"You can tell when someone's been a programmer for a while because they develop a deep-seated fear of time zone problems".
565 publicly visible posts • joined 14 Dec 2007
A lot of vertical-market companies spread chip manufacturing between many countries for various reasons. I used to work for a big European company (cough cough Eindhoven) at a design office in southern England. The chips we designed were fabbed in Germany, the wafers were shipped to Malaysia to be packaged and the chips went to various production plants scattered across the world (Spain, Mexico etc.) to be built into company-only branded devices.
I worked for a while providing ad-hoc IT support for a small building company that provided specialised refurbishment and installation services in a rather cut-throat market. When I was first employed by them to sort out their computers I found their network of desktops and Internet connectivity, such as it was, to be rather lacking in any kind of security. I demonstrated this to the company's boss.
"Look, XXXX (their main rival for Government contracts and blood enemies) could do THIS and THIS and there's no password set and..." as I brought up the company's current contract list on the screen, followed by the unencrypted tender document he was writing for a future project in the six-digit price region.
"Oh." he said. "Uh, could you do that to THEIR system, you know, speaking hypothetically..."
The only way to disable MCAS on the 737MAX is to disable electric trim via the yoke switches -- there are cutoff switches for this on the central console. Doing that leaves only a manual trim option, winding small wheels down by the pilot and copilot's knees that drive metal cables that run all the way to the aft stabiliser. It takes a lot of turns of these wheels to correct a runaway trim, a lot of effort and both crewmembers are needed to wind the small wheels if the stabiliser is under heavy aerodynamic load.
There's a manoeuvre pilots can carry out to unload the stabiliser and make it easier to trim mechanically but it requires a lot of height since they have to put the nose down deliberately while trimming. The Ethiopian Air flight had just taken off when the MCAS cut in and trimmed the plane nose down, the pilots didn't have enough altitude and time to carry out this manouevre even if they knew it was necessary.
Don't hold your breathe on the cutting edge fab from China - they are at least 10 years away from current lithography and I would be surprised if they get close to ASML equipment in that timeframe.
Just knowing that something like EUV-lithography is actually possible to achieve and being delivered today is a big step on the road to developing a knockoff that matches or at least closely approaches that capability. There is definitely "secret sauce" in ASML's boxes somewhere since the People's Number Five Rubber Novelty Products Factory down by the docks in Shenzen isn't spinning out 7nm wafers today but it's not White Man's Magic, it's optics and physics. Espionage may also play a part, of course.
from a technical perspective, x86 should have been dead in the water next to the elegance of the 68k architecture,
The 8086 architecture allowed for backwards compatibility in both hardware and software to its predecessor 8080 -- the system bus could use existing 8080-family chips and a subset of the register and addressing modes matched the 8080's internals making it easy(-ish) to rewrite existing 8080 software and get it to work on 8086 hardware.
The 68000 was certainly an elegant clean-sheet design but it had no support chips for a long time after the first CPUs came out. The system bus wasn't directly compatible with any of the 68xx support chips or the 8080-family chips so it required bodges and/or lots of TTL around it to make it work at all. Any software for the 68k platform needed to be rewritten from the ground up rather than 'simply' being refactored.
The problem with long-lived NiFe batteries is the nickel and iron elements need to be very pure. Nickel, well 99.9% purity is feasible at a cost but commercial-grade iron is a chemical zoo at the best of times and getting ALL the impurities (carbon, silicon, sulphur, phosphorus, manganese, potassium etc.) out of iron feedstock takes a lot of expensive processing.
NiFe batteries cost about ten times the price of lead-acid batteries for the same capacity, in part because there's less demand for them but a lot of the price is due to the difficulty of making them to a decent standard i.e. the sort of batteries with a real fifty-year lifespan.
NGK in Japan has been developing and selling sodium-sulphur batteries for grid storage for a while now. The battery pack they had outside their offices as a technology demonstrator caught fire and it took about two weeks for the fire to be confirmed as extinguished. Four days for the lithium-ion battery pack mentioned in TFA is quite respectable in comparison.
There's a sodium-sulphur battery storage facility in Rokkasho in Japan used to buffer the energy from a small windfarm. It held the world record for a long time as having the largest capacity of a grid-connected battery (245MWh). The batteries are arranged in a large open compound, well separated to make it easy for firefighters to get at any battery units that light off without the chance of the others catching fire too.
California is setting limits on the efficiency of the power supplies in prebuilt computers, not the actual total power consumption of the computers themselves. Most prebuilt computers, including very powerful ones have good-quality PSUs which meet or exceed the efficiency specs so no problem. A handful of style-over-substance prebuilt computers from Dell including some of their Alienware 'gaming' PCs have PSUs that don't meet that efficiency specification so they can't be sold in California (and a handful of other states that use California regulations to save themselves the effort of writing their own).
Story I was told, it may not be true but the Facilities engineers installing a set of as-yet unpopulated racks in a new data centre had a 3kW heater where they were working, plugged in to one of the aisle phases because the DC hall was cold and they had been yelled at when they tried to switch off the hall's cooling earlier.
I upgraded my desktop from 8GB to 16GB to appease Firefox which was using 6GB, two GB of which was virtual RAM on the SSD. Firefox responded to this extra real estate by using 11GB, four GB of which was virtual RAM on the HD since for some reason it had taken a dislike to the SSD.
I built a new PC and outfitted it with 32GB to start with. Firefox is currently, as I type this, using 6GB in 12 threads and no virtual RAM so the mouse pointer doesn't freeze every few seconds when I switch tabs any more.
Inside the reactor vessel and piping and circulating pumps and steam generator loops filled with hundreds of tonnes of water at 200 bar pressure and a temperature of over 350 deg C there are a few milligrammes (if that) of radioactive material that shouldn't be there. The slick sci-fi style high-tech monitoring systems in place are detecting this miniscule anomaly in real-time and flagging up a warning and even identifying which kind of radioactive isotopes they are. That's actually kind of impressive.
The excellent Scott Manley of Youtube fame actually explains how to do what Rep. Gohmert was wittering on about, and indeed gives several different possible yet practical mechanisms to achieve this. The bad news for Rep. Gohmert is that it would take a lot of effort (literally) and tens if not hundreds of millions of years to succeed.
I was told this tale by an IT support guy I worked with back in the day. He did a lot of shop and small office work and one of his customers was a large store which had a couple of back-office servers plus PCs for stock control etc. which were supported by a smallish UPS. If there was a power cut the store would be closed but the UPS would allow the office staff to shut down everything tidily before the UPS power ran out. One of his regular duties any time he visited the site was to check out the UPS under load, make sure it worked and that the batteries were swapped out every couple of years to keep them functional.
Day came, there was a power cut, the UPS ran for less time than expected and the servers went bonk! before they could be soft-landed. Oops. Investigation discovered someone had taken the microwave out of the staff break room and plugged it in to a UPS-fed socket over by a stock control PC to heat up their lunch while they waited for the mains power to come back on. Ding!
...in the press report what processor family will be deployed in the BullSequana system Edinburgh Uni is getting for DiRAC. The PDF linked in TFA suggests the engineering chassis, interconnect and cooling solution is processor-agnostic so it may not be AMD Epyc blades that are chosen for this project.
What is this? Some interplanetary bus service?
Actually, yes. Every couple of years the stars align (so to speak) and there's a short window to fly missions to Mars using less fuel than usual in a reasonable amount of time. The American mission that delivered Perseverance and Curiosity directly to the Martian surface took off at around the same time the Chinese Tianwen-1 orbiter/lander mission to Mars was launched back in July 2020.
There isn't a lot of cloud on Mars and the atmosphere is thin so the sunlight that gets to the solar panel isn't attenuated as much as it would be on Earth. Saying that Mars is a lot further from the Sun and the actual intensity of sunlight there is less than half that at the top of Earth's atmosphere.
A Uni I worked at a long time back got a cluster of minis and serial terminals to replace the old punched-card-masticating mainframe that had catered for the student and academic computing workload previously. The minicomputers were very new -- rumour had it they had serial numbers starting 0001, 0002 etc. and the OS was similarly fresh from the oven and only half-baked.
A group of first-year students, doing Maths IIRC, discovered these wonderful toys and, being infantile to the extreme, proceeded to break them on a monotonously regular basis. For example the Rabbit team wrote scripts and later programs to create a subdirectory in the file system, copy itself into that subdirectory then launch an instance and create a sub-sub-directory, digging its way down until something broke. The Ferret team wrote scripts and programs to go after the Rabbit programs and kill them before something broke. The Rabbits usually won and then the people running the computing department had to go round and reset the computers which had gone Bonk!
Eventually the Problem Children were taken aside, it was made clear that they were getting in the way of people doing Real Work on their shiny toys and they should desist. The computer manufacturer, having heard of these exploits thanks to the error reports the Uni sent them donated another mini to the Uni as a crash-test-dummy machine since the information they were getting back was quite valuable for their OS developers. The Problem Children were allowed to indulge themselves on this firewalled machine, the academics and other students could get uninterrupted machine time on the cluster and everyone was happy, that is until the end-of-year exams when the Problem Children all failed to pass. Sad trombone.
It was not unknown to get a service call to a desk in the morning, "My computer doesn't work". Upon inspection I would find RAM missing, the CPU missing, the hard drive missing. Changing out the contract cleaning staff who, ahem, "serviced" the office spaces in the evening cured the problem, mostly. Fitting security screws to the cases reduced the failure rate a bit too.
The ISS (51.6 degrees, ca. 410km altitude) and the Tianhe module (41.0 degrees, 370km altitude) are in orbits at different inclinations with significant and always-changing velocities (speed and direction) to each other. It would take a significant fuel burn and time to move from one station to another.
IIRC the Chinese station uses the same sort of docking ports as the Russian and American manned capsules use so the ISS and the future complete Tiangong station are theoretically capable of accepting each other's vehicles and, possibly in the future, crews.
An old friend of mine once broke his leg sitting in an economy seat in a plane. He was a bit over 2 metres tall and not exactly slim and he usually got a free upgrade to business class when he turned up at the gate but not this time.
He had been hit by a car a few years back and this, along with general bone loss due to age meant that folding and twisting to lever himself into the plane seat resulted in his leg breaking. The guy who turned up with a wheelchair to get him off the aircraft took one look and said something to the equivalent of "we're going to need a bigger wheelchair."
The thing I actually found most worrying is the fact they couldn't even get a reliable video link to work over a line-of-site range of a few km.
Actually they can and do maintain reliable video links, multiple video streams for the engineering teams and post-launch analysis people to go over. What they make public is another matter, either live at the time of launch and throughout the flight or afterwards once all the debris has hit the ground and the fires have been put out. I've not seen any SpaceX streams of the Starship launches the way they provide launchpad streams for Falcon 9 and Falcon FT launches, just third-party long-lens shots from cameras positioned several kilometres from the Boca Chica launch site.
The crash-and-burns of early attempts to land the Falcon 9 first stages weren't making things look good for SpaceX so they often had "outages" of their live streams just as the stages were attempting to land on the barges followed occasionally by resumptions of the video when it was clear they had a standing rocket to show to the world and not a pile of smoking rubble.
Engineers who had an in with the muddle management sometimes got the chance to travel first-class to Malaysia and spend up to a week there on the company dime doing buggerall except waiting for a phone call. When that phone call came they took a hire car to the chip packaging plant, picked up a heavy suitcase of freshly-packaged prototype silicon and then headed for the airport where they'd get on the next flight back to Heathrow. The first-class ticket was, at least at that time, fungible in that any airline would honour it (in the off-chance the high-flyer would book with them next time). Saving a few days in prototype turnaround using a courier was well worth the trivial expense of the hotel, tickets and engineer downtime.
I'm sure lots of people think that the sudo code must have been closely studied over the years. That doesn't mean it has actually been closely studied, indeed that mistaken belief might have induced many people who could and would have gone through the code with a fine toothcomb to decide to spend their efforts elsewhere in the open-source world.
A lot of Japanese appliances use DC motors because of the 50Hz/60Hz split in their two grids. It's easier to convert the AC mains to DC to drive a motor rather than have something run at differing speeds depending on where in the country you plug it in.
We discovered this when autopsying a dead Japanese coffee grinder someone had tried wiring up to British 240V mains. Sorry, but it wasn't ever going to run again after that...
If the aircraft had no nose-up tendency what was the purpose of MACS ?
There are legally-enforceable flight regulations that limit the amount of force needed on the control stick for stall and approach-to-stall. The 737MAX enlarged engine nacelles and their forward positioning on the wings means the pilots need to input more force than the legal limits in certain (and rare) flight conditions. MCAS, when it works properly, automatically prevents the plane from going nose-up too fast hence reducing the amount of control force needed. When MCAS goes wrong it pushes the plane's nose down even if that's a bad idea at the time.
Until MCAS was introduced the AoA sensors only provided a warning indication to pilots that they were approaching a stall. If the AoA sensor failed it could be ignored by the pilots who could fly the plane perfectly well without it.
MCAS as implemented on the 737MAX uses inputs from one of the two AoA sensors to drive control surfaces without pilot input if the sensor reports an approach to stall condition. If the sensor fails and reports an approach to stall in certain circumstances MCAS will pitch the plane's nose down forcefully even if it's flying level or in a climb.
Fitting an extra AoA sensor would require a lot of recertification paperwork, engineering changes etc. After that the MCAS system would need to be restructured and rewritten to deal with two-out-of-three voting and even then it still might fail on occasion. Doing all this would have indicated to the licencing authorities that something was wrong and it's possible the 737MAX would have required certification as a new aircraft type. Preventing this certification change from happening was the entire point of MCAS in the first place.
I do have a second monitor on my system, a 27" 2560x1440 IPS display (semi)-permanently in portrait mode to one side of my main monitor. It's handy for proofing page layouts and displaying tools I don't need all the time in my direct field of view.
Cheap monitors are the Devil's bargain as your aching eyes will confirm. I paid 800 quid for my 32" 4k display six years ago when they were at a premium, I have not regretted it once.
I have a pair of spectacles specifically for use with my computer monitor (32" diagonal 4k IPS 16:9 ratio). The optical prescription provides me with a good no-strain focus at an eye relief of about 75cm or so. Visually speaking this monitor's display field just about fills my peripheral vision at that distance, a little smaller than the frames of my specs. However my eyes aren't what they were even six years ago when I bought this desktop monitor and replacing it with a 43" 4k display is very tempting.
I'd happily use a 4:3 monitor again, almost everything I do is long, not wide...
You could tape a couple of bits of cardboard to the sides of a 16:9 monitor if you want a 4:3 ratio display. Alternatively there's probably someone out there with a Kickstarter for motorised side-curtains for monitors (like the old cinemas had for their screens) so you can adjust the ratio to suit whatever you're doing.
Myself I'm only considering replacing my 32" 4k 16:9 monitor because 43" models with the same resolution and screen ratio are becoming affordable. A 17" 4:3 monitor just wouldn't cut it for me today.
The F-35 is way late.
There are more than 500 F-35s of all variants in service with several air forces at the moment -- the USAF has about 250 of the F-35A variant alone in training, testing and operational squadrons. There are plans to build about 3,500 more F-35s over the next twenty years or so. I don't see how that makes it "late".
During WWII one of the most secret weapons developed by the Allies was the proximity fuze for anti-aircraft use and, later, land bombardment. The fuzes had delicate components inside (small valves and batteries etc.) and antenna connections to the external cap. In some theatres of operation protection for these fuzes for storage and deployment was needed, basically snug-fitting tapering plastic cones. They couldn't just put out a contract for fuze protectors due to the need for secrecy so they used a "back door" connection, so to speak, with the John Hopkins hospital to order five hundred thousand rectal spreaders.
> what kind of twat carries nunchaku?
Oi don't be like that! They're great for entertainment value.
They were all the rage in the 80s
Not with the British Board of Film Censors (the BBFC). The head hocho of that august organisation was seriously down on nunchuks, and ninjas too -- we in the UK were presented with something called the Teenage HERO Mutant Turtles movie. The Turtle character that normally used nunchuks was mostly edited out of the action scenes and only appeared during a comedy skit where he wielded a string of sausages.
Kung-fu movies with Bruce Lee and others where nunchuks were used were also cut to shreds for commercial release on DVD and the like.
China isn't working on a Liquid Fluoride Thorium Reactor (LFTR). They're testing a bunch of other nuclear power plant technologies, including a version of the Russian fast-spectrum sodium-cooled reactor (the fabled Waste Eater) as well as a helium-cooled high-temperature pebble bed reactor that might use thorium in some fuel mixes, but molten-salt fuel stream thorium breeder reactors, nope.
Saying that I'm sure there are in China, like in the West, a lot of Powerpoint Cowboys with theoretical models by nucleonics PhD students on how LFTR technology is really great and has no downsides. What doesn't exist is any concrete and steel, any licences, any approved designs, any funding and construction approvals for LFTRs. Instead Chinese companies are building pressurised-water reactors, nowadays mostly home-grown designs like the 1100MWe Hualong One and the newer 1400MWe Guohe One reactors.
The US had a reactor in operation in the Antarctic for a few years back in the late 60s and early 70s. It was at their base in McMurdo Sound near the coast, not at their permanent base at the Pole. It turned out to be more trouble than it was worth, basically.
They did put another reactor in at another base in Greenland as part of the Iceworm engineering experiment. Again it turned out to be not worth the effort and it was removed when the Iceworm project was abandoned.
Geothermal energy is usable in certain geological areas but it has limitations -- rock isn't a good conductor of heat so once a well has reduced the temperature of a volume of hot rock deep underground the efficiency and the amount of recoverable energy drops off unless more heat can be transferred quickly to the well bottom. The best sites for geothermal energy have lots of fractured rock strata and hot underground water that can be extracted while more hot water flows towards the well bottom. That sort of geology is a lot rarer than simply large volumes of hot dry basalt a couple of kilometres down.
Iceland had a magma-powered geothermal well at one time, they were drilling a regular hot-water geothermal well and hit lava instead. Oops. After retooling they got it to work somewhat -- the lava ate the drillhead and damaged the piping, not surprisingly. It's been shut down since then, IIRC. Maintenance and operating costs were the problem, the lava pool the drill string ran into produced highly alkaline soluble salts and gases and other crud that chewed up the heat exchangers at the surface, I think.
The current (ha ha ha) price under "Contract for Difference" rules for the power from the Hinkley EPR1600 reactors if and when they come into commercial operation is £104.50 per MWh, not the original "strike" price of £92.50 per MWh.
That sounds really high, doesn't it? A quick look though the public data published on the web for CfD agreements shows the big Hornsea offshore wind farms -- enough electricity for a million homes! as the press reported when the first Hornsea array came on stream last year -- had an initial strike price of about £140 quid but its CfD price is now £162.47 per MWh, over 50% more expensive than the Hinkley reactors. The other two uncompleted Hornsea wind turbine arrays have the same CfD price.