Posts by John Smith 19 16330 publicly visible posts • joined 10 Jun 2009
1st rate creative engineering skills.
3rd rate production management and engineering.
A detailed description of using computers to run the building of computers was given in 1956 in
"A progress report on computer applications in computer design"
authored by an R. Kisch and one Seymour Cray.
Trouble always seemed to be with the production side of things.
I've heard figures of around 190+
A government business cloud system hosted in the UK sounds like a good way to consolidate a lot of stuff.
But can you handle all that proprietary Unix?
And those big iron IBM (and perhaps some ICL/Fujitsu) mainframes apps as well?
A really big data centre on the the English/Scottish or English/Welsh borders should be well supplied with lots of cold water for cooling (and lots of power if it's next to an HEP dam). Probably best not to site it below the dam for safety however.
"So as SpaceX comes into the government contracting fold, their costs are going to rise. It is possible that ULA costs will fall, as if SpaceX is excused from certain requirements inherent in the EELV program, ULA can make a good case for also not having to abide by them."
Err you need to be aware of something.
This is the OSP-3 programme. AFAIK this is an "indefinite delivery/indefinite quantity" delivery programme.
It is not the cost plus EELV programme (which relates to the development of Atlas V and Delta IV)
But ULA have charged a great deal for "mission assurance" and been rather coy about what that actually means. Hopefully Spacex's involvement will bring a little transparency to this "magical" process by which they improve mission reliability.
I will raise a pint to that as well.
"Going into classified work will suck out all the life. It's the same as touching a undead body in need of a soul. "
I think it's more a case of how much the people at DoD and NRO insist on things being done their way.
Unlike ULA (who appear to have no commercial business in DIV or AV) spacex has a healthy business. If DoD or NRO complain (IIRC you're not meant to charge the US gov any higher prices because it's a USG order) about the price Spacex will split out the NASA/DoD/NRO "Special Sauce" (c Lewis Page) requirements as a separate item, something ULA seem most reluctant to do.
Market fail. No one wants what your selling. Difficult when running say a water supply utility, not so difficult for trips to the Moon.
Technology fail. It's simply impossible to go to the Moon. Just accept that my opinion of this is that it's BS. I think their architecture is viable provided all the bought in sections are standard straight off the line hardware.
Company fail. Tech can work but this company cannot get it's act together.
NASA's recent (as in the last 30 years) history of implementing major human spaceflight launch systems has not been one of 100% success. The last group of ex NASA types who got into a private company en-masse was Kistler.
The VC's who put $900m into that company will probably not be going anywhere near this lot.
The fact tech startups have 3 failure modes may explain why Warren Buffet tends to stay away from them.
I wish them every success. I hope they do achieve it but just keep in mind NASA's cost analysis of Spacex against how much Spacex really spent. There's a reason those figures were a minimum of 4x out.
"Also NASA needs to stop being a feeding trough for the Aerospace industry and actually require functional hardware or improvements instead of launching million dollar studies into dubious concepts."
This may come as a shock to you but that's exactly how their supporters in Con-gress see them.
If it comes to a toss up of sacrifice some jobs on a programme (but save the programme) so something will ultimately be achieved by the programme or keep all the jobs now (and cancel the programme before it achieves anything guess which they will choose?
Fiscal cliff 24 days away.
"harryreadme wasn't about modelling"
It was about the way you reconcile, adjust, normalize, re-scale and otherwise mash up the raw data into datasets that act as input for those models.
And weather you can recover the precise processing chain used to turn raw data into your dataset.
And I'd say the answer is you can't. But stuff the dataset through a model and you do get an alarming result.
So the cleverness of the model you're driving does not matter. GIGO applies.
That's not really good enough when you're looking at billions being asked for to fix this problem.
Elephants are fairly docile but bears are pretty dangerous, even the cuddly panda's could rip your face off with a gentle slap.
Will China change in time?
Will people go after the low of other green house gases?
Could anaerobic digestion supply 50% of the UK's gas needs in a carbon neutral manner?
Or could the software development clusterf**k that was the CRU "modelling" effort (described in the harryreadme file) finally be sorted out and some real predictions be made?
It's a speech. No new field research or analysis, so no thumbs up from me.
"Industrial are mostly Linux or VxWorks - the latter is properly real-time, but modern hardware is so fast that's starting to matter less in real equipment."
There is also an effort within the Linux community to gradually improve real time response for Linux. There goal has been to get their changes into the core distribution. Coupled with faster processors what's needed to give acceptable RT performance before you need to break out the specialist RT OS has been changing for years.
I unit is 1 cm/sec^2. The map is in mili's or 0.001 of this. That's 0.001 of those. IE 10 micro metres /sec^2.
In terms of the Earths gravity field that's a variation of 1 micro g.
Which is pretty impressive, and the key to doing things like weighing the ice sheet of Greenland using the GRACE satellite pair.
Note that orbital anomalies in Lunar orbiting satellites since the 1960s pointed to "Mass concentrations" or masscons of above average density. What they would be is unclear.
"I would love to see something like Skylon built, but it's either going to be a money issue or a political one or both."
REL know this. As you track through their history you notice they have been very good at forming partnerships and alliances with other organisations (companies and higher education) to leverage their limited resources to get progress.
REL are quite blunt in that they want to see an "industrial consortium" formed to mfg Skylon and most of SABRE, except the pre-cooler (I'm guessing they'd also like to do most of the other heat exchangers in the design, of which there are several).
I think EADS Astrium would be a key player. Senior REL staff have experience of the parts of the company that dealt with Airbus.
I'd hope they steer clear of BAe. It's basically a government con-tractor and dumped it's commercial parts onto EADS Astrium at the first opportunity.
A model could be the way the Tornado was built through formation of "Panavia" and its engines through "TurboUnion." The Channel Tunnel might also be a candidate for the structure (obviously not a perfect one).
"It's important to note that this is revolutionary. As in, it might very well change the way that people do things. i.e., revolutionary."
True. And therefor risky in a way that makes eurocrats very nervous.
Note Skylon targets the same core customers (communications satellite operators) as Ariane, but it's cost model is completely different. I'm not quite sure how the Ariane designs are formerly handed over to Arianespace but that would not happen in REL's case. They would just sell them a Skylon (or 2) and up to Ariansepace to find something for them to do.
"Will ESA man up, back Skylon and change to a model that has fairly small numbers of jobs actually building Skylons/SABREs etc s"
I'm not sure about the jobs ratio. skylon is more like an aircraft. People use the analogy of an Airbus 380 in size and they take quite a lot of people to mfg and assemble, although structurally Skylon should be simpler (fewish unique parts, but lots of them.
Note that REL engaged with ESA on this. The study contract has already been let but I think it was more for REL to get themselves in front of the key ESA staff and get them on their radar.
REL want ESA's support but are very weary of any govt having a substantial investment in them. It's the Concorde thing.
"The rocket-bodies needed for launch-from-altitude weigh MUCH less than rocket-bodes that can sustain a sea-level launch."
Doubtful. Sea level rocket have mass stresses aligned with the thrust direction. A rocket here has to be strong both longways (from where the thrust comes) and crossways (mass on the mounting fixture).
A pressure fed could be lighter because it would need a lower tank pressure to support the same sized nozzle at sea level before flow separation started (roughly 0.4x ambient pressure).
" they will debug their troubled liquid rockets."
what AFAIK OSC only major liquid rocket stage is the Antares 1st stage, which is being done by Russia. Those engines are meant to be versions of the highest performance LOX/RP1 engines ever built.
"But the money behind SKYLON and LAPCAT. And return Britain to the forefront of aviation"
You do know that REL do not want much more than their current 15% to be govt (of any country) funded?
Some of the REL staff remember just how much the men from the ministry (of any country) can f**k up a programme.
However if anyone has a spare £10m VC they are prepared to loose....
This is high cost/risk/reward stuff.
"How about spending that amount on a cure for cancer, who cares if there was life on Mars in the past. Its irrelevant!"
Which one?
There are about 200.
Do you want to remove all cancer genes form people? That would mean re engineering the human race.
Or eliminate the top 5 cancer triggers in the environment. So that requires an outright ban on tobacco for a start.
People know how to substantially lower their cancer risks. It's not a secret.
They just don't want to do it.
"While I am all up for space exploration and scientific research for its own sake, in this current economic climate and with the ever widening wealth/health gap in the US, can these expenditures really be justified?"
Well....
What the US gov bailed out the US banking industry with (a disaster entirely of their own making) was about NASA's accumulated budget for fifty
Or it's annual budget is < 1/2 what Bernie Madoff stole, which is about what the DoD spends on AirCon for it's overseas bases.
Or 2/3 what Americans spend annually on home delivered pizza. Merkins do like their pies.
It's really not that expensive, creates huge awareness of USA Inc's technology and reputation abroad (like Predator drones and Apache strikes but in a good way) and brings in a fair bit of tourist revenue as well, meanwhile keeping a lot of deeply vested interests more or less happy.
And I'm not even that much of a fan.
" If some new designs prove to be more efficient, NASA may be able to get by with what they have, but the clock is ticking and other programs would love to get their hands on the supply for other missions"
If Curiosity is running on the older RTG design a new (Stirling) generator is roughly 4x as efficient but it does have (some) moving parts (although they are all sealed inside).
IIRC the big problem with Pu238 is twofold. The law has always been interpreted to mean that the DoE (and its predecessors) control everything nuclear in the US but only NASA use this. Cue huge congressional bun fight as both NASA and DoD ask for money and Con-gress says it's only really NASA that wants it but only DoE can have the money (because US + Nuclear -> DoE)
The other problem is (apparently) to keep the extraction process simple you want to make the starting material very pure and it's something odd like Neptunium.
Where nuclear is concerned NASA cannot just stuff some material in a fuel pin sized flask, drop it off at their nearest reactor operator, hand them some cash and have them drop it in the nuke to cook for a bit.
but boy would it save a lot of trouble if it could.
* Provided SLS (The only programme NASA is legally forced to perform) doesn't eat the budget for it and everything else of course (subject to the US Legislature agreeing a budget in the 1st place. 26 days and counting ).
NASA has knocked out responsibility for the MPCV service module to ESA, Normally NASA administrators would rather have a live polecat dropped down their pants before considering this option.
8 years is a long in the US federal govt space sector.
Re-use of proven systems is a good idea and if the learning curve kicks in Rover #2 should be lots cheaper than rover 1. Historically doubling the product ->15% cost reduction but this stuff is so unusual they may do better. If they do it at all.
"the utterly useless "escalation" rabble seem to be Americans, even if based in the UK."
I wouldn't be too sure of that.
If they were only available in the PM (or they only called back then) they might well be in the US.
Centralised shared services. It's the way of the future.
I can hardly wait.
"Really don't understand why everyone is so against giving the state more power? After all if you have nothing to hide you have nothing to fear.
I have complete faith and trust in our law enforcement, judicial and political system and beleive that no abuse would be allowed top happen. After all the Police, Judges and MP's are all honest, dedicated anand would never seek to gain personally from public office............
Oh, wait...."
A little too subtle for some people.
""How many lives is it estimated that this £1.8bn project will save? "
Lives. No. Money. Yes
The claim is it will save £500m/year in the first 10 years, substantially more the £180m costs.
However you can't ask them how they worked out the costs.
And they won't say where these savings come from. Me suspicious? Damm right.
http://www.theregister.co.uk/2012/10/10/communications_data_bill_estimated_costs/
And what about that "regular review of the business case" that's supposed to be happening.
7/7 and its aftermath killed 57 people, 53 victims, 3 terrorists and 1 Argentinian electrician.
That's £175m a life.
There are calculations about the value of a human life. IIRC none is above $2.5m
"This all sounds a lot like the arguments the USA used to pass the Patriot act and any number of other laws that violate peoples human rights."
It is.
Factoid about the PATRIOT ACT. 100s of paragraphs drafted and passed 6 weeks after 9/11/01/
And 5 of them were probably to think up the BS backronym that makes up its name.
"These are powerful types of communications data that are already available to the police (and very important they are too) and which are very unlikely to be affected by any future technological change."
But they have weaknesses.
They need RIPA approval.
They are not real time.
Her advisors want to be able to call up a map and show anyone moving around at will.
They've seen spooks. They know it's possible.
"OMFG TERRORPEADOS!!!"
Exactly. The militant wing of the ***.
Notorious for their graffiti campaign "P******r seal of approval"*
*Actually there has been no such campaign. But that's not to say there might not be at some point. Which is of course why everybody must be watched 24/7/365 forever.
"Actually, I'd be more impressed if they were switching it on and off frequently - it would demonstrate that the thermal stresses and electrical surges involved were having little or no effect."
Good point. The other issue is at that lamps are switched on at random phase in the mains cycle and cold incandescents have have a start resistance 1/10 to 1/12 that of their on resistance (Allegro microsystems data sheet). IE inrush current 10x to 12x normal at full potential difference across the terminals.
I've long suspected that bulb mfg took advantage of this fact to ensure their life tests showed bulbs could last their 1000s of hours while ensuring an acceptable (to them) number of failures to keep the cash coming in.
That would be anybody who wanted to know about the internet topology of a country.
In descending order that would be nation states, sub national groups (Anyone from Syrian protestors to the militant wing of the League Against Cruel Sports), criminal gangs, individuals.
"Stirling engines and thermocouples have a lot of untapped applications on the ground. Clawing back a chunk of the ~60% of thermal energy which is currently tossed overboard in a nuke plant springs to mind as a first thought."
Chicken and egg. Power generated < infrastructure cost over life of plant.
The Stirling engine that can handle the coolant flow from a GW sized power plant is not exactly COTS technology. Keep in mind the efficiency will be limited by the temperature difference the system works with. You'd want to stick the Stirling between the outlet coolant flow of the generators and the coolant flow from out side for biggest temperature difference. By then I think it's <100c for water so maybe an 80c temp difference.
You want flexibility in it to cope with all the stupid options people insist your password complies with. (case important/irrelevant numbers allowed/not allowed repeated digits etc).
I think people would like to be able to pronounce it in their home language as well. I think this is probably the toughest problem, especially if you want to avoid real words in that language.
Some kind of soundex algorithm in reverse?
But seriously WTF with "password" after decades of warnings? sure for disposables but not long term.
My personal suggestion is (any) obscenity and the words "thieving" "lying" "cheating""parasites" mixed up with any random digits are quite good choices for any utility, credit card or telephone accounts
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
