Developer waits two years for management to define project

Electrical noise causing crashes used to be a bigger problem than it appears to be now. There was the electric pump that started up after someone flushed the toilet. The electric kettle when someone made a cup of tea. All indicated by throwaway remarks by people in general conversation.

One long computer program run came at the end of a weekend overtime shift. As it started its final phase's print the shift leader started to power off the unneeded peripherals. When he did that to the papertape reader - the system crashed. There was a different, but similar, human failure in the next attempt. Eventually everyone left the computer room until the job finished so they could finally power down and go home many hours later than expected.

An intermittent problem of program and system crashes on a mainframe was investigated during the Saturday's whole day of engineering maintenance. It was quickly noticed that the problem occurred when an engineer closed a tape deck door during his maintenance. That was the point when the tape automatically loaded. You could have the mainframe stopped - and watch the engineer's panel logic monitoring lights change as a door was repeatedly opened and closed.

This was reported to the chief engineer - but a few minutes later it could not be reproduced. What had changed? The answer was that a floor tile had been lifted as part of the commissioning of a new tape deck. The tape deck cables each had a metal box partway along them - and the latest one of these of these was sitting on top of the bundle of cables under the lifted tile. The floor tile was reinforced by a bare steel plate on its underside. The false floor supports were not the usual corner posts - but a matrix of horizontal steel bars. This matrix was connected to the building's "dirty earth". Putting the floor tile back recreated the problem as the slightly protruding metal box was shorting the computer "clean earth" to the building's earth.

A university mainframe was being run through its final formal acceptance trials. It kept crashing at near enough exactly the same place every time - while thrashing the high speed papertape reader. After several failures someone heard a coincidental "click". The large tape collecting basket was metal - but not earthed - and was being charged by the fast moving papertape spooling into it. The static discharge spark was responsible for the mainframe crashing.

Crash Pad

At a former job we had a keyfob and access code to a building security system operated and maintained by the usual outsourced security provider of last resort popular with industry and olympics. The security cameras didn't really work, logging was "disabled" - probably never worked at all or the dum-dum's could not be arsed to fix it - the best part: The four-same-digit login was never changed and they did not track the number of key-fobs, if you lost one, you just got a spare. Some PA would ask occasionally, but, since we didn't sign for it, one could just say that "I returned that already".

Now, at the top floor of the premises was a large recreational area with a well-stocked bar, cooking area, shower, toilet, sofas, fat-boys, games, music and beds.

This was practical, because if one needed a place to crash after a heavy night out - or - a diversion flat for the girlfriend-of-the-day - one could just pop in "after work".

Indeed, this was possible for many *years* after having quit working there. Sometimes one would go to crash, meet former colleagues by accident and have a drink and a smoke.

"The bars' relay switched .. let loose an EMP ..and scrambled the CPU's registers."

Something like this happened by purest coincidence on the first industrial job I ever did, i.e. a relay wired directly to a solenoid valve. Ever after, I developed an obsession with zero crossing AC relays, trapezoidal waveform DC drivers, and eventually found myself being paid by HMG to go to meetings on EMC.

If the klutz who designed the electrical hardware on my first job had been more competent, I'd have missed out on some fun trips to exotic places.

Re: Management...

"70% of the time passes before any development is even planned."

OTOH any developer worth his salt should be able to work out what they need fairly quickly and get on with that. The last 30% of the time can then be spent on working out how to persuade them that that's what they said they wanted.

Re: Management...

Plus, when management do finally make up their mind regarding their requirements, and you deliver the project, they then start demanding to know where all the features that they never asked for were...

Re: Management...

Management and sales are two sides of the same valueless coin.

I have frequently sat in meetings and listened to them swap buzzwords and gradually worked out something that is relatively easy to implement and goes a long way to solving a lot of problems. Then the shit hits the fan when, having been painted into a corner and having some software that actually solves 90 percent of the clients needs, it becomes apparent that the clients needs are secondary to sales and managements needs to obfuscate things.

I wrote a client facing web app that allowed customers to see all their orders and buy all our products that was largely scrapped because a couple of our major customers would have found out they had been overcharged by 3 or 400% for a basic product. The sales and management teams of course got huge bonuses as a result of overcharging and weren't convinced they wouldnt be out of pocket.

Re: Ghana?

Huh. Turns out it looks pretty good. Seriously considering that now.

Re: "While workers sorted out desks and networks and other niceties..."

I helped build a test platform in a DC in London, and there were only two small cages tucked away to one side where customer equipment could sit - the rest of the entire floor was empty - lots of football played with printer paper balled up and wrapped in tape ensued :)

Oh, and did I mention there was a pub about 10ft outside the door with a pool table and an excellent cook doing authentic Thai curries? Lovely :)

That is until M$ came along and bought the entire floor for xbox live :(

Re: "While workers sorted out desks and networks and other niceties..."

My last job was a bit like that... The previous sysadmin/netgod disappeared eighteen months earlier and the whole place had gone to crap IT-wise.

I turned up on my first day, and they said "Sorry, we haven't had time to fit your office out yet..."

Bare walls (by which I mean "Bare Brick"), threadbare, stained carpet, third-hand office furniture, and a PC that burst into flames within a week of me starting. Running out of the computer centre with a burning tower unit after the building was closed DID rather attract the attention of Mr Plod, who just happened to be driving through campus at the time !

A few days later, I also ended up with an insect infestation as a bonus... Flying ants in the office, bats in the roof cavity above me, and Lord of the Flies waiting out for me on the network.

(Bonus points: Piggy was already dead and the Conch had been smashed)

Re: My story is this...

"b) actually thinking that the deadline would still be applicable"

This appears to be quite common. Project starts, rough scope is set and rough plan and timelines agreed.

Then comes a couple of months of dicking around with getting the project in the system, changes to scope etc. etc. When it all finally kicks off they seem to think the original deadline still holds.

Now we just quote a number of days from 'kick-off' (i.e. when all the money and scope is agreed and it's all signed in writing).

Re: My story is this...

Quite normal. I once had to do a project. The CEO (a total asshat) asked how long it would take. I said 6 months, conservatively. He said you have 3. I said no, it'll take 6. He said it WILL be done in 3.

Fast forward 2 months and they finally iron out the contracts and I'm allowed to start. And, of course, I now have 1 month to do it. At the end of the month I was off on holiday.

He told me I couldn't go. I basically said well I'm going, fire me when I get back if you want and off I went. I didn't get fired and he didn't get his project finished to his timescales either. In the end, with a lot of work and a very reduced set of testing, we got it done in about 5 months IIRC.

Re: My story is this...

I worked with what I would call sensible people for a long contract. At the start of it I was asked to come in early on one particular day each week. I agreed as I was sure I could get there on public transport. Near the end of the contract the company moved location further away. They were absolutely astonished when I said I didn't know if I could still get to the new office early and said 'but you agreed months ago to come in early' (to the old office) . I don't know what it is that turns these people into idiots who think you can either break the laws of physics or should be willing to turn your whole life upside down to the point it costs more to go to work than what they are willing to pay.

Re: My first real job

I learned early on that the "Look Busy Kit" (TM pending) is essential for places like this. It consists of a pile of blueprints or docs. What they actually are is irrelevant.. just have the pile and scatter it about the desk. Periodically re-shuffle the pile. Then assorted tech book cover in various sizes. Use these to cover the real book you are reading. Lastly, schedule several off-site meetings weekly with "colleagues".

OTOH, I had a boss early on who would tell me to grab my notepad, we were going to "engineering". We get up with notepads in hand, go out to his car, get in, and go fishing for several hours.

Re: project spec

Hey,

That's pretty OK for a project spec, its not a design you know.

Could be an issue if there is more than one instance of either of those but often the job involves getting different data formats from different platforms, mapping them, arguing for ages about exceptions and edge cases and who's responsibility they are etc.

I don't think I have ever worked somewhere, even in a software house, when I could simply code from the document I was given with no fear of gaps arising.

Re: Signal to noise

A colleague had a student holiday job at a large scientific computer establishment. His job was to run a plotting program. The plotter was a fixed speed device that started spewing paper as soon as it was switched on. All the computer did was jiggle the pens.

"Run" was the operative word - as the plotter and computer were at opposite ends of the large room. He had to start the plotter - then sprint back to the computer to output its results before the plotter ran out of paper.

vacuum & UPS

I had a project run on 2 Winchester disks (75+75MB, I think); the tape backup was delayed &delayed. I was told "back up 1 disk to the other". "what if there's a power-cut?" -- "you're on an uninterruptible power supply!"

Then the same technician plugged a vacuum cleaner 4 times running into that UPS, killed it. I inquired as ti why she was not shot at dawn.

Been there seen that. Total fuckwits.

Re: In the bowels of the Pentagon

Kewl! I knowing this, I would make sure to mail some magnets once in a while.