Posts by John Smith 19 16330 publicly visible posts • joined 10 Jun 2009
Yup, pretty much the unvoiced thinking of so many brexiteers.
Thatcher (the only PM ever with an actual hard science degree) understood the UK's real place in the world and that it was way better inside a big tent (of 550m people) p**sing out than being outside the tent getting doused.
May. Somewere on the spectrum or I'll eat my hat.
Truss. Johnson-lite.
Which is how it's done in Switzerland (and needs a majority of the cantons, so decisions are not biased by towns having a higher population density).
But a) The UK has very little experience of running referendums. b) Cameron was a gormless f**ktard who was p**sing himself at the prospect of Farage hoovering up his back bench majority. In a documentary on Farage he admits the party actually had enough funds for about 15 constituencies (out of 635). No wonder the Johnson told him to p**soff.
A classic case of the Tory elephant stampeeded by the UKIP mouse.
There is a conspriacy theory that it also stopped the UK from complying with tighter money laundering rules, but surely not.
You don't appear to have heard of "Big O " notation. :-(
If the size of the problem is big enough it will still bring the SoA hardware to its knees. That's where this comes in. The problem size you can tackle with better algorithm goes up regardless of how fast the cores (or how many of them) there are to begin with. It's like the Fast Fourier Transform's effect on signal processing software. Turning overnight runs into minutes of computer time.
CFD problems (for example) use lots of matrix math and continue to bring SoA HPC arrays to their knees.
In the early 70's Strassen took that down to something like n^2.8.
BTW Doug Lenat was using heuristic AI in a program called "Eurisko" in the 80's to do mathematical discovery IE to discover various laws of mathematics from basic axioms.
So yes, better than what's taught in standard level textbooks is possible. Apparently it costs a lot more additions, but those are relatively cheap compared to mults.
Can the AI do better than this?
Fooknose. I've seen so many BS AI claims over the years I doubt it's anywhere near as good as they are claiming.
And won the first time because some of the electorial college (a holdover from the days when it took weeks to cross the continent on horseback which has never been sorted out) voted Republican instead of reflecting how their state actually voted.
Fun fact about the Supreme Court in the US. It receives 7000 petitions for appeal.
It actions 70 (1%) of them.
It had a reputation for improving US society. Now that shrub and trumpf packed it with a bunch of corporate and evangelical friendly judges that has gone down the pan. :-(
1) F**k the poor *
2) Blame mortgate hikes on the BoE for doing the job they are required to do.
3) Pray to whatever Dark Lord she believes in that rich people will use the extra money by buying useful stufff that a)Injects cash into the economy (hiring expensive escorts. Doing lots of recreational drugs, legal or otherwise) b) Buys "useful" stuff, like machinery for their companies, or even whole new companies.
*And by "poor" I do mean anyone on < £150k, the point at which someone starts to see a benefit from all of the last 12 years UK tax and benefit changes. Anyone on < c£35k is likely in fuel poverty as well.
Which are supposed (in the UK) to have a 15 yr life, as opposed to the 40 years of "dumb" meters (before they are re-certified and can be reused)
OTOH the head of IT for a certain US energy company (in Congressional testimony) said they are computers, with a lifespan of about 7yrs before they need replacing.
But these are UK smart meters, which will be a special order.
Yeah. Right.
Still good start for the rest. And I do like the "Offline mode required" so if (when) the company goes TITSUP the product has some usability.
this thing runs with a (small) register set and instruction set.
And of course once you a Turing complete processor....
There have been some work on "language security" and it's pretty amazing what you can coopt into running a program once you find something is Turning complete.
Unfortunately :-(
Not entirely true of animals, definitely not true of humans.
Some animals certainly learn things from their parents while humans have this thing called "language"
AFAIK unlike computer memory human memory results in physical changes to brain structures (not just chemical levels) so language literally re-structures the brain.
The real question is how does a large number of neural networks (as shown by fMRI) linked together in various ways turn from a pattern recognition engine (kitten/not kitten) to a "thinking" machine ?
My instinct says the idea in the novel "Snowcrash" is right. The brain has evolved a micro-machine architecture. I don't think the
language/physical behaviour --> brain re-structures --> learns how to do new things (including make new language strings it's never heard before) paradign is powerful enough.
But it'll take someone way smarter than me to figure out how it works. :-(
The Manch U AMULET project under Steve Furball for a clockless 1st generation ARM demonstrated this decades ago.
However then you cannot do chip differentiation by clock speed (you could say all chips have a clock frequency of 0 Hz) which chip mfgs are unhappy about doing.
Asynchrounous chips were also trickier to design (but the design tools have gottent better) and normally used more transistors (although I think this has also gotten lower over time).
So yes, better options exist if you want to do that. OTOH if you wanted to get brain like power consumption (IIRC a brain runs several Petaflops of processing on about 400W, so all those "Electronic brain" stories of the 1960's were complete BS) there are already other ways (Carver Mead's team used a stand CMOS process in non-standar ways to get highly non-linear signal processing, which is necessary as it turns out is most human input signals have enormous ranges)
This seems to do with something called hard clutch engagement where AIUI the rotor disengages from the engine in flight, the other engine takes over partial rotor drive, then the original engine re-engages, so lots of torque on the rotor over whats on the normal running rotor. So far there is no root-cause reason for this happen but then the V22 is the "Talking horse" of air/rotor craft :-(
BTW fun fact from a 1985 NASA report on helicopters (not including the V-22) the rule of thumb for gearbox weight is 0.3-0.5lb (of gearbox)/Hp tranferred. Given each V22 engine is about 6000shp that transfer gearbox is about 900lb (it's only transferring 1/2 the power). Of course materials have gotten better since then (but not that much) but I'm betting that's still a serious lump of mass. About the only gearbox the V22 doesn't have is one for a tail rotor, but it's got several others.
What does have me puzzled is they've folded the rotor blades, but not the wings. This just seems odd.
Ahh.
The light dawns.
With a scanner at each end of the plane that's going to be a looong shaft. To keep it really rigid you'll need a) Lots of supports b) Really thick (heavy) shaft.
The other obvious answer would be limit switches. Scanner hits the switch, generates an interrupt, computer cuts off Rx (or Tx) to that scanner, starts next scanner moving in correct mode.
Sometimes KISS is overrated. :-(
Also I heard that they tried all the HW in a Hercules, but had to throw some boxes away when it was fitted in the Nimrod airframe. Although those huge round radar displays looked pretty cool. I could imagine one under glass coffee table making a really nice video game or analogue TV :-)
There's a similar shaft linking the engines on the V-22 Osprey. It and the associated gearboxes make up a 14 segment drive chain (the wing is not flat) that allows either to drive both props if an engine fails.
It's a massive PITA. Gearbox overheating is the reason why the Osprey cannot stay in prolonged hover. The central GB cooks in near static airflow.
Thanks for raising my understanding. Maybe one day someone will make it work.
More or less the box that ran (or tried to run) the Nimrod AEW that was such a fiasco, also probably mostly written in Coral66.
Known (IIRC) for a semi-hardwired scheduler, like the Alto.
But in the words of Edward Morrow. "Good night, and good luck."
And now check these.
The longest is in Brazil at 2543Km.
But that still the issue of how you store all that power.
Personally I think high speed flywheels running in vacuum have much better long life due to the limited failure modes and fairly ready ability to scale. Plus they can implemented in many different materials, ranging from the low tech (how about the wheels from a TGV? most mass is at the rim and they are used to speed) to the tightly-wound-ribbon-of-carbon-fibre that if it cracks shatters into millions of low mass shards.
The problem with solar (aside from on earth it's only available during the day) is the huge surface area you need, but that's an issue with all the renewables (not necessarily all sustainable energy options however).
When just predicitng how a string of amino acids would fold was viewed as combinatorically impossible.
Which it no longer is. This is great.
If you want to make a protein shape on demand (which you might want to)
Turns out that's the wrong question.
What you wanted to ask is "Given this existing, natural protein, what shapes do I need to make (possibly with an amino acid string, possibly by straight organic chemistry) that lock onto it"
There is a phrase, "The inverse problem" that kind of covers this but doesn't quite. The classic example is where you want to design a wing shape (say for a fluid pump) that delivers certain specific performance parameters, rather than try shapes and analyse them and check what parameters they give. Software to solve this problem does exist but the optimal wing may demand tolerances, or razor edges, that cannot be made by any existing tools.
Still good to know the SoA has substantially advanced in this corner of the problem.
Lead by the buffoon-in-chief AKA Captain Combover.
I discovered that Alaska changed it's election method for Govenor to a PR system where you state your preferences and they don't have to be all with the same party.
So the Reps had a good Rep candidate, and then Sarah Palin and the Dems had a good candidate.
The good Rep candidate wasn't quite good enough and it turned out most Reps 2nd choice was in fact the good Dem candidate.
Palin is playing the "I-was-robbed-this-is-unconstitutional-that-a-republican-candidate-can-loose-the-governorship-the-the-people-didnt-understand-what-they-were-doing" card.
However this election was only because of the premature exit of the last governor and it's unclear if there will be another election at the standard time as well in a few months.
Which is the (apparent) chaos of democracy.
If their s**t stirring elects a narcisistic bully who can be played like a banjo with enough flattery so much the better.
Unfortunately no one teaches a course called "Critical thinking skills" or "Bu***hit detection" to give it a pithier name.
Than rad hardness and power consumption.
You have to ask wha'ts the mission of these processors?
Normally the mission of a probe is to send data back home to Earth.
If you want to collect a great deal of data, crunch it down and then send the synopsis to Earth that's something else.
I can see use cases for it, but possibly not as many as people might think.
The processors that ran Shuttle were versions of the IBM S360 architecture developed for militar apps, not specifically the Shuttle programme (although the IO was).
We'll see how "successful" the design is by how many organisations outside NASA use it for their missions as well.
Why start now?*
<CAM>
I try everything in Britian first. If it works here I try it in a real country.
Britain. Petri dish to the world.
</CAM>
*Except of course BAE, which it has a "Golden share" in. F**k knows why.
Damm right.
Only 3 groups of people voted for Brexit.
The delusional, the greedy and the gullible.
I'd put most of the (remaining) 15 million left who voted Leave (after about the 2 million dead of old age) to be gullible.
Sooner or later the young will realize how much their elders (because it was quite well split by age, as well as wheather you were a numerate graduate) have f**ked them.
And they might take quite stronlgy against quite a lot of the codergly class.
Curiously I know exactly of who you speak.
And yes tha is one British histories great "What if...." questions.
I saw about 10 mins of the Johnson's farewell speech and thought "Started with bu***hit, ended with more bu***hit." "50Gw of offshore wind power (when it's blowing at the right leve, otherwise 50Gw of gas turbine to start up, and feed). "
Hoped for better.
Not worth the wait.
And so "Will the real Ms Truss please stand up?"
Why am I hearing the words "Now the Cobra will reveal itself" in my head?
Simple.
Every time they change this s**t it means people (and by people I mean large corporations) to retrain their staff.
Books and other media need to be rewriten to reflect the changes.
And of course it means anyone trying to make a compatible UI has to chase your changes.
No matter how non-sensical they are.
And of course corporations can't give it up because (all those con-sultants say so) the costs will be soooooo much higher than just bending over and taking it.
Wrong.
The Space Shuttle Main Engines run a Fuel Rich Staged combustion cycle. The J-2s ran a straight gas generator. Totally different.
You might like to think of the Saturn V US as Merlins, and the SSME's as Raptors, although the SSME has substantially better Isp and poorer T/W ratio. LH2/LO2 is simply the highest Isp combo that's (relatively) safe to handle.
And of course the SSME's each made orbit at least 35 times, whereas Starship has yet to reach orbit, although I'm sure we're all expecting it to make orbit at least once before the end of the year.
Basically persistent leaks that lasted IIRC about 18 months.
Root cause. Rocketdyne (the SSME contractors) used LN2 to simulate LH2 at 4x the operating temperature of LH2
No leaks at 80K. Only when they tried the tests (eventually) at LH2 temps did they find the leaks. :-(
Nothing simulates LH2 except a)LH2 or b)GHe.
BTW North American Aviation (Rocketdyne's parent) who also built Shuttle built like a warplane. IOW you have so much redundancy you fly the mission anyway if 1 sensor's a dud and you have 2 others (as they had in the base of the ET LH2 tank) and a sensor in the downpipe and a timeout coded into the engine management (so the SSME's never ran O2 rich) launch the f**king mission. It's got to hold its s**t together for about 6 minutes (like the booster on SLS in fact) and like the booster it's completely expendable
Is anyone else wondering if SX is holding off trying to launch SS to orbit before NASA because it's bad policy to embarrass the customer who's missions are probablly much more profitable than their regular lauches?
Including it's pre-history as part of the Constellation Programme Aries V this thing has been under continuous funding since 2004. IE 18 years.
And note the con-tractors on the Orion capsule sucked up so much cash that the only way NASA could get what is the Service Module done was to get ESA to supply it as their fee for access to the ISS
Nope.
You might think a company with the word "Scottish" in their name might test the system with regional accents, but they didn't, according to friends who experienced the system.
This sort of bu***hit prediction is easy to figure out because the meat sacks are such a huge part of the costs.
The trouble is that all humans have an almost unlimited way to say the same things in different ways.
Unlike the demonstration conversations they also tend to pad the "conversation" with all sorts of extraneous, random garbage, which humans ignore but the system will continually chomp through, desperately trying to make sense of the non-sensical.
TBH companies could save a shedload of cash if they just didn't f**k things up when they set the account up to begin with.
Except.
It can go bang at levels from 4% to 96% in air.
It can diffuse quite easily through most metals.
It makes welded joints brittle even at room temperature
You need to store it either at -253c or 5000Psi to get decent amounts of usable H2 (and the USAF range safety specs class 5000psi tanks in "lbs of TNT equivalent" regardless of what gas they are storing)
It takes about 3x the energy to compress or cool it as it does to mfg it.
Other than those it's just like gasoline.
Much like the $ US taxpayers have been coughing up to get this thing built for most of the last 20+ years (SLS is the latest incarnation of the HLV-that-will-not-die).
Let's keep in mind yes it's bigger than Saturn V.
But at the end of the day it's yet another TSTO ELV, only with added SRB's just to make it more stupid challenging.
Well givne the 10 yrs from 2011 cost the US taxpayers $23Bn and lets say a NASA specific FH launch was $200m (with all the NASA specific "mission assurance" special sauce they need) that's 115 launches worth. Of course FH has only been flying since 2018 so tha's only $9.2Bn pro rata. Only 46 launches.
But this misses the key metric.
The number of jobs in Alabama that the Senator for Alabama can claim he's bought to the state.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
