re: What is it with managers and training costs?

Still what is it with Managers and training costs?

A sense of control, of being in charge. Training costs is something the hugest simpleton amongst manglement can "understand" and "manage".

Such as the company that compiled a exhaustive list of trainings, which would be supported in full or partially. On this list are also a considerable number of CAS, which change their names nearly every semester. Countless hours have been wasted with a) changing some relevant training on that crap list and b) justifying and approving specific trainings.

[Ochib]

The Place Where The Sun Does Not Shine

Isn't that a place in Slice?

[KittenHuffer]

Re: The Place Where The Sun Does Not Shine

I thought it was in Lancre, in the Ramtops!

[Michael H.F. Wilkinson]

Re: The Place Where The Sun Does Not Shine

Slice is in the kingdom of Lancre, if I recall correctly

[blu3b3rry]

Re: The Place Where The Sun Does Not Shine

Turn left at Bad Ass, it's just up there hill there. You can't miss it.

[Cyberian]

Re: The Place Where The Sun Does Not Shine

Isn't that 99% of the data centres?

[StewartWhite]

Re: The Place Where The Sun Does Not Shine

"The Place Where The Sun Does Not Shine"

Nope, I think you'll find it's Cheadle Hulme (see https://www.youtube.com/watch?v=20XLWEjN9eI for The Justified Ancients of Mu Mu's full explanation).

[Korev]

Good to see that George didn't Bungle it

[Will Godfrey]

I wonder how many here will get that one.

[Anonymous Custard]

Good to see that George didn't Bungle it

And he was quick about it, indeed positively zippy...

I wonder how many here will get that one.

At least one...

[neilg]

Two.

[Will Godfrey]

Buckle my shoe

yes I am that old

[tiggity]

At least 3 (Rod, Jane* & Freddy)

* interesting personal life with Rod & Freddy!

[Anonymous Coward]

Well, the sun is shining while things are raining down...

[Anonymous Custard]

And then they ask you to paint the whole world...

[KittenHuffer]

Zip(py) it!

Edit - Damn! Got beaten to it!

[Anonymous Coward]

I did

https://www.youtube.com/watch?v=CgbcQIT7BMc

[Bebu sa Ware]

Good to see that George didn't Bungle it

He clearly didn't need to be peeled off a tree. ;)

[CountCadaver]

Maybe had help from Rod, Jane and Freddie?

[jake]

Tricks of the trade

"What's your preferred method for finding cable faults?"

I simply ask the Boss's personal secretary. 80% of the time she'll point me in the right direction. 15% of the time, she'll know who I should talk to. The final 5% is where I make money troubleshooting.

_{At least the troubleshooting is pure profit these days ... the TDRs were paid off decades ago.}

[swm]

Re: Tricks of the trade

A TDR can locate the distance to a fault very quickly.

[M.V. Lipvig]

Re: Tricks of the trade

Electrical/optical distance is not necessarily actual distance, if the cabler stuck the company for a few extra miles of cabling and just looped it on itself in the cable racks. With optical, you might see a fault a thousand feet away while standing under the break.

[Korev]

Things that didn't happen

The story in the factory seems a bit fishy to me

[Inventor of the Marmite Laser]

Re: Things that didn't happen

Still spawns comments though.

[Korev]

Re: Things that didn't happen

That was off the scale

[Mishak]

Re: Things that didn't happen

You really are trawling the depths.

[Inventor of the Marmite Laser]

Re: Things that didn't happen

It's freely available on the net

[Roger Greenwood]

Re: Things that didn't happen

Yes I wouldn't like to perch on top of a forklift

[Richard 12]

Re: Things that didn't happen

You'd be shocked what people haddock to do in work plaices.

Standing on forks used to be very common. Sometimes they'd put a pallet on the forks, if you were lucky. And if you were very lucky, someone would put a pin in the end of the fork to stop the pallet sliding off.

Never forget that Health and Safety laws and regulations are written in literal blood and guts.

BTW, There are now cages designed specifically for this. Never used one though, still feels like a bad idea.

[CountCadaver]

Re: Things that didn't happen

Pretty much should be a scissor lift or a cherry picker (With a trained operator)

[Phil O'Sophical]

Re: Things that didn't happen

Standing on forks used to be very common. Sometimes they'd put a pallet on the forks, if you were lucky.

Or: https://thekilpatrickgrouppa.com/wp-content/uploads/2021/06/forklift.jpg

[ArguablyShrugs]

Re: Things that didn't happen

Staplerfahrer Klaus wholeheadedly (?) agrees

[Anonymous Coward]

Re: Things that didn't happen

Doch!

[NXM]

Re: Things that didn't happen

That made me laugh so much I had to buy a disc of it.

For similar fun, try Black Sheep or Brain Dead.

[Anonymous Coward]

Re: Things that didn't happen

Working briefly at a home improvement store, I had a co-worker get fired for this. Standing on the rolling stairs, he couldn't quite reach that last item on the pallet on the forklift, so he stepped onto the pallet... right as Health and Safety walked by. That was the last straw.

[Hazmoid]

Re: Things that didn't happen

used a cage on a forklift many times. You have to trust the operator though, and most of them have a requirement now to wear a harness which becomes a problem when the item you want to access is just out of reach.

[David Hicklin]

Re: Things that didn't happen

> Standing on forks used to be very common

Back in my first job after school the place I worked for built diesel generator sets, and one they had was too heavy/bulky for the forklift to move it from assembly to the test beds as the backend of the fork lift just lifted off the ground

Piling some steel work on the back did not do the trick so several apprentices (they were expendable in the 1970's) climbed on the back and were just about enough to balance it

[Anonymous Anti-ANC South African Coward]

Any Chronicles of George folks around here?

[Andy A]

Yes. I used to maintain our company's own version of ICL's George 2, known as George 2++. The assembler used to compile it (called GIN) had a wonderful feature - you could write code to be executed at compile time, so you could teach the old dog new tricks!

[Bebu sa Ware]

Ah, the joys (?) of 10Base2

Really scaled (not!) Once you had a hundred or more user PCs on the network with bridges and repeaters there was hardly a day when some user induced problem didn't take a segment of the network down. You always carried a couple of 50Ω terminators to locate the offender with a binary section — the network not the offender although tempted.

How a piece of 75Ω TV coax came to be used to connect the PC into the network was a mystery that the user was unable or unwilling to illuminate.

The commonest offence was a user relocating a PC and taking T Piece and two coax cables. When the PC was the only one in the room understandable, I suppose.

[Conrad Longmore]

Re: Ah, the joys (?) of 10Base2

First job was in a college, they decided to roll out a lab connected to a NetWare server for teaching. It was quite neat, the coax cable ran through the trunking on the walls and then broke out and the coax went through the cable management system at the back of the desks, connected up the PCs and then went back into the trunking on the same double connector it came out of.

To give the lab more flexibility, there were several of these connectors in the lab which where unused unless we change the desk configuration. Here there was a tiny U-shaped cable from connector to connector. We discovered that the students found these fascinating and would unscrew them to see what they did. What they did was summon an angry technician. I did have to laugh because at the time I was running VAXes which were all connected by screwed-in 25-way D-type connectors just as God intended.

[Doctor Syntax]

Re: Ah, the joys (?) of 10Base2

It was a lot more manageable than the previous generation of cable.

[jake]

Re: Ah, the joys (?) of 10Base2

I have three runs of thicknet strung through the house, and out into the machineroom/museum/mausoleum/morgue ... seemed to make sense to put it in while the walls were open and before the closed-cell insulation was sprayed. All are in use occasionally, with the one between here in the office and M4 used near daily.

[Doctor Syntax]

Re: Ah, the joys (?) of 10Base2

"All are in use"

Apart from anything else, they'll hold the house together.

[Dwarf]

Re: Ah, the joys (?) of 10Base2

Why would you need a network connection in a morgue ??

I guess the residents dont need the Internet much, other than to do their tax returns.

[John Brown (no body)]

Re: Ah, the joys (?) of 10Base2

"I guess the residents dont need the Internet much, other than to do their tax returns."

That made oi larf :-)

[Anonymous Coward]

Re: Ah, the joys (?) of 10Base2

Ah, memories...

A University PC open access area, full of diskless workstations, booting from LAN, all "nicely" cabled up running off a LAN card at the back of a Novell server. One terminator at the server, one at the end of the run in the room.

Student turns up to use PC, finds none free. Sneakily removes terminator from end of run, and watches as the PCs stop working and people give up in disgust. Replaces terminator, reboots a PC and voila, gets a quiet lab at the same time.

Only, Novell servers didn't like the network being mangled like this, and would start racking up packet receive buffers, once that hit the defined maximum, goodnight network on all cards, not just the one with the dodgy terminator. The command "set maximum packet receive buffers=" to increase it just enough to keep things running until a convenient time to reboot the server after replacing the terminator became somewhat muscle memory (particularly avoiding the obvious typo, which could well refer to the remover of the terminator...)

Did wonders for the reputation of the service and although sometimes that reputation was warranted, we got a lot more stick than we deserved.

Posted anon otherwise someone will probably recognise the Uni...

[PickledAardvark]

Re: Ah, the joys (?) of 10Base2

The university where I worked first ethernet wired the buildings housing science and engineering with 10Base2. 75 ohm cables were still popular on some scientific instruments -- oscilloscopes and signal analysers. Their effect on ethernet was very intermittent.

[Doctor Syntax]

Re: Ah, the joys (?) of 10Base2

The first Unix system I used, back in the 1980s was all RS232 terminals but there was also a TV coax-based network to connect the terminals into the cupboard next to the lift with predictable results machine room. I can't remember how they were all broken out into D connectors at that end but the network was a collection of little boxes, each with a Z90 inside (allegedly, I didn't open one), a D connector, a TV coax socket and a short length of TV coak with a plug on it. I can#t remember a power supply for them - perhaps they were powered through the coax.

[Roland6]

Re: Ah, the joys (?) of 10Base2

This summer had a garage tidy and rediscovered a box of networking stuff: complete 10Base2 network for 5 systems, including a couple of 3Com ISA bus and PCMCIA adaptors and the coax "loops" for the wall sockets. Still deciding what to do with them, as I also rediscovered my first laptop a Zenith 386 WfWG machine, which still works.

[Roland6]

Re: Ah, the joys (?) of 10Base2

>"How a piece of 75Ω TV coax came to be used to connect the PC into the network was a mystery that the user was unable or unwilling to illuminate."

Assuming this was UK, I suspect they read an article about Ethernet networking which would of had a US-bias and thus read that the cable used was TV coax, not being aware the UK and US have different standards for "TV coax"...

[Herby]

Re: Ah, the joys (?) of 10Base2

TV Coax:

In both the USA and the UK, TV coax is 75 ohm (it started out as RG-59, and eventually became TG-6 (quad shielded). Ethernet uses 50 coax. For 10BASE2 the coax used was double shielded RG58, and for 10BASE5 it was RG213 (double shielded). Most RF things that transmit use 50 ohm coax (like ham radio setups). It is a balance between velocity factor and power requirements (if I remember correctly). The 75 ohm coax was used because folded dipoles are naturally 300 ohm things, and they used 4:1 baluns which got things down to 75 ohms.

Thank you for following this small history lesson.

[Antron Argaiv]

Re: Ah, the joys (?) of 10Base2

Old Ethernet designer here. It was *never* (75 ohm) "TV coax". Metcalfe and Boggs initially used RG-213 (double-shielded 50 ohm solid center) and the specs for Yellow Ethernet coax are remarkably similar to those of RG-213. All of it was always 50 ohm impedance. Orange Ethernet is the plenum safe stuff and two 100+ foot lengths are currently in my backyard connected to HF antennas, which work very well. 10BASE-2 was something similar to (50 ohm) RG-58, but slightly thinner and with a 100% coverage foil shield. I have some and some of the crimp-on BNCs, and it makes wonderful test equipment jumper cables.

As long as we're on the subject, the preferred method at Data General of connecting to yellow or orange "thicknet", was to avoid vampire taps entirely, and use crimp-on type N connectors and 3Com transceivers. Plenum AUI cables were even more inflexible than the thicknet coax, and the silly slide latch on the AUI connectors was never used as it didn't do diddly. I forget what we did, probably just waited for 10BASE-T to appear. AUI cables were responsible for most of the connectivity issues, but they didn't take down the whole network. Thicknet was only used for a few years, we all rapidly transitioned to 10BASE-T as it was far easier to install and maintain, and being rid of shared media made made the network much more reliable and far easier to reconfigure.

There *was*, BTW, an IEEE 802.[somethng] "Token Bus" (not "ring" network made up of CATV (75 ohm) hardline, amplifiers, splitters and drops. AKA "GM MAP" for "General Motors Manufacturing Automation Protocol". We at DG experimented with it briefly, had a couple of huge rolls of CATV hardline in the lab. Chipcom made a whole data network using this technology, but I think it used the Ethernet protocol with a translating headend. The early cable company ISPs used something similar, and it's still around today, and as poor a choice for wide area data networks as it was back then.

[Roland6]

Re: Ah, the joys (?) of 10Base2

>Plenum AUI cables were even more inflexible than the thicknet coax

The memory seems to remember all cables being much thicker and stiffer back then: RS-232, Centronics, SCSI...

> the silly slide latch on the AUI connectors was never used as it didn't do diddly.

Needed to be in the correct orientation. On Sun 3 "under desk" workstations the connection was vertical, with the slide having to be raised to unlatch. This prevented the AUI cable dropping out under its own weight, it also meant a journalist who thought they would be clever and show our demonstration was canned and simply "knock" the cable had to invest a little more effort, drawing her to our attention and so enabling us to tell her to go ahead and remove the cable with the expected effect of things stopping.

>we all rapidly transitioned to 10BASE-T as it was far easier to install and maintain

it did take a while to get there and appreciate telephone style wiring closet cabling had its advantages.

>There *was*, BTW, an IEEE 802.[somethng] "Token Bus" AKA "GM MAP"

Yes, its a shame IEEE802.4 seems to have disappeared (ie. yes I know the IEEE withdrew it decades back, but I haven't found a PDF of the Standard lurking on the Internet), as the concept was sound and had benefits (and disadvantages) compared to Ethernet. At the time, I think cable TV style networking components were readily obtainable and certainly a plumbed in network made sense within the GM production line environment. Today I expect the requirements are more easily and cheaply satisfied by running twisted pair within a metal conduit.

>The early cable company ISPs used something similar, and it's still around today

Milton Keynes cable TV network used similar. The trouble was such networks had been designed as broadcast only, so conversion to data network would involve the replacement of thousands of splitters a sub £0.50 component.

> and as poor a choice for wide area data networks as it was back then.

The concept of Metropolitan Area Network seems to have dropped out of currency, I see IEEE 802.6 was withdrawn in 2003.

I think the reason for 10Broad36 being considered for MANs was the supposed synergy with established cable TV networks.

VirginMedia's UK network still uses DOCSIS as its subscriber data interface to its "CATV" network.

[John Brown (no body)]

Re: Ah, the joys (?) of 10Base2

"The concept of Metropolitan Area Network seems to have dropped out of currency, I see IEEE 802.6 was withdrawn in 2003."

It was cancelled due to the inherent misogyny. It's due to be replaced by Persistent Existential Radial Scope Overhead Network any decade now.

[Andy The Hat]

Re: Ah, the joys (?) of 10Base2

"avoid vampire taps entirely, and use crimp-on type N connectors" ...

Wuss ... what's wrong with soldered N-types? We used them because we couldn't afford the stupidly expensive crimp tools at the time.

Terminating cables for an extension on a bouncy, damp, 2nd floor, flat roof in the middle of winter with a 240V soldering station and a hot air gun (to shrink sleeving over the joint as I was even conscientious back then) was fun.

[J.G.Harston]

Re: Ah, the joys (?) of 10Base2

Would of what?

[mikecoppicegreen]

Re: Ah, the joys (?) of 10Base2

I worked in a company for a while that implemented a 10Base2 network in a semi-permanent arrangement of cubicles set up on the factory floor of a small industrial unit, so there was coax literally everywhere.

There was fairly regular crashing taking place, and we soon correlated this with our test facility - also within the industrial unit we had a 200kV setup to test electrical testgear, and flashovers generated a fair pulse.

We never really solved the problem - we just learned the benefits of saving our work very frequently and doing transactions on the accounts system during lunchtimes.

[Jaxx]

Re: Ah, the joys (?) of 10Base2

We had a really weird intermittent problem, did the usual splitting of the chain, eventually finding the problem near the start. A user decided that having 2 coaxes and a tee piece attached to his PC was untidy, so he obtained a length of coax and a male to male adaptor, and dropped the tee piece to the floor behind his desk.

[Anonymous Anti-ANC South African Coward]

Ahhh, 10base-2 networks and Hunt-the-Terminator... one of my favourite pastimes when at a new client with a wonky 10base-2 network...

Luckily said Terminator was not a deadly one, hence the icon for this post.

I do miss those fun days, those times was a lot better than today's fun and games with Billware crap.

[Anonymous Custard]

Luckily said Terminator was not a deadly one, hence the icon for this post.

Oh I don't know. From what I can recall, they could be quite capable of killing the network stone dead with ease if mishandled.

[Joe W]

...or if you put a bunch in a sock and make an impromptu LART (luser attitude readjustment tool).

[blu3b3rry]

"Oh yeah, I just push it back in when it falls out"

Ugh. I think I've lost count of the number of times I've heard that phrase when trying to diagnose wiring issues on our kit.

That and the hallowed words "It just broke."

I've probably mentioned before about trying to diagnose a dead monitor (and a very, very expensive one it was too - about £8k at the time) only to find someone had fitted the DVI connector into it backwards.

[PB90210]

Re: "Oh yeah, I just push it back in when it falls out"

Had a job to provide a couple of hundred routerts that came with line cards fitted with RJ45 connectors that needed to be connected to BNC presented line. They tried to save money and outsourced the cables to a 3rd party rather than paying Cisco's rates. Unfortunately they were crap because some BNCs were not crimped properly, so after a few install failures we had fun with the rest of the cables, tugging on the BNCs... if they came away then the cable was faulty and was binned.

(it was a toss up between getting them returned to be fixed or replaced but the PM decided it was easier and safer to get replacements from another supplier)

[Zoopy]

Re: "Oh yeah, I just push it back in when it falls out"

> someone had fitted the DVI connector into it backwards.

With what, an industrial metal press???

[Giles C]

Re: "Oh yeah, I just push it back in when it falls out"

Never underestimate the ingenuity or determination to just get the cable to fit….

All of us have seen provably seen some abused connector and a user saying no nobody has touched it….

Bent pins, bent housings, completely broken connectors used to be a regular occurrence.

USB-b into Ethernet has been seen before…

Why do you think usbc is a orientation agnostic connector.

[Anonymous Coward]

Re: "Oh yeah, I just push it back in when it falls out"

Try an RS232 25-pin D plug upside down... he bent the D-shaped shroud to fit and bent or broke every single pin as none of them lined up with a socket

[blu3b3rry]

Re: "Oh yeah, I just push it back in when it falls out"

I can only assume someone had fitted it with the aid of a hammer - the only way to extract it was a very hard yank with my Leatherman! The pins inside the connector had been effectively mashed together, creating a single row.

I've seen similar done with RJ45 ethernet connections, too.

[phuzz]

Re: "Oh yeah, I just push it back in when it falls out"

I used to work at a PC builder, and one day the guy responsible for putting CPU's and RAM into the crate for each build order, ran out of the little antistatic boxes for putting CPUs in, so decided to help out by just installing them straight into the motherboards. Alas, he'd never install an Athlon 64 before, and didn't realise that the arrow on the CPU had to face away from the leaver on the socket (the opposite of most CPUs at the time).

He got through twenty machines, all of which went down the line and had coolers jammed on top, before the first ones hit QA and were found to not power on. Every single CPU had a bunch of mangled pins from being inserted the wrong way around.

IIRC we managed to salvage over half, by carefully bending the pins back with a knife blade.

[Prst. V.Jeltz]

I'm just about old enough to remember BNC networking being a ring that could not be broken lest the whole network vanish.

Seems incredibly fragile way to operate through modern eyes. Then again giving each machine its own wire must have seemed like a ridiculous proposition back then.

Odd though , cars are going the opposite way these days with their CANBUS . Used to be everything had its own wiring , now all the bits hang off the one (or 2 or 3 ) network line , which seems simpler tbf .

When it breaks only one car stops though , not the entire workforce

[Richard 12]

Depends who's behind, though!

To be fair, CANBUS is very reliable and tends to show warning of degradation before it actually fails.

The really important runs are also very short.

[ascii bandit]

Otherwise it would have been cantbus

... Badumchi

[The Organ Grinder's Monkey]

"Tbf canbus is very reliable"

Hmm. It can be, but whether it actually is or not is very much in the hands of the designers that deploy it.

I'd got off the spanners by the time that canbus filtered down to the Saabs that I was working on exclusively at that end of my motor trade career, but the people in the workshop spent a lot of time doing warranty replacements of complete wiring looms on new & nearly new cars for faults that were designed in by halfwit designers that had never worked on a car in their lives.

Mostly that was speccing wire & connectors entirely on the basis of the current that they needed to carry (almost none for canbus data lines) without allowing for the effects of moisture, vibration & temperature variation that are a fundamental fact of life for any car component.

Far too often the poor sod in the workshop tasked with finding the latest novel electrical fault would take the suspect connector apart & find that one or more of the microscopically-thin pins connecting a data line together had ceased to exist. (Yes, really, not just corroded but essentially vaporised.)

The incentive for all this was weight & cost saving of course. The wiring loom for a full-spec pre-canbus car like a 9000 Griffin would completely fill the largest size of domestic UK wheelie bin & leave a suitcase-sized pile of wiring on top. The loom for an equivalently-specced canbus car would fit entirely in the suitcase, so the saving wasn't trivial.

(Props to autocorrect for trying to insert "cancanbus" into this post, but only at the last instance of canbus for some reason.)

[42656e4d203239]

>>BNC networking being a ring

Nope - or rather nope when it comes to Thinwire Ethernet (as the marketing used to call 10Base2) - that was a washing line with 50Ω terminator at each end; each machine had a T-piece to connect.

The only ring I can remember is IBM's Token Ring but I don't know if that was actually a ring (two wasn't it? really don't know - I never thad the misfortune) or a logical one...

[jake]

Yes, TokenRing has a ring topology, both logical and physical. So does FDDI.

[Heimdall27]

Logical ring, but ultimately physical star. You'd run cables from the computer to an MAU/MSAU/CAU/LAM (Token Ring) or a concentrator (FDDI) in the wiring closet. Topologically it was almost identical to a modern wired Ethernet over Category X structured cabling system, certainly in second generation equipment. For resilience you'd run a pair of root concentrators, so you'd have two stars, with each device connected to both.

[Apocalypso - a cheery end to the world]

> Logical ring, but ultimately physical star.

You've reminded me of a former client, a power company servicing an industrial estate with 15-20 odd substations. The control and monitoring network was designed as a loop so that if there was a break in the cable (i.e. a man in digger doing what he does best) then packets would still flow the other way around the loop.

Until we found that the loop actually ran around the wall of a single room on the substation more or less in the middle of the estate, with a star pattern "out and back" pair in a single duct to each substation. D'oh.

[Andy A]

One place I worked had a ring main connecting the various buildings with power. Substations as appropriate, with Big Substations at each end of the site. Two rows of pylons connected the site to the National Grid. They actually thought about resiliency in 1940.

One of my jobs during the annual maintenance week was to reduce the IT load to a minimum (just in case) while the sparks tested the changeover to the "other" set of pylons. I was invited to watch once. Those big breakers are IMPRESSIVE!

[Doctor Syntax]

Obligatory Dilbert: https://3.bp.blogspot.com/-4CIb4Gw7WtQ/Wvb2HbKBdtI/AAAAAAAA1LU/hUiOInPurPIQtj5nsI41c9sInOz27LiBwCLcBGAs/s1600/TR1.gif

[ITMA]

Why has nobody (that I've seen) mentiond Cambridge Ring?

Oh, and there is Peter's post right under my post LOL

[Zoopy]

I'm sorry ITMA, due to the pun quality we've decided to eliminate your position.

[ITMA]

Oh no you haven't ;)

[Peter Gathercole]

FDDI was a ring (actually two, running in the opposite direction from each other to allow for a single break without totally breaking the network).

And of course there was the Cambridge Ring.

Talking about rings, IBM SSA disk connections were rings as well.

[Andy The Hat]

Stealing terminators could be avoided with a bit of heatshrink.

I had a fully working mixed 10base2/10base5 line - except for three unconnectable machines in a class room (often the machines would move along the line so one day it's be number 7, 8 and 9, the next it'd be number 8, 9 and 10 or no issues at all). Done my head in until I found a bad connection under a secretary's desk (proper IT crowd moment ...) which was causing a reflection and killed the signal for about 6m, 50m or more back up the line ... Depending on temperature, activity or what it felt like on the day, the dead spot wandered a bit so either killed machine 7 or 10 but never both.

[jake]

"BNC networking being a ring"

A bus, not a ring.

"Then again giving each machine its own wire must have seemed like a ridiculous proposition back then."

ARCnet existed before Thinnet.

[Phil O'Sophical]

ARCnet existed before Thinnet.

And woe betide the network admin who didn't throw away every bit of 93Ω ARCnet cable when upgrading to 50Ω 10base2. I spent the best part of a week trying to diagnose a problem between two clustered μVAXen, which would work fine until network traffic reached a certain peak, at which point collision numbers would go through the roof and everything stopped. The problem was eventually traced to a short leftover piece of ARCnet cable in an underfloor duct.

[Anonymous Coward]

Oooh, fancy, we only had the short run 75Ω variant, which was reasonably able to keep things connected.

The only issue I had on my first network ever was that I didn't have a server with quota and an office full of users who would fill up every bit of space available if you let them.

A bit of Turbo Pascal to write large files full of a single character fixed that - I had a directory which only I could access which basically absorbed a lot of space with files that would compress to pretty much nothing during the backup. If someone genuinely needed more space I just zapped one or more of them and presto, space.

Pragmatism: it works :)

[Dwarf]

Turd files

I used to do the exact same thing on Netware 3.x servers and NT 3.51 servers.

I had an amazing trick on these that allowed you to create the files nearly instantly - from the app perspective, but it flat lined the server for a bit, it went like this

Create the file, set the file pointer to XMb out, Write a 0, then close the file

The great thing was that sparse files didn't exist yet and when you fseek() to where you want and write one byte, the OS would handle all the rest of the IO for you.

The other great thing about the approach of filling the disk with "turd files" as I used to call them, is that if anything is running and is running away, when the disk is full, it will generally abort, sometimes deleting its huge file as it goes.

You can them simply delete a turd file, to give yourself free space and get the company working again, whilst you go and find out which user is creating massive files and go for a quiet word.

[Erix]

ARCnet? Wasn't it this special little metal thingy that Griff gave the Men in Black that they had to get on top of that rocket to prevent an alien invasion?

Definitely existed before 10Base2.

[Dwarf]

I think I've still got an Arcnet hub in my roof somewhere, along with a bunch of old adapters. They are still doing useful work as insulation.

They used 75R cable too iif I recall correctly. Star network out from the switch, no terminators necessary, but slow ....

[Anonymous Coward]

That's what the discussion above is about, if I recall correctly. It was 75Ω for short runs, but if you needed longer runs you'd have to deploy 93Ω gear.

[cosmodrome]

That was token ring. BNC based ethernet just needed to be terminated with 50Ω resistors on both ends, no ring required (or allowed). The BNC connector and coax cable just was the state of the art connection for high frequencies. It still is, btw, in many cases as BNC still is quite alive everywhere where you're dealing with word clocks (digital audio) oscilloscopes etc.

[Anonymous Coward]

It was just annoying to switch from ARCnet to ethernet because of managers who didn't understand that a 50Ω cable may LOOK the same as a 75Ω one but isn't and thus didn't understand why wire had to be replaced with identical one/

Eventually we told them that network wire wears out after a while, somehow that seemed to work..

[Aladdin Sane]

This was back when everybody had desktops and the machines rarely if ever moved location, and if they did it was by IT guys. Now that everybody works on laptops it's a very different proposition.

[Doctor Syntax]

"Seems incredibly fragile way to operate through modern eyes."

Well, nobody ever accused the previous generation of Ethernet of being fragile. It was about the size of a garden hose but less flexible. It was accesses with vampire taps which had to be at least 1.5 metres apart and the cable was marked with suitable points. There was, IIRC a 15 core ribbon between the receiver clamped at the tap and the card in the computer that handled the rest of the interface.

Fortunately it wasn't my team that handled the thicknet connected stuff.

[Phil O'Sophical]

> It was about the size of a garden hose but less flexible.

It had a minimum bend radius of around 25cm, IIRC. Didn't stop people trying to pull it into standard conduits, though...

[Doctor Syntax]

I thought it was more like a metre. You'd probably need a pipe bending jig to get it tighter than that.

[Roland6]

You are confusing it with broadband networking (IEEE 802.4 10Broad36) where you did need "plumbers" tools such as pipe bending kit as the backbone "cable" had a solid aluminuim shield making it more of a pipe..

[Roland6]

>Then again giving each machine its own wire must have seemed like a ridiculous proposition back then.

With 10Base5 Ethenet each machine had it's own cable to a transceiver attached to a common cable and each machine could be connected or disconnected without breaking the network. 10Base2 was sold as "cheapernet" as it both reduced costs and simplified matters by effectively putting the transceiver on to the network adaptor and replacing the "yellow peril" by cabling and components already widely used in other applications such as radio and lab equipment.

It took some development work to achieve 10Mbps Cat3(*) (IEEE 802.3 10Base-T ratified in 1991). So no it didn't seem like a ridiculous proposition back then, just something we didn't have off-the-shelf technology to achieve.

(*) US telephone twisted pair.

[Herby]

Ring??

If you did the 10BASE2 up in a ring, it would fail. It would fail BAD. Reflections and all that. It needed to be a line with terminators at each end.

Of course then there was IBM's thing: Token ring. Big unwieldy cables that wouldn't bend easily, MAU things. Ring adapters. Generally a mess. It also had funny connectors that were "sexless". It ran at 4MHz, and 16MHz, and got outclassed by 100BASE-T (100MHz Ethernet). We haven't looked back.

[Andy A]

Unfortunately vehicle manufacturers seem to think that CANBUS is not a bus, and they wire everything in a star configuration. Every control unit (and there are dozens) has connectors with loads of wires.

I remember an article in New Scientist in the 80s where someone had designed a car wiring scheme with a proper bus, and a control system which sent messages such as "left front indicator, turn yourself ON". Nobody wanted to implement it because you would need a network analyser to diagnose faults. Now every workshop has a raft of network analysers, each for a different brand (or even model) of car.

[The Organ Grinder's Monkey]

Further to my gentle rant above about automotive canbus (it's in domestic boilers now, too, apparently) I remember a Saab 9.3 which had had a lot of the customer's money spent on it trying to cure a combo of spurious ABS captions & random stalling. The original dealers had just done what they always do, & thrown a lot of (the customer's) money at it by randomly changing most of the control units in the car, & then pronounced the fault incurable.

We weren't a dealership, but a non-franchised specialist, & kept a yard-ful of wrecked cars that we could borrow parts from to trial-&-error test theories on such problem cars without having to charge the customer 3 or 4 figure sums each time for new components.

Long story short, it was the 6-disc CD changer in the boot that was the cause of both faults. We never really worked out how or why, but replacing it completely resolved all it's problems. Later chats with someone at Saab who had detailed knowledge of the canbus tech suggested that there were several tiers to the canbus implementation in the car, & that there was a hierarchy to the devices on the network. Evidently the CD changer had been given greater status than it needed & when it faulted it provoked a panic response from the central controller rather than being ignored & simply logging a code. Note that the CD changer worked perfectly throughout, whatever the fault it had was, it wasn't a functional one.

[Prst. V.Jeltz]

I threw a BNC connector away last week in a mass "This wire will come in one day" reality check

I had to be harsh! I Got 4 plastic crates down to one, amongst the ejected was my collection of every spare wall plug dc transformer thingies I ever owned with all voltages covered .

The BNC was an early VCR connector rather than networking

[Anonymous Custard]

And if course is now the one you'll have an urgent need for next week...

[Doctor Syntax]

The trick is not to throw all of them away. Retain just enough to ensure you'll never need them again.

[jake]

"Retain just enough to ensure you'll never need them again."

I still have a couple contracts that use 8" floppy disks and paper/mylar punch tape on a daily basis.

So how long, exactly, is long enough?

_{Sometimes being a packrat helps pay the bills. You'd be surprised what the exactly correct 386 or RAM chip can bring occasionally ...}

[Anonymous Coward]

Nah, that won't work either.

If you save n connectors it's a pretty safe bet the next time you need them your need will be n + 1. Murphy's Law and all that.

:)

[Roland6]

The trouble is gauging what is sufficient. Too many times, I've reduced my stock keeping "just enough" only when I found a need to discover I hadn't kept enough.

[Anonymous Coward]

If it was a VCR one then it should have been 75 ohm. There is a subtle difference between the 50 and 75 ohm connectors, imperceptible to the untrained eye, but whereby the 50 ohm pin* can, err, stretch the 75 ohm socket making it unreliable.

* I don't dare genderise connectors these days...

[JackHarveyCan]

Early video monitor networking(late 70’s, 80’s), which was a multiplexed signal passing through daisy-chained BNC video monitors, was accomplished using 75 and 50ohm cabling, wherein the very last(terminating) connection was created using a low impedance cable, and all connections between the other monitors would use high impedance cables. If the text hasn’t rubbed off the cable, they’re usually labeled, but a quick glance at the pin or quick measurement with a multimeter will do. Very important to check if you’re planning on running any power through it, transmitting for example, as you’ll likely damage your equipment if the SWR is too far off.

[Phil O'Sophical]

The BNC was an early VCR connector rather than networking

It long predates VCRs. It's a smaller (Baby) connector derived from the wartime radio "N" connector, for military antenna cables. It's still in wide use today because both the 50ohm and 75ohm variants have constant impedance beyond 1GHz, so are very popular for professional radio installations (as are "N" connectors, for lower-loss applications).

[Doctor Syntax]

You learn something every day. I'd always thought it must be a contraction of bayonet connector.

[Heimdall27]

BNC is Baby Neill Constant. Paul Neill of Bell Labs was the inventor (N type, BNC and TNC, among others), Baby because it was a smaller variant of the N type, Constant because it had constant impedance above 1GHz.

[Herby]

Connector types

the B in BNC stands for Bayonet, the T in TNC stands for Threaded. Of course this isn't what Wikipedia says.

Wang Net used two 75 ohm cables. One for transmit, one for receive, and a head end that did the amplification. It used two different types of connectors so they wouldn't get mixed up and kill everything. I believe it used RF style things, not baseband (that's what the BASE in 10BASE-T stands for) like Ethernet.

[Anonymous Coward]

Re: Connector types

From a thread on Reddit: "In 1941, the US Navy used a smaller version of the threaded Type N connector, the Type BN (Baby N), as the UG-85/U, UG-86/U, UG-114/U and UG-115/U. In 1943, the British introduced a 1/4 inch 50 ohm coaxial cable. Immediately, companies developed many connectors for it. In 1944, the US Navy called the designers together to standardize on a single standard design acceptable to all. Paul Neill developed a prototype connector with a constant 50 Ω impedance beyond 1 GHz, a bayonet fastening (faster than threaded), and easier to manufacture and assemble. The Bell Laboratory drawing ESL 662916, dated March 2, 1944, specified it. In April 1944, the Navy took Neill’s design, added a ‘C’ to BN (for constant impedance) to get Type BNC ( Baby Neil Constant), and assigned part numbers UG 88/U to the male and UG 89/U to the female (Amphenol still makes them). The Navy issued a final drawing RE 49F 246, dated December 2, 1944.

The picture [not included here] is from Salati's patent application from 1945, who patented the BNC after the Navy had them. A court invalidated his patent in 1958. The court records are the source for the above information."

[ChrisElvidge]

I first came across BNC as an oscilloscope connector

[Sam not the Viking]

Not quite terminal

On slightly larger cables than coaxial, our customer complained that the cables to the three-phase motor were "running red-hot". These were not small cables, four-off, four-core, 70 mm2 connected in parallel. On checking the installation, our engineer found that one of the cables was indeed very hot. Opening up the local junction box, looking for loose connections, all looked well until he gave the cables a tug and some of the cores just pulled away. Three cables had been fitted with the correct terminal lugs, but not been crimped-up so were not conducting properly. Most of the power was running through a single cable. Although crimping the terminals corrected the immediate problem, the high temperatures caused other damage which resulted in considerable expense.

Although the issue was caused by the customer's contractor, it's always the equipment manufacturer who gets the blame.

[Mishak]

Re: Not quite terminal

That's one of the reasons I use a thermal imaging camera - makes it really easy to spot where there are bad connections.

I used to work on high power motors, and once spotted a wire swing out of a test rig and leave a trail of molten metal behind it - a 10mm steel bolt had been carrying about 500 amps between two cables when it was supposed to have been keeping them in direct contact. Those rigs used to loose a few kW of power in the connections even when they were tight...

[CountCadaver]

Re: Not quite terminal

How many megawatts / gigawatts are we apparently talking here?

[Maximus Decimus Meridius]

Re: Not quite terminal

1.21 Jigawatts

[Mishak]

Re: Not quite terminal

It was running at 500 amps, so a resistance of 10 mR creates 2.5 kW.

The specific heat capacity of steel is ~500 J/kg·K, and it takes another 275 kJ/kg to melt at ~1400 C.

Assuming the bolt was 50 g, then it would take ~14 s to get to the melting point, plus about 6 s to then melt (all of it, but it would fail before that happens).

[42656e4d203239]

Re: Not quite terminal

We had a similar incident, only it was an actual fire at the three phase incoming termination from the grid.

Someone had used an incorrect crimp and, becasue the incoming cable had been stretched to fit (or cut off really close to the exact length required), had, instead of faffing around getting the incorrect crimp off, just drilled the fixing bolt hole out a bit so the larger (correct sized) fixing bolts would fit.

The load in the building was particularly high one day, becasue of reasons, and with a high ambient temp the insulation started to get very warm and started letting out the magic smoke... Much fun was had by all and many curses uttered towards the installer by our, by then, tame sparks.

Oddly he was an apprentice on the job when the building was built and knew who installed it though he never let on (I always suspected it might have been he who made the error of judgement)

[ArguablyShrugs]

Re: Not quite terminal

Nothing like that much power, but could have made a nasty fire nevertheless – I'd been just clacking on the keyboard when my whole apartment's lights went out, along with the UPS chiming in.

In the building, all the master switches are in a "locked" (square key) box outside in the hallway. Out I come raging with fury, and indeed, there is a sparky meddling with the box.

"Did you just turn my whole apartment off without any warning?!?"

"No, I am just installing a new meter for your neighbours, never even touched your switch"

Turns out some previous sparky from a different provider forgot to properly tighten the 400/230V three‑phase wires at my meter. Just loose enough that they sparkled when he bumped the box installing the other meter. Probably a good thing, as it could have been quite a ticking fire bomb otherwise, as it must have been loose for years.

Needless to say, the good sparky tightened the mess left by the bad sparky properly and I gave him a beer.

[Doctor Syntax]

Re: Not quite terminal

Just been clacking at my keyboard reading this thread when we had an hour long power cut. I think there's some spooky quantum entanglement going on here.

[PRR]

Re: Not quite terminal

> reading this thread when we had an hour long power cut. I think there's some spooky quantum entanglement

Going both ways. Ask Schrödinger's cat. The day -before- this was posted, all the power on nearby Tourist Island went out. For a couple minutes. Power company says "transmission fault" but is coy about the exact trouble. FWIW, that transmission line runs past my town but we never blinked. Prolly loose wires everywhere "Not our fault man!"

[Mishak]

Similar experience

The voltage to my house was regularly going over the statutory maximum of 253v, causing my UPS to go into "trim" mode.

After about 18 months of pestering the supply company, they came round to install a data logger to see if I was right.

Guy told me he would need to switch off the power to connect his kit and asked if I needed to shut anything down first. I did, but as I went back in to do so, the power went off.

"Not me", he said, showing the wire that had just fallen out of the meter when a brushed it with his hand.

Turns out the metering company didn't do a very good job when they replaced the old meter with a "smart" one.

Again, that could have caused a fire. Luckily, there were no long duration, high current appliance in that house.

[PRR]

Re: Similar experience

> 253v, ....to install a data logger ....replaced the old meter with a "smart" one.

Gosh, I'd think any "recent' "Smart meter" would know what voltage it saw and be able to report that back to HQ in near-real-time.

Is an issue here b/c we run AT 250VAC near all the time.

[Mishak]

250 is "ok"

The nominal supply voltage in the UK is 230 V, with a permissible variation of -6% to +10% (~216 V to 253 V).

The unsymmetrical tolerance was introduced to allow the UK and EU to "harmonise" and adopt the same supply voltage of 230 V when they were using 220 V and 240 V. Of course, no country has actually changed the voltage it uses.

This can have an impact on devices that were optimised to work at 220 V - for example, incandescent lamps that were designed for 220 V use in Europe will have a much shorter service life if they are used on a 253 V supply in the UK.

[The Organ Grinder's Monkey]

Re: 250 is "ok"

From the "Tourist Island" reference, iirc PRR must be "way down under Australia, (very different from over here, all the animals look so strange...)" so possibly has different nominal voltage to us up here in the UK?

Edit, misremembered, was thinking of Magnetic Island, so I'm probably wrong.

[Mage]

Re: report that back to HQ

I was told the ones in Ireland don't report voltage. They also use GSM, not 4G.

[Stratman]

Re: Not quite terminal

Not strictly on topic, but 'interesting' nevertheless.

Many moons ago a colleague always carried a modified 13A mains plug in his staying-away-in-a-hotel kit. The fuse was replaced by the thickest copper wire he could fit in it, the same wire joined L and N. When he was kept awake by a television blasting at full volume because the occupant of a nearby room had fallen asleep, he'd plug it in and take out the breaker giving him a good night's sleep.

[Mishak]

Much safer...

To use a plug-sized RCD tester!

[GlenP]

Been there...

In one role we had a 10-base2 network but they had installed the make-before-break wall sockets which reduced the problems. One afternoon the network got decidedly flaky, nothing definite but there were intermittent disconnections and slow running.

Rather than cause more problems I told everyone in the main office to leave their computers on at the end of the day and then started fault finding. Disconnecting the patch cables to all the computers cured the problem so it was then a case of plug each one back in while soak-testing the network until I found the faulty connection; it turned out to be a failing network card* rather than the cable which slightly surprised me. Fortunately I had spares so a quick swap and all was well.

*This was back in the day when computers had discrete network cards (NE-2000 based mainly), not integral to the mother board.

[Anonymous Coward]

Re: Been there...

Still prefer discrete network cards even if the motherboard has network ports.

I have had integrated network ports go 'flaky' which is a pain.

I tend to leave the integrated ports as emergency backup if a network card fails and I don't have a card to replace it.

(I do tend to have spares so have probably only had to use the integrated network port once or twice for half a day.)

:)

[Roland6]

Re: Been there...

I've also use discrete USB PCI cards on servers where high capacity backup devices etc. will be attached. An advantage is that these cards tend to be significantly more performant, although it is strongly advisable to use powered USB devices...

[David Hicklin]

Re: Been there...

Ah those were the days of choosing the DMA and Interrupt channel settings, often with DIP switches, and if memory serves me right the shared upper memory region.

[IGotOut]

Ahh "the good old days"

First up in impressed they were so advanced using TCP/IP rather than IPX/SPX .

But boy do I remember those days. My first IT job was desktop support and what an utter mishmash of systems that was.

It was the "switch over" days, so the estate consisted of Novell servers running IPX/SPX, NT servers using TCP/IP and IPX, desktops running Win 3.11 and 95, Wyse terminals, 10base2, 10baseT and 100baseT, Token ring 16/4, Baystack switches, ShivaRas servers and not to mention Compaq desktops that would happily slice a finger off if you dared try to remove a network card.

Those couple of years probably taught me more about trouble shooting than the other 20 odd years, not just in IT, but life in general.

[Dwarf]

Re: Ahh "the good old days"

But did you run 802.2, 802.3 and SNAP on the same IPX network ?

[Random42]

Crimp pin

I lost cound of the number of BNC connections I remade after my boss/predecessor failed to push the crimp pin in to the beyonet fitting far enough to get the click... happy days....

[TimMaher]

About George.

Surely they should have been working on ICL mainframes?

[RMclan]

Pretty sure I've posted this before. Working as IT Manager for some holiday resorts in Spain. I was in Tenerife and due to visit a resort near Malaga at the weekend.

Got a call to say nobody at the mainland resort could access anything on the network (Novell 3.12 server, 10base2 cabling). I asked the nearest person I had to an on-site assitant if anything had changed, computers moved, etc.and was reliably informed nothing had.

I arrived on the Friday evening and after a bit of troubleshooting moving terminators around I isolated the problem to the locked accounts office. I wa sable to bypass the office and get the rest of the network working with some spare cables I had.

Monday morning one of the 2 sisters who worked in accounts turned up and unlocked the office. The other sister's computer was missing, and so was the T-piece which connected it to the network. Apparently the one sister had gone off on maternity leave and taken her computer with her so she could do some work from home.

[KittenHuffer]

The pregnant pause waiting for the network to come back up would have been quite significant!

[Anonymous Coward]

HR is the one thing that is absolutely ripe to be wiped out by a chatbot.

[Anonymous Coward]

I sure hope not.

"Your email reply uses the word "the" a great deal. This word shows up in 99% of messages deemed offensive. You have therefore been fired for sending offensive messages."

"Can I talk to a supervisor?"

"Supervisor chatbot here. I agree with my subordinate."

[Roland6]

SCO Unix...

In use I found SCO Unix to be a reasonable PC Unix. Looking back I do wonder just how much the post breakup AT&T killed Unix through its approach to the commercialisation of Unix and its licencing, laying the ground for the subsequent protracted SCO Group litigation.

[Doctor Syntax]

Re: SCO Unix...

AT&T had little option - it was a regulatory issue. Killing SCO? Imagine somebody with two competitors. Offer to hold the coat for one of them while they slug it out.

[Roland6]

Re: SCO Unix...

AT&T most certainly had the option, they decided to try and bolt the stable door, by having a legal fight with UoC Berkeley and BSDi - instead of buying them out, which lead directly to MINUX which in turn gave rise to LINUX...

Killing SCO Group was only really necessary because AT&T and Novell hadn't open sourced Unix and been sensible about trademarks, giving the opening for SCO Group...

At least with Linux it seems there will be no opening for a SCO Group style litigation...

[JohnnyS777]

Fun with coax!

Ah, 10Base2. The memories, the pain, the triggering, the PTSD...

[Daniel B.]

10Base2

I got real bad memories of coax LAN from the early 90s. That story about the fallen cable? That was super common back in the day. Check all the PCs to find where the hell did one of those connectors have a cable slip out. ARRRRGH

[David Hicklin]

Re: 10Base2

Worst case was when they did an office desk shuffle in a room where the power and network connections were in fixed floor mounted boxes.

[martinusher]

What's with the BNC hate?

BNC connectors might be "1940s era" connectors but they're still in widespread use. Just not for networks. The most common place to find them is testgear but they're also widely used on lower power radios where the case is too big for a PL259 'UHF' connector.

I've had a few run-ins with 10Base2. In the early days of networking it was a lot more convenient than the triaxial cable used with early Ethernet. This coaxial cable was tapped using a device that clamped over the cable and inserted a probe that connected to the three layers. It included the network PHY -- the physical layer interface -- which interfaced to the computer using a multicore cable. Expensive, inflexible and a pain to work with. Other offerings included Token Ring connectors which looked as if they had been designed for use in a WW2 bomber. The only simpler alternatives were the 75 ohm TV antenna cable with 'F' type connectors used by PC-NET or 50 ohm using BNC connectors, somewhat better and potentially more reliable if you didn't mess with it. Networking using phone cable such as Starlink wasn't fast or particularly well developed.

People forget that in order to get where we are today there had to be a fair amount of engineering iteration -- after all, back in the 1980s the bulk of 'communications' was still through external modems connected through DB-25 connectors.

[TeeCee]

Re: What's with the BNC hate?

The token ring plugs for Type 1 and Type 6 (??) had the unusual feature of being unisex (i.e. there were no "sockets" only "plugs") Even the wallports were just a bit of plastic with a square hole in it, into the back of which you stuck one of those plugs.

They were also screwless, for (alleged) ease of assembly. The only trouble being that with the metal bits, the two halves of the plastic shell, the endplate and the blanking roundels for the holes you weren't using in the plug, you really needed all the fingers of both hands to keep it all in place while your third one squeezed it together...

[NetMage]

Re: What's with the BNC hate?

I have a brand new in box DEC vampire tap left over from a few years before I started from the original 10BASE5 VAX network that were the first computers installed there.

[Roland6]

Re: What's with the BNC hate?

Might be worth asking TNMOC @ Bletchley Park, if they want to give it a home; I think this has long past the point where it might come in useful…

In fact if you have some yellow peril etc. they might accept the donation.

[zb42]

Long ago, when coax ethernet was already considered obsolete, I started diagnosing a cable run through a few different rooms in an office building by checking resistance with my multimeter and found a couple of ohms short circuit.

A person had stapled a run of cable to the wall along a corridor, with staples through the middle of the coax.

[Anonymous Anti-ANC South African Coward]

Here's a Cat5 o' nine tails... have fun!

[JackHarveyCan]

All hail the one Token Ring

The advantage of 10Base2 coax/BNC networking is strictly in its simplicity, requiring only one shielded cable as opposed to twisted pair; at the time, standard 10BaseT RJ45 network cables only provided 10MB/s, which was equivalent to the capabilities of earlier token ring networks as well as the “newer” 10Base2. Token ring required special hardware and processing; Novell and StarLAN were popular providers of both, although 10Base2 would later simplify the required infrastructure.

[Anonymous Coward]

Re: All hail the one Token Ring

That's 10 *Mb/s*, not MB/s. Funnily enough, I have an Xbox 360 shaped AV switch that also includes a 10Base-T hub.

[John Tserkezis]

Back in the day, before TDR (Time Domain Reflectometry) where test kit was crazy expesive, and without a signal generator and CRO, the only way to diagnose cable breaks was to do it manually - we got down on our hands and knees to man-handle each connector to find out which connection was fault.

I was at a client who had this type of fault, the fault took a large part of the day to nut out.

Happily, I finally found the dodgy connector, and just finished repairs, when the user mentioned his knowledge of the fault - where he used to kick the cables to bring the network up again.

The user would have been helpful if he mentioned this earlier - but, there you go.

[Excused Boots]

Although, on the other hand, you could now legitimately charge said client for a full days’ attendance on site to ‘fix’ the issue, rather than just an hour or so!

[Anonymous John]

"Wherever George was sent, he strung up cables and connected devices using BNC connectors, a 1940s-vintage connector that somehow survived into the early LAN age. "

That's nothing. We're still using 1/4 inch jack plugs (6.35mm), and they were invented in the 1870s.

Icon as he was alive then.

[Mage]

Re: invented in the 1870s

The 4mm plugs are mere children, only dating from about mid 1920s and still used. The approximately 1/8″ wander plug must be older, but it's gone out of use. It was used on pre-WWII battery radios and up to 1960s for aerial, earth etc on radios.

The 75 ohm BNC is now rare compared to 50 Ohm. The 75 Ohm N is still used on some specialist cable broadband gear. You can connect an N plug to BNC plug using a "barrel". The inner bits of N are similar. TNC seems very rare now.