Re: One of my cousins ...
They do that with books too - you lend them a book and never see it again because they read it (sometimes) then bin it or give it away. And act all surprised that anyone would want a 'used' book back.
54 publicly visible posts • joined 13 Aug 2010
Agreed. I used to work on systems that were in maintenance for a couple of decades after delivery. Royalties, number of developer seats, per-named-developer seats, eol libraries, deceased suppliers, annual subscriptions,... the stuff of fing nightmares. OS could be awkward to get through company policies, but once you had ... it was so not a headache.
Yes , but the point is that these days you mostly don't have to be concerned about whether your app will fit into the available memory/storage. Back in the day, the estimates you had to get signed off before starting included memory and processor budgets for every process, plus you had to show the tasks would complete within their time constraints (num_instructions * tick < time-limit). This gets very expensive but you do it because it's painful if you finish the s/w and discover at that point that - it won't fit into the required system, or barely runs.
(The actual need to do all this expensive guessing went away when computers grew faster than software, but it took a long time after this for the requirement to do it to be removed from the QA processes)
In the 80's. I was writing real realtime software with Oregon Pascal and pSOS. There was lots of bit-twiddling and endian stuff to talk to other systems; surprisingly little had to drop down to 68000 m/c due to the wonders of variant records. Worst bit was the linker only able to handle the 1st eight characters of the names leading to ugly prefix warts, e.g. 4x15x1_S(omeName)
Around 2005, i got a call from a defence customer about a system they'd had from us in about 1990. The system had a few more years to go on support, and one of the 10MB Winchester drives had just failed. They still had the spare drive we'd supplied but weren't confident to set it up. I could still remember the pain I had setting the things up first time round - stone age tools, useless fragmented instructions, and sacred details residing in the heads of annoying gits. And the bollocking after for 'wasting time' writing up some decent instructions and trying them out.
Anyway, I was suitably impressed by my younger self''s scribblings because it went relatively smoothly. I was a bit twitchy about getting some data for it off a 15 year old TK50 cartridge that had spent the time in a box in succession of security cabinets on three different sites, but the worst problem was only I'd completely forgotten how to get the cartridge into the drive.
In the 80's there were lots of libraries that had royalties. It was a nightmare to track, especially when software went into PROMs and could be stuffed away in cupboards and drawers where it's existence could become ambiguous. It was a major influence on the impulse to move to pay-once-per-lifetime-anonymous-seat (whatever that's called), open source or shareware equivalents.
Yes but how long from the pilot inspection till completing cockpit checks/paperwork and actually taking off, and is that long enough for a wasp to sneak in. I don't know, but I'd suspect the pilot checks the covers are on during the walkround. Maybe the groundcrew should wave the covers at the pilot before he sets off?
Dell used to ship Windows XP with an interesting behaviour. You would, as normal, set your password when doing the initial setup of the new box and would be admin. However it also had an admin account which was hidden from the pretty XP logon screen and this had a blank password - you got to the login prompt for this by bashing Ctrl-Alt-Del about three times, then hit return to login and you were in.
Nah. Pick the desktop first - Plasma/KDE for the transparent wall-panel users or Gnome for fruity peeps, and then pick a 'currently popular' no-brainer distro in that area. Please, not the niche L-thing you actually use and definitely not something that 'looks the same' (because it won;t work the same in various important ways).
Consider the rules change twice a year with the budgets - three or four months to roll out the changes for these (on the legacy and shiny systems), then you're playing catchup with the grand unification for a little, then the next change hits. And the incentive for the companies doing the work would be to keep it going just like that.
It is actually quite easy to write the minutes up front, based on the last set. Make sure there are a few typos to give them some incentive to play. It turns the meeting into a collective go at a multiple-choice questionnaire. If anyone acts up, they get to dictate the changes they want - failure to agree a set of words means the first jerk collects the action to resolve it.
FFS - adding a USB charger (to a lunatic concept) is going to up the regulatory effort, which is the sort of thing you'd expect an engineer to notice.
"According to the court filing [PDF], Mills was hired to develop a battery-powered pump for an umbrella that incorporates a mist spraying system. After the client informed Mills that the device should support USB charging, Mills raised his initial estimate from $4,000 to $4,800. The client in response complained to the Arizona Board of Technical Registration that Mills was not a registered engineer and demanded a refund of payments made."