Perfect? Shuttles? [laughs] Oh my god, you're adorable. The shuttles were a hideously compromised design from day one.
The wear and tear on the system from entry and reentry is to high in most cases, where you'll see stress cracks places you might not expect. For every one stress crack you can see how many are forming that you can't see?
Neither Challenger nor Columbia failed because of cumulative fatigue. Challenger's O-Ring was a new part, replaced every time they recycled the SRB casings. Likewise, Columbia's ET and foam were brand new for each launch, and the RCC leading edge on the orbiter would have been holed even if it was brand new.
However, to your point, SpaceX has no intention of flying the same set of engines for 25 years. They will determine the MTBF (say 25-30 launches) and will retire each batch of engines when they reach the point of diminished returns. Same with the tankage, thrust frame, legs, etc. Engines will be tested between launches, as they are already tested three times before launches. The first stage accounts for 70% of the Falcon 9's costs. But testing costs 10% and fuel just 1%. A single reuse will cut a third off their launch costs, two will cut half.
And if reuse somehow ends up costing more than building new... guess what? They won't reuse. They already have the cheapest launcher on the market, they are well on their way to having their own private HLV that'll be every cheaper. With the shuttle, reuse/refurbishment was a requirement. For SpaceX, reuse is gravy.
On the other hand you seem to be completely discounting the wisdom of learning from other people's experience. Having already tried to do this maybe NASA knows something about it. Or should every other rocket maker from NASA to the end of time all repeat the same expensive mistakes just in case they can find a way.
From what I've seen, it's NASA (and their primary contractors) that fail to learn from experience and repeats the same expensive mistakes over and over.
SpaceX started with NASA's cancelled Fastrac engine program. NASA cancelled it, NASA's primary contractors ignored it, SpaceX turned it into a successful business. Hmmm.
SpaceX will have a 50 tonne to LEO HLV in a year or two that they are pricing at less than $100m per launch. Their existing 13 tonne to LEO launcher is priced at less than $60m per launch. And they spent less than $500m to get to their first F9/Dragon launch.
Meanwhile, NASA is struggling to recycle shuttle technology to develop a 60 tonne to LEO launcher (SSME+SRB+Centaur version) for $3b/yr, and maybe $1b per launch, which they hope to launch by 2017. Afterwards, they want to increase that to 70 tonnes by 2021. 130 tonnes by 2027-2030.
For the price of a single year of SLS/Orion development, you could buy 30 FH launches. Say 1,500 tonnes to LEO, for a single year of SLS/Orion funding. For half of just SLS's funding, you could buy 10 FH launches every year. 500 tonnes to LEO each year. And still have half the annual SLS funding to develop something to actually launch. And you could do that for at least seven years before SLS would have allowed NASA to develop a single piece of mission payload.
Doesn't that suggest that maybe, just maybe, that it's NASA doing something wrong?
Exactly, SpaceX thinks *outside the box*. This innovative thinking allows them to toss aside all learned wisdom and knowledge from those old dinosaurs at NASA. Innovative thinking, used to build synergy and form a new paradigm, THAT'S what it's all about. The physical reality will follow from that.
Except this is the opposite of what SpaceX did. They took a tested but cancelled design from NASA for a simplified Kero/LOx engine (Fastrac) and built an engine based on it. They gradually upgraded it, learning as they went. If they hit a wall they hired engineers from NASA and the Primes who'd been involved in rocket development. They incrementally improved their technology, they proved their engines first on the test stand, then, and this is important, they proved them on a single-engine launcher. Which failed three times. Once they solved the problems, they moved to the larger launcher using the same engines. Only once they had proved their 9-engine version, did they take what they had learned and redesigned it to make it better.
Likewise, when they built Grasshopper, they didn't create a bespoke test vehicle with a fundamentally new type of engine (like DC-X), they just used a first stage from their now-flying commercial launcher. That meant that everything they learnt applied directly to their main program, and vice-versa. Each time they fly-back a first stage, they learn more. And it's basically a free test, since the risky stuff only comes after the main mission has staged. F9R tests the re-entry and approach. Grasshopper tests the last-mile and landing. Each getting a little closer to closing the loop.
When NASA created the space shuttle, they threw away everything they learned from the previous two decades. They required a brand new type of engine (LH/LOx), with a radically improved Isp, which was reusable in spite of burning for the entire launch, and had to be powerful enough to loft a 100 tonne space plane to orbit. They required a brand new heat-shield material, which was only theoretically possible, applied in a brand new way. The shape of the space plane had no resemblance to any prior space program or X-Plane research, and right up to that first manned launch in '81, was only theoretically possible. They required a new launch configuration (side-mount), and SRBs larger than anything built before (which were only ever tested in a horizontal static firing. They didn't even do a sub-orbital test launch.) And again, the whole configuration wasn't tested before that first manned launch. There were no intermediate steps, except the Enterprise glider, and no way to do incremental development.
Many excuses are made for this to deflect blame to others, but we can see from subsequent programs that NASA repeated exactly the same mistakes...
They did it with Space Station Freedom, wanting to build a giant all-singing-all-dancing space station with no stepping stones, no test programs, no connection to Skylab, and promising to do it fast and cheap.
Likewise, when they created the shuttle replacement program, VentureStar, they picked an even harder design (SSTO), so that even their sub-scale test vehicle (X-33) required a dozen cutting edge technologies that had never been tried before and were only theoretically possible. (One of which was only developed a decade after the X-33 program was cancelled.) Exactly as they did with the shuttle. Because they totally learned their lesson there.
They did it with MSL-Curiosity by trying to jump ten sizes in one step, which blew out its budget 100%. They are doing it with JWST, which tries to shoe-horn every technology they can into a single program, which has blown out it's budget 400% so far and still going climbing.
Is NASA learning it's lesson? I've been told straight out by NASA science guys that a project is "worthless" unless it includes as much new technology as possible. Oh, there's a few engineers within NASA who get this stuff, and occasionally you see a micro-p
They've still got the cheapest launcher on the market
The ESA's Vega is cheaper per launch to orbit but with smaller payloads
[sigh] Really? That's how you interpreted my comment?
You're comparing a $40m/launch (+$700m/dev for three launches) 2 tonnes to LEO launcher with a $57m/launch (+$500m/dev for 9 launches) 13 tonne to LEO commercial launcher?
Hey, Falcon 1 launched half a tonne and cost $6m.
How low can we go? Is there a nano-sat launcher somewhere that costs $3m?
Idiot.
SpaceX has a bigger promotional budget though.
SpaceX would spend less on their promotional budget than the Primes are spending trying to undermine them.
The problem is that if they launch over the sea and then try to recover a first stage back to land it is going to burn a lot of fuel reversing course before it falls out of the sky
About 30%. Which is part of the reason for the recent F9 v1.1 upgrade. Fuel is cheap, and the rocket equation is kind when adding weight to the first stage. [Mass added to the first stage has vastly less effect on payload than payload mass has on the first stage.]
The first stage is responsible for 70% of the cost of launch, according to Musk. But the fuel is only 1% of the cost. Provided the refurbishment cost can be kept below the build cost (which their Grasshopper program clearly says they can, or why would they be continuing?), they will save money by reuse. Over time, a lot of money.
And if it doesn't work, so what? They've still got the cheapest launcher on the market, and a heavy lifter on the way.
What makes the Merlin engine so interesting is precisely because it is bland. SpaceX hasn't been trying to push the envelope
It surprises me how few people get this. Even people within NASA.
Saturn V was Kero/LOx, and it had a crappy Isp. But it built upon the experience of the previous two decades of rocket (and missile) research.
When NASA went to the shuttle, it threw away everything it had learned and started again. A LH/LOx engine (much harder), with the highest Isp of any engine to date (even harder), which was reusable (seriously?), on a 100 tonne space-plane (I mean, seriously!), launched side-mount on an entirely new configuration. There were no stepping stones, no way to learn your "craft", to understand the limits of materials and techniques. It was completely unrelated to either previous rocket programs or X-Plane research. Much of the proposed design was only theoretically possible, such as the heat-shield, but every single piece had to work right on the very first test launch, manned, in 1981. That is an Apollo level challenge and it's stunning that they were able to get anything that flew, let alone flew for over two decades. But it's not how you build practical systems. It's not how you build affordable systems.
GP is referring to Freedom Space Station, not Skylab. You know that, why are you lying and pretending you don't?
Freedom was proposed to Reagan in '86 and accepted in '88 (continuing under Bush I). It was supposed to cost $8 billion all up and take 8 years.
8 years and $8 billion later, nothing had been launched; hell, nothing had been built. The design had shrunk from a 12 man space station much larger (internally) than Skylab, to a 4 man station much smaller than Skylab. (The in-house joke at the time was that they had to call it "Fred" because they could no longer fit all the letters of "Freedom" on the side.)
So in '96, Clinton forced an enquiry which required them to take what had been designed up to that point, pick one of the three leading configurations, and just build it. This led to the current design, then nicknamed "Alpha". In '98, a requirement was added to merge the station with Russian modules (and to a lesser degree Europa and Japan). This would reduce NASA's cost in developing some of the core modules, get modules launched earlier, and buy access to Mir technology/experience; it was also thought beneficial to keep Russia's space program intact, to prevent rocket engineers going to work for Iran/Iraq/etc. That became the ISS.
As for "Powerpoint", it's clearly being used as a euphemism for the endless "paper studies" and over-management that infest NASA. You also know that, why are you lying and pretending that you don't?
A radio station I listen to recently rebranded. Their "improved" web site does not deliver content without the WebTrends tracking code being allowed through
That's bizarre. None of the other ABC properties requires trackers to function. Nor even do the sibling digital radio sites, Triple J and Triple J Unearthed. It's weird that Double J is so completely locked down. ABC is not a commercial company, they're government funded non-commercial.
I wonder if Double J is some kind of commercial partnership? Perhaps the ABC has outsourced it? It wouldn't be the first time they've done stupid things like that, but I thought they'd been burned enough to learn their lesson.
[I listen to their streams through my ISP's mirror, so I don't need the site.]
If you don't live in a swing state, a vote for the lesser of 2 evils doesn't really come with a benefit.
If you live in a non-swing state, you join the dominant party (even if you hate them... especially if you hate them) and vote religiously in every single primary you can. And you push your politically disenchanted friends and neighbours to do the same. If the major election is a fait accompli, then the true election is the primary.
One thing I just thought of, it would make a Terminator-style AI launch impossible...unless the AI first injected code into the floppies, then waited for them to be loaded into the various ICBM silos...
No, Skynet doesn't need to get into the silo systems, it only needs to hijack the more centralised authorisation system. Ie, it would send the authorisation codes to all the silos and subs. The silo and sub crews would mostly follow their orders. If even as few as a third of the crews didn't question their orders via another (unauthorised but hander to spoof) channel, you'd still cause Russia to respond, triggering Judgement Day.
It may be possible for the primary missile console. It's in a vault, manned by two specially chosen and armed airmen who are authorised to shoot each other if their partner causes a problem, designed so that it's physically impossible for a single person to operate alone, etc etc. It wouldn't surprise me if they had actual honest-to-god booby-traps in the console itself.
When doing maintenance, you switch out missile ops to the second control room, send in bomb-techs to turn off the booby-traps, then and only then send in your console maintenance tech to replace the malfunctioning board/etc. All under continuous armed guard.
Bypassing a pressure-based system is probably easier than spoofing the missile console. If the attackers are capable of the latter, I wouldn't count of the former defeating them.
But your talking about that does bring up the idea that the attacker might not need access to the hyper-protected console room. There may be a cable running down a utility corridor somewhere. Bypass the anti-tamper system, bypass the console. Your agent only needs to pose as a maintenance person, not a missile airman. It may reduce the difficulty of breaching the silo security. (Hopefully they are paranoid enough that this is impossible, but it wouldn't be the first important security system that invested all their effort in securing the front door.)
The scenario you describe means that they would have had to bypass several layers of physical security and also remove/compromise the two people at the missile command consoles, which are probably also armed.
And that was the point I, and others, were trying to make. The only security was that it was airgapped and had obsessive, paranoid, over the top physical security. The age of the systems (the existence of 8" floppies for example) added nothing to the system security. Because it likely was never designed with any computer security beyond the console itself, bypassing the console could well bypasses the entire missile security system. For example, the two-man activation system, the activation codes, etc. If the security is in the console itself, being able to bypass the console may allow direct access to the missile controls.
In the '70s, that might have meant lugging your own mainframe onto the base, or something equally ridiculous. Today, it's the sort of thing you could hide in a watch.
It may be impossible because of physical security (and god knows I hope it is), but the age of the system adds nothing and may actually have created a false sense of security. A false faith in the system, which often leads to weaker security elsewhere.
The guys who created Stuxnet were not script-kiddies. They knew exactly what they were sabotaging. It seems reasonable to assume that someone attacking a nuclear missile silo is motivated.
With old systems, a lot of the details have likely escaped into the public domain, simply because no-one else realised 30 year old hardware was still being used for such systems. Your own intelligence services probably wouldn't register if someone started collecting information about such old system (the civilian versions, anyway), collecting hardware, discussing it online, and quizzing old timers. They would be noise amongst other old-system hobbyists and collectors.
What that means is, as others have said, the only real security is the well-defended airgap. If someone figures out a way to breach that, you are pwnd.
it would require an attacker to jump through a lot of hoops just to even procure a computer that can read an 8" floppy drive, the drive itself, and the exact media used
When they want to read old wax cylinders, they don't look for a wax cylinder player. They rig up a laser from an old DVD drive, use a bit of software cleverness on a laptop, and read the wax-wobbles via a non-contact interface.
The domains on an 8" disk are, by modern standards, huge. You could not only read them, you could do the equivalent of forensic analysis of layers of previous writes, using the heads salvaged off of a modern HDD. Hand held widget, place over the exposed part of the disk, spin the disk once by hand, everything else is handled in software.
Likewise the idea that there are no "network ports", hence no way for modern systems to get access. This probably also that the whole system has no "network" security, bypass the security console and you have direct access to the entire launch system, because it never occurred to the creators that you could spoof the entire console. (The equivalent of the old Windows password you could bypass by hitting "cancel" on the "Try again: Yes/No/Cancel".) So if someone can smuggle something small past the, probably impressive secured, airgap, there is no second line of defence. Unplug the existing terminal, plug in a tiny portable bit of modern, hard-hacked kit, pwn the whole system.
You might argue that the techniques necessary are not routine hacker knowledge. But Stuxnet was not created by a script-kiddy. They had a deep understanding of the system they were trying to sabotage. This is a nuclear missile silo, you can reasonably assume a motivated attacker.
That's the issue with old planes. As they pass down through the airline food-chain, they tend to inherit owners who don't or can't pay for sufficient maintenance, and tend to "make do".
More generally, the perception of old things being better is that designers and engineers tended to have a more wildly variable margin of error. The bad side was that you had many lemons, flawed by design, the good side was that anything that survived its first years was going to be rock solid. As we got better at understanding the engineering, we could target the lifespan much better. Less lemons, but also less solid rocks. Nice predictable failure curves (usually U-shaped.)
However, the entire specs are available. There are probably physical examples in hardware museums to steal if anything isn't already documented. And you raise no eye-brows (well, prior to the theft) when asking old-timers about such systems.
Once mapped, you can simulate the whole thing on a trivial amount of modern hardware.
Old systems do not provide even security-through-obscurity since information is more likely to have leaked into the public domain. The silos are protected by a well-secured (and probably deeply paranoid) airgap. That's pretty much it.
However, you can simulate an entire old hardware system in software on a tiny new system. Open a panel, swap out a cable and bypass all your antique security entirely.
Slashdot Mirror
one tiny flaw in an otherwise perfect system.
Perfect? Shuttles? [laughs] Oh my god, you're adorable. The shuttles were a hideously compromised design from day one.
The wear and tear on the system from entry and reentry is to high in most cases, where you'll see stress cracks places you might not expect. For every one stress crack you can see how many are forming that you can't see?
Neither Challenger nor Columbia failed because of cumulative fatigue. Challenger's O-Ring was a new part, replaced every time they recycled the SRB casings. Likewise, Columbia's ET and foam were brand new for each launch, and the RCC leading edge on the orbiter would have been holed even if it was brand new.
However, to your point, SpaceX has no intention of flying the same set of engines for 25 years. They will determine the MTBF (say 25-30 launches) and will retire each batch of engines when they reach the point of diminished returns. Same with the tankage, thrust frame, legs, etc. Engines will be tested between launches, as they are already tested three times before launches. The first stage accounts for 70% of the Falcon 9's costs. But testing costs 10% and fuel just 1%. A single reuse will cut a third off their launch costs, two will cut half.
And if reuse somehow ends up costing more than building new... guess what? They won't reuse. They already have the cheapest launcher on the market, they are well on their way to having their own private HLV that'll be every cheaper. With the shuttle, reuse/refurbishment was a requirement. For SpaceX, reuse is gravy.
On the other hand you seem to be completely discounting the wisdom of learning from other people's experience. Having already tried to do this maybe NASA knows something about it. Or should every other rocket maker from NASA to the end of time all repeat the same expensive mistakes just in case they can find a way.
From what I've seen, it's NASA (and their primary contractors) that fail to learn from experience and repeats the same expensive mistakes over and over.
SpaceX started with NASA's cancelled Fastrac engine program. NASA cancelled it, NASA's primary contractors ignored it, SpaceX turned it into a successful business. Hmmm.
SpaceX will have a 50 tonne to LEO HLV in a year or two that they are pricing at less than $100m per launch. Their existing 13 tonne to LEO launcher is priced at less than $60m per launch. And they spent less than $500m to get to their first F9/Dragon launch.
Meanwhile, NASA is struggling to recycle shuttle technology to develop a 60 tonne to LEO launcher (SSME+SRB+Centaur version) for $3b/yr, and maybe $1b per launch, which they hope to launch by 2017. Afterwards, they want to increase that to 70 tonnes by 2021. 130 tonnes by 2027-2030.
For the price of a single year of SLS/Orion development, you could buy 30 FH launches. Say 1,500 tonnes to LEO, for a single year of SLS/Orion funding. For half of just SLS's funding, you could buy 10 FH launches every year. 500 tonnes to LEO each year. And still have half the annual SLS funding to develop something to actually launch. And you could do that for at least seven years before SLS would have allowed NASA to develop a single piece of mission payload.
Doesn't that suggest that maybe, just maybe, that it's NASA doing something wrong?
Exactly, SpaceX thinks *outside the box*. This innovative thinking allows them to toss aside all learned wisdom and knowledge from those old dinosaurs at NASA. Innovative thinking, used to build synergy and form a new paradigm, THAT'S what it's all about. The physical reality will follow from that.
Except this is the opposite of what SpaceX did. They took a tested but cancelled design from NASA for a simplified Kero/LOx engine (Fastrac) and built an engine based on it. They gradually upgraded it, learning as they went. If they hit a wall they hired engineers from NASA and the Primes who'd been involved in rocket development. They incrementally improved their technology, they proved their engines first on the test stand, then, and this is important, they proved them on a single-engine launcher. Which failed three times. Once they solved the problems, they moved to the larger launcher using the same engines. Only once they had proved their 9-engine version, did they take what they had learned and redesigned it to make it better.
Likewise, when they built Grasshopper, they didn't create a bespoke test vehicle with a fundamentally new type of engine (like DC-X), they just used a first stage from their now-flying commercial launcher. That meant that everything they learnt applied directly to their main program, and vice-versa. Each time they fly-back a first stage, they learn more. And it's basically a free test, since the risky stuff only comes after the main mission has staged. F9R tests the re-entry and approach. Grasshopper tests the last-mile and landing. Each getting a little closer to closing the loop.
When NASA created the space shuttle, they threw away everything they learned from the previous two decades. They required a brand new type of engine (LH/LOx), with a radically improved Isp, which was reusable in spite of burning for the entire launch, and had to be powerful enough to loft a 100 tonne space plane to orbit. They required a brand new heat-shield material, which was only theoretically possible, applied in a brand new way. The shape of the space plane had no resemblance to any prior space program or X-Plane research, and right up to that first manned launch in '81, was only theoretically possible. They required a new launch configuration (side-mount), and SRBs larger than anything built before (which were only ever tested in a horizontal static firing. They didn't even do a sub-orbital test launch.) And again, the whole configuration wasn't tested before that first manned launch. There were no intermediate steps, except the Enterprise glider, and no way to do incremental development.
Many excuses are made for this to deflect blame to others, but we can see from subsequent programs that NASA repeated exactly the same mistakes...
They did it with Space Station Freedom, wanting to build a giant all-singing-all-dancing space station with no stepping stones, no test programs, no connection to Skylab, and promising to do it fast and cheap.
Likewise, when they created the shuttle replacement program, VentureStar, they picked an even harder design (SSTO), so that even their sub-scale test vehicle (X-33) required a dozen cutting edge technologies that had never been tried before and were only theoretically possible. (One of which was only developed a decade after the X-33 program was cancelled.) Exactly as they did with the shuttle. Because they totally learned their lesson there.
They did it with MSL-Curiosity by trying to jump ten sizes in one step, which blew out its budget 100%. They are doing it with JWST, which tries to shoe-horn every technology they can into a single program, which has blown out it's budget 400% so far and still going climbing.
Is NASA learning it's lesson? I've been told straight out by NASA science guys that a project is "worthless" unless it includes as much new technology as possible. Oh, there's a few engineers within NASA who get this stuff, and occasionally you see a micro-p
They've still got the cheapest launcher on the market
The ESA's Vega is cheaper per launch to orbit but with smaller payloads
[sigh] Really? That's how you interpreted my comment?
You're comparing a $40m/launch (+$700m/dev for three launches) 2 tonnes to LEO launcher with a $57m/launch (+$500m/dev for 9 launches) 13 tonne to LEO commercial launcher?
Hey, Falcon 1 launched half a tonne and cost $6m.
How low can we go? Is there a nano-sat launcher somewhere that costs $3m?
Idiot.
SpaceX has a bigger promotional budget though.
SpaceX would spend less on their promotional budget than the Primes are spending trying to undermine them.
The problem is that if they launch over the sea and then try to recover a first stage back to land it is going to burn a lot of fuel reversing course before it falls out of the sky
About 30%. Which is part of the reason for the recent F9 v1.1 upgrade. Fuel is cheap, and the rocket equation is kind when adding weight to the first stage. [Mass added to the first stage has vastly less effect on payload than payload mass has on the first stage.]
The first stage is responsible for 70% of the cost of launch, according to Musk. But the fuel is only 1% of the cost. Provided the refurbishment cost can be kept below the build cost (which their Grasshopper program clearly says they can, or why would they be continuing?), they will save money by reuse. Over time, a lot of money.
And if it doesn't work, so what? They've still got the cheapest launcher on the market, and a heavy lifter on the way.
What makes the Merlin engine so interesting is precisely because it is bland. SpaceX hasn't been trying to push the envelope
It surprises me how few people get this. Even people within NASA.
Saturn V was Kero/LOx, and it had a crappy Isp. But it built upon the experience of the previous two decades of rocket (and missile) research.
When NASA went to the shuttle, it threw away everything it had learned and started again. A LH/LOx engine (much harder), with the highest Isp of any engine to date (even harder), which was reusable (seriously?), on a 100 tonne space-plane (I mean, seriously!), launched side-mount on an entirely new configuration. There were no stepping stones, no way to learn your "craft", to understand the limits of materials and techniques. It was completely unrelated to either previous rocket programs or X-Plane research. Much of the proposed design was only theoretically possible, such as the heat-shield, but every single piece had to work right on the very first test launch, manned, in 1981. That is an Apollo level challenge and it's stunning that they were able to get anything that flew, let alone flew for over two decades. But it's not how you build practical systems. It's not how you build affordable systems.
NReitzel is referring to ISS, not Skylab. You know that, why are you Lie Lie Lying and pretending that you don't?
GP is referring to Freedom Space Station, not Skylab. You know that, why are you lying and pretending you don't?
Freedom was proposed to Reagan in '86 and accepted in '88 (continuing under Bush I). It was supposed to cost $8 billion all up and take 8 years.
8 years and $8 billion later, nothing had been launched; hell, nothing had been built. The design had shrunk from a 12 man space station much larger (internally) than Skylab, to a 4 man station much smaller than Skylab. (The in-house joke at the time was that they had to call it "Fred" because they could no longer fit all the letters of "Freedom" on the side.)
So in '96, Clinton forced an enquiry which required them to take what had been designed up to that point, pick one of the three leading configurations, and just build it. This led to the current design, then nicknamed "Alpha". In '98, a requirement was added to merge the station with Russian modules (and to a lesser degree Europa and Japan). This would reduce NASA's cost in developing some of the core modules, get modules launched earlier, and buy access to Mir technology/experience; it was also thought beneficial to keep Russia's space program intact, to prevent rocket engineers going to work for Iran/Iraq/etc. That became the ISS.
As for "Powerpoint", it's clearly being used as a euphemism for the endless "paper studies" and over-management that infest NASA. You also know that, why are you lying and pretending that you don't?
No, NASA management is done in Powerpoint.
(NASA engineering is done in Excel.)
Correction, it's now spread to every ABC site.
Would it kill the President to take a stand here?
A radio station I listen to recently rebranded. Their "improved" web site does not deliver content without the WebTrends tracking code being allowed through
That's bizarre. None of the other ABC properties requires trackers to function. Nor even do the sibling digital radio sites, Triple J and Triple J Unearthed. It's weird that Double J is so completely locked down. ABC is not a commercial company, they're government funded non-commercial.
I wonder if Double J is some kind of commercial partnership? Perhaps the ABC has outsourced it? It wouldn't be the first time they've done stupid things like that, but I thought they'd been burned enough to learn their lesson.
[I listen to their streams through my ISP's mirror, so I don't need the site.]
Yes.
Clinton spent 12 years in the White House
Wow, you can't even get through your first four words without showing how ignorant you are.
If you don't live in a swing state, a vote for the lesser of 2 evils doesn't really come with a benefit.
If you live in a non-swing state, you join the dominant party (even if you hate them... especially if you hate them) and vote religiously in every single primary you can. And you push your politically disenchanted friends and neighbours to do the same. If the major election is a fait accompli, then the true election is the primary.
or simply not voting.
Writing "none of the above" is not voting.
One thing I just thought of, it would make a Terminator-style AI launch impossible...unless the AI first injected code into the floppies, then waited for them to be loaded into the various ICBM silos...
No, Skynet doesn't need to get into the silo systems, it only needs to hijack the more centralised authorisation system. Ie, it would send the authorisation codes to all the silos and subs. The silo and sub crews would mostly follow their orders. If even as few as a third of the crews didn't question their orders via another (unauthorised but hander to spoof) channel, you'd still cause Russia to respond, triggering Judgement Day.
It may be possible for the primary missile console. It's in a vault, manned by two specially chosen and armed airmen who are authorised to shoot each other if their partner causes a problem, designed so that it's physically impossible for a single person to operate alone, etc etc. It wouldn't surprise me if they had actual honest-to-god booby-traps in the console itself.
When doing maintenance, you switch out missile ops to the second control room, send in bomb-techs to turn off the booby-traps, then and only then send in your console maintenance tech to replace the malfunctioning board/etc. All under continuous armed guard.
Bypassing a pressure-based system is probably easier than spoofing the missile console. If the attackers are capable of the latter, I wouldn't count of the former defeating them.
But your talking about that does bring up the idea that the attacker might not need access to the hyper-protected console room. There may be a cable running down a utility corridor somewhere. Bypass the anti-tamper system, bypass the console. Your agent only needs to pose as a maintenance person, not a missile airman. It may reduce the difficulty of breaching the silo security. (Hopefully they are paranoid enough that this is impossible, but it wouldn't be the first important security system that invested all their effort in securing the front door.)
The scenario you describe means that they would have had to bypass several layers of physical security and also remove/compromise the two people at the missile command consoles, which are probably also armed.
And that was the point I, and others, were trying to make. The only security was that it was airgapped and had obsessive, paranoid, over the top physical security. The age of the systems (the existence of 8" floppies for example) added nothing to the system security. Because it likely was never designed with any computer security beyond the console itself, bypassing the console could well bypasses the entire missile security system. For example, the two-man activation system, the activation codes, etc. If the security is in the console itself, being able to bypass the console may allow direct access to the missile controls.
In the '70s, that might have meant lugging your own mainframe onto the base, or something equally ridiculous. Today, it's the sort of thing you could hide in a watch.
It may be impossible because of physical security (and god knows I hope it is), but the age of the system adds nothing and may actually have created a false sense of security. A false faith in the system, which often leads to weaker security elsewhere.
The guys who created Stuxnet were not script-kiddies. They knew exactly what they were sabotaging. It seems reasonable to assume that someone attacking a nuclear missile silo is motivated.
With old systems, a lot of the details have likely escaped into the public domain, simply because no-one else realised 30 year old hardware was still being used for such systems. Your own intelligence services probably wouldn't register if someone started collecting information about such old system (the civilian versions, anyway), collecting hardware, discussing it online, and quizzing old timers. They would be noise amongst other old-system hobbyists and collectors.
What that means is, as others have said, the only real security is the well-defended airgap. If someone figures out a way to breach that, you are pwnd.
it would require an attacker to jump through a lot of hoops just to even procure a computer that can read an 8" floppy drive, the drive itself, and the exact media used
When they want to read old wax cylinders, they don't look for a wax cylinder player. They rig up a laser from an old DVD drive, use a bit of software cleverness on a laptop, and read the wax-wobbles via a non-contact interface.
The domains on an 8" disk are, by modern standards, huge. You could not only read them, you could do the equivalent of forensic analysis of layers of previous writes, using the heads salvaged off of a modern HDD. Hand held widget, place over the exposed part of the disk, spin the disk once by hand, everything else is handled in software.
Likewise the idea that there are no "network ports", hence no way for modern systems to get access. This probably also that the whole system has no "network" security, bypass the security console and you have direct access to the entire launch system, because it never occurred to the creators that you could spoof the entire console. (The equivalent of the old Windows password you could bypass by hitting "cancel" on the "Try again: Yes/No/Cancel".) So if someone can smuggle something small past the, probably impressive secured, airgap, there is no second line of defence. Unplug the existing terminal, plug in a tiny portable bit of modern, hard-hacked kit, pwn the whole system.
You might argue that the techniques necessary are not routine hacker knowledge. But Stuxnet was not created by a script-kiddy. They had a deep understanding of the system they were trying to sabotage. This is a nuclear missile silo, you can reasonably assume a motivated attacker.
assuming proper maintenance
That's the issue with old planes. As they pass down through the airline food-chain, they tend to inherit owners who don't or can't pay for sufficient maintenance, and tend to "make do".
More generally, the perception of old things being better is that designers and engineers tended to have a more wildly variable margin of error. The bad side was that you had many lemons, flawed by design, the good side was that anything that survived its first years was going to be rock solid. As we got better at understanding the engineering, we could target the lifespan much better. Less lemons, but also less solid rocks. Nice predictable failure curves (usually U-shaped.)
However, the entire specs are available. There are probably physical examples in hardware museums to steal if anything isn't already documented. And you raise no eye-brows (well, prior to the theft) when asking old-timers about such systems.
Once mapped, you can simulate the whole thing on a trivial amount of modern hardware.
Old systems do not provide even security-through-obscurity since information is more likely to have leaked into the public domain. The silos are protected by a well-secured (and probably deeply paranoid) airgap. That's pretty much it.
However, you can simulate an entire old hardware system in software on a tiny new system. Open a panel, swap out a cable and bypass all your antique security entirely.
pwnd.