Posts by molletts 114 publicly visible posts • joined 20 Jan 2016
Some of the teachers at my school used to either not notice or forget to retract the pull-down projection screen in front of the new-fangled drywipe boards that were being gradually deployed. There were numerous screens around the school with the odd letter or two written on them where someone had started writing then realised it was the wrong surface.
Careful examination of the copious small print reveals that the free T-shirt is, in fact, the only benefit of the Treadmill Pro Silver Subscription when compared with the basic version and that there is a minimum contract period of 5 years...
(Oh, crap! I should have posted this anonymously! Their lawyers will be all over me - I had to sign an NDA before I could view the T&Cs of the contract!)
I have a vague recollection from probably about 20 years ago of someone demonstrating actually using the NT kernel to run Gentoo. I don't mean running Gentoo Linux in a WSL2 VM or having a Portage-managed userspace under Cygwin, but actually using it as the kernel's userspace - booting the NT kernel then presenting a CLI at the HAL console. Presumably, the old POSIX subsystem was a major enabler, providing enough of a libc to get the basic stuff running.
It was very experimental and really only one those "because we can"-type things, like the recent demonstration of installing Linux via the Windows installation environment, but it was cool nevertheless.
I can't remember where I came across it (or if I may still have it somewhere) but I once encountered a ball mouse with entirely electromechanical quadrature detectors - the principle was the same as the "slotted disc" optical type but they had little round PCBs with radial stripes of copper which pressed against springy wipers.
It wasn't terribly reliable and didn't track very well either, probably because of the relatively high torque that was required to turn the rollers.
I seem to recall workstation names including "fanny" and "breast" in one of the computer labs at uni. I'm sure the IT team would have claimed they were randomly-chosen according to an entirely innocent documented pattern of names if asked.
I've had pairs of servers called Romulus & Remus and Castor & Pollux in the past, before boring but informative names like "host1" and "host2" became de rigeur.
On "old code" - I was absolutely gobsmacked a year or two back when doing something in Office 2019 (I think it might have been a mail merge in Word) and a vintage Windows 3-style file dialog popped up, complete with the drop down list of drive letters! I think I might even have had to map a drive to the UNC path from which I was planning to open a file before I could access it.
They seemed to sync themselves with what was on the PC pretty enthusiastically in the early days, in my recollection.
I recall the school where I worked at the time getting some shiny new servers in summer 2003 with a whopping 72GB of storage on the main file server (minus the capacity needed for the Windows system drive, which was probably about 8GB at the time for Server 2003). After the old 18GB NT4 box, that seemed pretty roomy. All staff were also issued with laptops for the first time, rather than having to share a departmental desktop.
A few months later, the free space on the file server suddenly dropped from a healthy 40-odd percent to just 1%. I deployed SequoiaView that evening to see where all the space had gone and immediately spotted a 25GB chunk of media files which turned out to be someone's iTunes library. As it wasn't a music teacher, who might have been able to justify having an extensive music collection on the system (although I would certainly have approached them and arranged to reconfigure iTunes so that the files were stored locally on their laptop, as indeed I did when a boy at another school where I volunteered did a similar thing), and the situation was a bit of an emergency as people would soon be getting "disk full" messages when saving, I simply deleted the folder. No problem, I thought, it'll still be on the iPod and I can go and see them tomorrow to move the library folder before it backs up again.
They came and saw me first! They had plugged their iPod into their laptop when they got to school and it had promptly synced from the empty folder!
I made the appropriate soothing and apologetic noises while pointing out that they had nearly "crashed the server" (easy to understand even if not terribly accurate) and I'd had to take emergency action. Thankfully, they were very understanding and agreed to let me move their iTunes library before they restored their collection to it.
I filed that bit of knowledge about iPods away under "useful, for good or evil".
I had a hire car a few years ago that had a similar problem. One of the roads I was using at the time to get to work is quite bendy but also quite busy. Every time I approached a bend as something was coming around it the other way, the car would detect another vehicle directly ahead and slam on the brakes, much to my annoyance (and that of the person who was following me in a BMW, who retaliated after the third incident by overtaking me then slamming on his own brakes as he cut in). If I'd had the car for more than a day, I'd have worked out how to disable the auto-brake system.
+1 for the Firefox Multi-Account Containers. It's an official add-on from Mozilla; why they don't just build it into the browser, I have no idea. It's probably the biggest of the many reasons I use FF as my main work browser despite being in a "Microsoft shop" - I have to juggle four Office 365 accounts on two different tenants (plus occasionally needing to get a user to log into their account or logging myself into a test account) so it helps to preserve what's left of my sanity.
I used XFCE for some years but it started becoming more and more like "Gnome Lite" with menu bars vanishing, massive wide title bars containing very little, no OK/Cancel buttons in dialogs, etc. so I took the plunge and switched back to KDE last year. I had previously used KDE 3 up until the bitter end when it started succumbing to bit-rot, then KDE 4 which, by that time, had regained most of KDE 3's functionality, before switching to XFCE mainly because I wasn't really using much of the KDE ecosystem any more and wanted something more lightweight that didn't need nearly half a gig of RAM just to log in.
I was pleasantly surprised to find that KDE 5 actually used less memory than XFCE (!), as well as looking and feeling generally more polished (which was expected), thus avoiding the snide comments about "that crappy old Linux thing you use that looks like Windows 95". Someone at work actually mistook it for Windows 11 last summer. How insulting :)
I was, however, a little disappointed to find that the desktop cube animation is no longer available, so when I say, "KDE," and people respond, "Oh, is that the one with the spinny cube thing?" I have to reply, "Yes, but no." :(
I don't think Canadarm would be able to give the modules enough delta-V to de-orbit them. They would just end up in an orbit very near that of the remaining pieces. (Also, it would impart equal and opposite momentum to the bit the arm was attached to.)
The simplest method would be to send up capsules to dock with each piece and use their engines to de-orbit, which sounds like what they're planning. I assume they have a good reason for wanting to come up with a new craft (maybe more taxpayer $$$ for the contractors, or am I too cynical?) rather than using an existing one (the Russians will probably just use Soyuz for their bits).
The one (and only) thing I can say for Apple's blatantly-anticompetitive policy on browsers is that it blows a hole in any website's claim of only working with Chrome. There's a site we use at work that, even if you visit it using Edge (which, let's face it, is just a re-skinned variant of Chrome that phones home to a different mothership), simply displays page that says, "you need to be using Chrome to access this site; click here to download it." Funnily enough, it works fine on an iPad regardless of what Safari skin you happen to be using, despite refusing to run on Safari on a Mac.
I get that too and I'm also far too young to remember it. I don't even recall seeing my Nan react to it (simply because I don't think I was ever with her when a siren was tested nearby). I know she hated thunderstorms, though, because they reminded her of the Blitz - she would go and sit in the airing cupboard which was in the centre of her maisonette well away from the outer walls so the sound was quietest there.
I think, at least in my case, it's a kind of "cultural conditioning" - I've seen so much footage of the Blitz, almost invariably accompanied by the sound, that it's become indelibly associated with death and destruction. It's also a very mournful, rather eerie sound anyway (although some of that may just be because of the association).
At a previous employer, I got told off in front of all the support staff because the site manager had found the IT office locked when he swept the building (it was an "unscheduled fire drill" - i.e. a disgruntled teenager hitting a few break-glass callpoints as he stormed out of the school after being asked to do something he didn't want to do - so had to be treated as a real fire).
I pointed out that I hadn't been in the IT office when the alarm went off and asked whether, in future, they would like me to a) leave it unlocked at all times and accept that expensive things would get damaged and/or go missing regularly or b) on hearing the alarm, return to the office and unlock it before leaving the building.
It was rather grudgingly agreed that I could continue to lock the office when I left it and would not be required to go back and unlock it in the event of a fire alarm. I think the intention was that I should not lock it as I left if I was evacuating but I continued to do so anyway.
It'll probably be able to run Crysis with software rendering (see the various demos of this on Threadripper).
As for Windows 12, its physical address space limits may be a little snug and you'll need a dual-socket system if you're going to have any cores left over for doing actual work ;)
Point 5 is an interesting one.
On the one hand, presenting a simple, straightforward message in plain English sounds like a great idea: when the brown stuff hits the (turbo)fan, even the best-trained pilot is ultimately still a funny-looking ape and is susceptible to brain failure - hence the occasional tendency of pilots to pull up when their aircraft begins to stall and descend despite "knowing better". (That's also why you should always pay attention during the safety briefing, even if you've heard it a thousand times before - each additional hearing decreases the chance of your brain entering panic mode and going blank if you do ever need to take action.) A nice, easy message that doesn't require interpretation could be just what's needed to elicit an appropriate response.
On the other hand, though, we need to be wary of letting the automatic systems over-interpret the sensor inputs (which may be wrong anyway) and potentially display an incorrect diagnosis which could then influence the pilots' own diagnosis (or, indeed, short-circuit it when the pressure is on). Sometimes, it's better to let the meatbags think it through for themselves, taking into account all their observations and experience.
The other big problem with any automated system, be it the autopilot of a complex airliner or simply a high level of driver assistance in a car, is what happens when something occurs that the computer can't handle. It can take a few seconds for a person who has been largely out of the loop to overcome the startle response and re-establish a basic level of interactive control when the computer throws in the towel and dumps the controls back into their hands, and quite a bit longer to get back to full capacity. There's usually a bit more time before things go from bad to worse in an airliner with lots of empty air around it than there is in a car on a busy road, which is why we don't insist that pilots hand-fly the aircraft for hours on end, but it can still be an issue.
I'm a little surprised that, having gone to the trouble of taking two samples at each location and caching one of each at the depot in case of problems retrieving the ones kept on the rover, it sounds like they're going to put both sets of samples on the same return craft. OK, they can get twice as much material back in one go but if the ascent to Martian orbit, capture by the orbiting return stage, transfer to Earth orbit or descent to Earth surface goes wrong, they lose both sets.
I guess they've done the maths and decided that the chance of something going wrong during any of these (really hard-sounding but probably quite run-of-the-mill for NASA) operations is much lower than the chance of a problem with transferring samples from the rover (which sounds really simple but is probably filed under "never attempted before").
A "do not disturb, except phone calls" feature would be nice.
I don't know whether iPhones have anything like that (I've never had one) but I've had it on my Android phones for many years. I can't remember whether it was in the stock ROM on my Galaxy Nexus but it's in LineageOS 14.1 (Android 7.1). It was also in the stock OxygenOS on my current OnePlus 6 and is in LineageOS 19.1 (Android 12.1).
I have it set a little stricter than "just phone calls" - silent except calls/texts from starred contacts - and it comes on automatically overnight.
I knew someone who didn't believe that neural networks existed before the current "AI" obsession.
It blew his mind when I showed him my copy of "Explorations in Parallel Distributed Processing" from 1988, revealing an already-established field of study. As for the 5¼" floppy disks tucked into the back cover, they truly blitzed his chakras.
The possibility of emergent behaviour is something we cannot and should not dismiss out of hand in systems of this level of complexity and, indeed, for which we should be vigilant as we dial up the parameter count to ever more mind-boggling numbers, but we must also remain sceptical and remember that this kind of human-like conversation is exactly what these models are "designed" to do (whatever that means in the context of the ultra-high-volume statistical data-mashing that we refer to as "machine learning" and "AI").
And anyway - nobody has yet managed to formulate an unambiguous definition of consciousness so how can we say for certain whether something is or is not "conscious"?
A simple "No features of this device work when OFF " message would explain the issue and allow the user to make a choice.
And yet there would still be people who would whine that the alarm didn't wake them up or they can't pay for stuff or they have to cope with the unbearable hardship of having to carry a separate key for their car or whatever when they've used the "turn off fully" function...
I can't remember whether it was Amstrad Action or one (or more) of the books I had that used it but I seem to remember having a program that echoed a 16-bit checksum (I think it was literally just a sum of the bytes, not anything fancy like a CRC) every time you entered a line of code or after every line when LISTing a program. You could then quickly cross-check that against the one printed in the mag/book.
Very useful when entering a long listing.
I sometimes get weird download issues with the LibreOffice torrents (I seed all the "current" versions using spare bandwidth on my home server) - they get to 99% then go back to, say, 92% and re-download the last few MB. If one of them does this, it'll likely get stuck doing it repeatedly for hours before eventually completing successfully. I've never had this with other torrents (I also seed a few Linux distros' ISOs and some big X-Plane sccenery files).
At one of my previous workplaces, I sometimes used to move my old Volvo 440 (which had power steering) from my allocated space near the top of the fairly-steep car park down to the (un)loading bay at the bottom without starting it. Both steering and brakes required significant effort when there was no hydraulic assistance but fortunately my previous car had been a 1986 Vauxhall Astra which was rather tank-like in construction and I still had the muscles from it.
I've never tried doing this in my S60... I suspect it would be nearly impossible, partly because of the extra weight and partly because I've gone soft in the intervening years!
(To get back on topic, the 440 was in "Silver Gold Metallic" IIRC. I think it was registered as "silver" and that's what it looked like - until you parked it next to an actual silver car. The S60 is something like "black sapphire", registered as "black" and, again, it just looks black except when it's beside a black car.)
The type of phone auto-attendant I hate most is the bastard child of the two - a limited selection of options which you choose not by pressing a number but by saying a word or phrase. Someone invariably comes and hangs around by my desk just as I'm navigating one of these and gives me odd looks when I start saying seemingly random, disjointed words separated by long pauses. "Support"... "Business"... "Networking"... "Yes"... It's even worse if I have to say a long serial number or something: I know from experience that I can't just read it out naturally - for it to be reliably recognised, I have to say. each. digit. as. if. it. is. a. separate. sentence.
You think a 30-second loop is bad? The Volvo Insurance line had about a 10-second muzak loop last time I called it. I had to wait ages (and boy, does the short loop make it seem even longer) to activate the free "drive away cover" when I last bought a Volvo.
You should see the walnuts from my neighbours' tree. I picked some "fallers" up a few years back, eagerly anticipating lovely sweet, soft nuts just like the ones we used to get from "Auntie Pam"'s tree when I was a kid and nothing like the bitter, dry things you get in the supermarket. I had to use a bench vice to crack them. The bloody nuts were the size of peas; the rest was solid shell. Just for laughs I put one under the wheel of my car and drove over it - it survived intact.
Sounds familiar - I had to bring an unfamiliar system at a small customer site up from cold last year. The Hyper-V cluster wouldn't come up because it depended on DNS for the nodes to find each other (and probably AD for configuration); the DNS servers were VMs in the cluster...
Fortunately, I was able to attach a keyboard and monitor, log in locally on one of the hosts and manually start one of the DCs. Once both nodes noticed that they now had DNS, they started talking to each other and everything sorted itself out.
(Actually, "from cold" is probably not the best description - all the switches, the router and the SAN had started up automatically when the power came back on goodness-knows-how-long before I got there; the office-style air conditioner needed to be turned on manually by pressing a button. The tiny room was absolutely baking and the shrieking of the fans in the switches was deafening.)
What the hell is it with the Office 365 web version of Outlook where some random part of its UI suddenly gets focus while you're typing an email and not only swallows several keystrokes but acts on them as commands?! If you're lucky, they just get redirected to the browser UI and you find yourself doing a "search within this page" but sometimes you end up deleting the message you're replying to, archiving another one, snoozing one, emptying the folder...
At first I thought it was either me palming the touchpad on my laptop (although I've never had a problem with that in the past, but maybe the anti-palm thingy had stopped working...) or using an "unsupported configuration" (the laptop runs Linux). Then I had exactly the same problem while using it on a Windows 10 desktop so it can't be either of those things.
I'm seriously tempted to defy the company standard practices and use a real email client.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
