Sure the Ion drive is a really neat addition, but it's soooo slooooow. It's going to take them 15 MONTHS to get there! And the payload isn't really greater at all. It takes longer to get any large loads going. The US space program got people to the moon and back in what...2 weeks? It may be slightly more economical, but it just doesn't seem practical.
It's a lot more economical. It means that they can launch a vehicle on top of a commerical GEO booster, and that the vehicle can make its own way out of geostationary orbit into lunar orbit.
Yes, the Apollo program managed to do the job in four days. But it did it by launching a huge, custom-made multistage vehicle on top of a Saturn 5, which was already a huge, custom-made multistage vehicle. The part of the Apollo spacecraft that actually entered lunar orbit consisted of the lander, the command module, and the service module. The service module was by far the largest component and was mostly fuel tank. It weighed twenty-five tonnes.
SMART I, OTOH, weighs less than half a tonne. Slight difference. It's actually sharing its Ariane 5 booster with a bunch of other satellites.
What I would like to know is what SMART I is going to use for station-keeping around the moon. Luna is lumpy; the part that faces the Earth sticks out, and that makes for really weird gravitation fields around it. There are, basically, no stable orbits around it. If you don't keep adjusting your course you'll either get slung off into space or plough into the ground. Is the ion drive powerful enough to do this usefully, or does it have conventional thrusters as well?
Tube tickets are trackable anyway --- they're paper tickets with magnetic stripes. There's a serial number on the stripe. If you buy your ticket with a credit card, I have no doubt that the serial number gets linked to your credit card number in a database somewhere; which means it's easy to work out that card #2284752 left the Tube at King's Cross and was bought by J. Random Punter.
Frankly, I fail to see what the fuss is about. The Tube is public. You have no expectations of privacy. If you're worried about being tracked, buy your tickets with cash --- but remember your trenchcoat and false beard...
Seriously, to many people, a "space elevator" is going to sound like the "escalator to nowhere" from the Simpsons - a fairly frivolous-sounding projet, and not as inspiring as rockets. Okay, so it'll make space exploration cheaper - what benefits does it have for ordinary people?
Well, once you have cheap access to space, a whole bunch of things suddenly become much more profitable.
Example: most near-Earth asteroids contain very high quantities of heavy metals. There are all sorts of things you can do with iridium, platinum or gold alloys. How would you like a car that ran off ordinary petrol but used a fuel cell instead of an IC engine? Quieter, lighter, cheaper, more reliable --- provided you can get the palladium catalysts required to make it work.
Example: it would be possible to start mass producing things in microgravity. Defect-free crystal growth would lead to much cheaper electronics among other things. If you can get the cost of access cheap enough, even mundane things like steel refining will change: vacuum foam steel girders would be cheaper, lighter and stronger than conventional rolled girders.
Example: Outside geostationary orbit is a great place to be if you want to do something hazardous. Want to build a really messy experimental nuclear power reactor? Now you can do it and it won't be in anyone's back yard.
Example: there's more you can do with a space elevator than get to orbit. They provide an ideal anchoring point for telecommunications systems, among other things: put a communications complex 500km up and you've got LEO-quality satellite communications while still able to use fixed position dishes. Plus it's repairable. Cheaper satellite TV, anyone?
Example: low gee hospitals .
Example: Tourism!
These are just a few examples I can think of off the top of my head --- I'm sure that given a few minutes thought I could come up with some more. The great thing about a space elevator is not that it's directly profitable, but that it's an enabler. It makes a whole bunch of other things become profitable, and opens up the possibility for a whole variety of other industries, currently unthought of, that would be even more profitable. It provides new wealth to the economy, which produces long-term gains in the same way that feeding starving children (although an admirable goal in itself) or building aircraft carriers just don't do. It's the old teach-a-starving-man-to-fish argument: invest, don't spend.
I'm really glad this is getting political room. The shuttle was a waste of money, material and lives from the day it was conceived, and the really sad thing was that everyone involved knew it.
The Russian space industry is doing things right in a way that NASA have never managed. The Russians have focused on making spaceflight boring: so boring, in fact, that the last accident in a Soyuz capsule was in 1971. That's a safety record that makes the shuttle look a bit sick. It also helps that the cost is a tiny fraction of the shuttle; I worked out once that you for the price of a single shuttle launch, you could get the Russians to lift about four times the amount of cargo, plus people, in five seperate vehicles and still have change.
From an engineering point of view, the lesson is painfully obvious: generalisation means compromises. The shuttle is trying to be a heavylifter and a man-rated lifter and a space station and a reentry vehicle, so no wonder it sucks. Much better to focus on small, simple vehicles that do one thing very well.
The Russians have the best man-rated lifter in the world: the Soyuz. It doesn't do much, just takes people from the ground to LEO and back again, but it does it cheaply and reliably. They have the Progress, which I believe is the world's only orbital tug; it can launch, rendezvous with a vehicle, dock, undock and ditch safely, all by remote control. No-one else has anything like it. They have a whole selection of reliable heavylifters, although they are beginning to get competition in that area.
If the Russians with their, ah, mostly broken economy can do it, why are the Americans having so much trouble?
I just wish it were politically feasible for someone with money to just buy the entire Russian space industry, lock stock and barrel, and do some decent investment...
Also, preemption on some processors is an incredible pain: for example, the 6502 has a single, fixed address, 256-byte stack. In order to preempt a process on the 6502 you'd have to manually copy the entire stack somewhere else in memory, and copy another process' stack in. Not pleasant.
What I would love to see is a BSD userland running on a Linux kernel. The BSD userland is so elegant and modular --- I really love the way you can rebuild everything with just one command. OTOH, the Linux kernel has much better hardware support and has some nice features like an automatically sized buffer cache (which I was amazed to discover OpenBSD doesn't, or didn't, support).
One of Linux' big problems is a lack of modularity. Building an entire Linux system, from scratch, is an incredible pain; you have to buy books to explain how. This is particularly annoying when, say, you're building a Linux appliance and want to tweak things. You want to compile with -m686? On BSD you just change one setting, run one makefile and everything rebuilds. On Linux you have to configure a zillion packages independantly.
OTOH, one of BSD's big problems is hardware. I have a Hauppauge Nova-T DVB card. Is it supported under BSD? Need you ask? (Although, surprisingly, BT8x8-based TV cards are supported by OpenBSD.) I have a long-term project to build a PVR. If I could use BSD, I'd go for it like a shot --- it's just so much easier to configure. As it is, I have to go for Linux, which is so much of a pain that I haven't been able to muster the energy to get started yet...
The only problem is (as I see it) that a 32MB CF card (which should hold a full CD at almost CD quality "-q0") currently costs $15. On the other hand, Minidiscs are about $1 per disc, hold more than a CD, can record live audio (analog or digital--no computer required) in realtime, can edit the tracks on the fly, has better sound quality than Ogg at even the highest quality settings, has a longer battery life, puts off less heat, never skips, can be rewritten more times than a CF card, are more physically durable than a CF card, etc.
Actually, Minidiscs only store about 150MB. Standard MD ATRAC compression is around 300kb/s; MP3 and ogg can easily beat this --- just crank up the bit rate (although ATRAC does appear to be a technically superior compression algorithm).
And they're certainly not robust. I mean, come on, moving parts. An equivalent CF-based player will be smaller and tougher and consume much less power (no disk to keep spinning).
But that said, I really like Minidiscs. They're good enough for most purposes. They're cheap, as you say. I really like being able to losslessly edit on any device, even the smallest handheld player (even if the filesystem does suck and lose free space --- it can't coalesce disjoint blank areas shorter than about 12 seconds, it just loses them). I just wish that Sony had pushed data drives a bit harder; they'd make excellent floppy/ZIP disk replacements.
Lots and lots of technical information can be found on this FAQ page. Interesting reading.
BTW, how did you manage to fit a CD's worth of music onto a 32MB CF card? That must be some compression algorithm!
And no, I'm not being melodramatic. To be useful it needs to be 100%...
Um, you can't get 100% hydrogen peroxide. It exists in equilibrium with water; above a certain critical point it spontaneously (and slowly) decomposes to produce water and dissolved oxygen.
In fact, peroxide is a really great rocket fuel. It's cheap. It's easy to handle. It's environmentally friendly. It can be used in monoprop and biprop engines, depending on what you do with it. It's hypergolic, which means it's trivial to build restartable engines (the shuttle's engines aren't restartable; they can only start with assistance from the ground). It's safe, too --- much safer than hydrazine, the most common hypergolic fuel, which is horribly poisonous, carcinogenic and can be unstable, to boot.
Yes, hydrogen peroxide can be nasty. It's a rocket fuel, for gods' sake --- it's supposed to decompose violently. You just have to be careful, and it's a hell of a lot easier to manage than stuff like liquid oxygen. Now, that stuff really is painful to handle.
Peroxide isn't the best fuel; it's got a specific impulse of only about 160-190 seconds when used as a monoprop, but so does hydrazine. And, if you use it as a biprop with kerosene, it goes up to 200-230, which means your ship can have one small tank of kerosene for the main engines and one large tank of peroxide which runs the main engines plus the thrusters. Compare with the shuttle, which uses loads of different fuel types, each with their own storage and delivery systems.
(The best fuels on the referenced page are in the region of 300 to 385. Hydrogen and flourine. Ack!)
But hydrogen peroxide is the perfect choice for a small setup like Armadillo. All you need are a few simple safety precautions --- bleeder valves, non-reactive storage facilities, some basic technical expertise in handling the stuff --- and you're fine.
There is always some sort of plan B - in this case the most obvious one is dock with the ISS and look to the other shuttles or the Russians for extraction.
Actually, docking with the station was completely out of the question. Firstly, the shuttles have bugger all orbital manouevering capability; once the main engines are shut down, that's it. (They don't restart.) All you have is the OMS, which is strictly low power. Did you know that it takes the same amount of delta-vee to change your orbital plane by ninety degrees (that is, polar to equatorial or vice versa) as it does to launch in the first place? Columbia was nowhere near the station, and there was nowhere for it to go.
Secondly, SpaceLab was in the cargo hold connected to the airlock so they couldn't dock to anything anyway.
It may have been possible to launch another shuttle. The crash launch programme takes a bit over a week. Had they known for sure that there was a problem immediately after launch, this could have been done; Columbia's mission was 17 days, plenty of time.
But there wasn't really any way they could have known. Debris falls off the shuttles all the time, and it's been carefully investigated and --- up until now --- hasn't been a problem. The crew might have done a spacewalk and manually examined the bottom of the ship, but ST-107, being a SpaceLab mission, didn't carry jet packs; the bottom of the space shuttles is completely smooth. No hand holds. And I don't know if you can get off the shuttle at all when SpaceLab is installed.
No, Columbia's loss was just one of those things. You do your best, take every reasonable precaution, but sometimes... things go wrong.
Admittedly, things weren't helped by the shuttle's poor design. Wheel wells? Opening through the heat shield? Definitely an accident waiting to happen. Those tiles are a really bad idea, too. Ablative shielding is old tech and Just Works. The Russians use simple, dumb capsules and have never lost one on reentry due to heat shield failure. (They did lose a crew when a valve jammed open and the astronauts died in vacuum, but the capsule still landed perfectly.) For years their heat shields were made out of oak.
Re:Some words it needs to attract the slashdot cro
on
A Word a Day
·
· Score: 2, Funny
Perhaps my all time favorite common misspelling is "retarted" when used as an insult.
wier'd, verb, past tense (archaic). To turn someone into a small dam.
retarted, verb, past tense. Bill Gates after he has just been hit with another cream pie.
Just out of curiosity why is that? Are you afraid that in a couple of years Sony will abandon the memory stick and leave to stuck unable to get new ones? Do you want to use your memory stick in other devices? Do you just not like the concept of a "proprietary (whatever that means in this context)" format?
128MB Memory Stick: 58.73 UKP (not inc. VAT)
128MB Compact Flash: 30.27 UKP (not inc. VAT)
Both prices from Dabs. (Normally, I buy memory from Crucial, but they don't stock memory sticks.)
So: memory sticks are expensive, hard to get, and only work in Sony devices. That's a pretty fair reason not to use them in my book.
I downloaded one of the tests for the Mac a long, long time ago. Gameplay seemed intriguing, but what really hooked me was the music by Stefan Poiss (I play the three tracks on my MP3 player over and over). Anybody know if there's other tracks available by him? I haven't checked the site in a while and it seems to be well and truly slashdotted now.
Good, isn't it? Alas, he hasn't updated his MP3.com web site since 2001. A web search doesn't show up much for him, either. I think he just doesn't use the 'net much.
If anyone finds anything more by him, do let us know...
... setting up licensed cell-phone free zones (such as concert halls etc) where phones cannot ring. The zones could have a small very low-power transmitter/scrambler that would inform/interrupt the cell phone so that it just wouldn't ring in those areas.
A better solution than just blocking everything would be to set up a microcell inside the theatre/concert hall/etc. Any calls to a phone inside the microcell get routed to the theatre/auditorium/whatever's reception, where a message can be left. If it's a genuine emergency, the message can be forwarded on.
I gather this is actually possible, but I don't know why no-one's tried it.
I remember hearing, but can't verify, a particularly inspired hack on, IIRC, a Pioneer: something needed adjustment, possibly an antenna arm. So they turned on all the heaters around the part in question. The resin softened. Then they fired the thrusters to jolt the spacecraft, and switched off the heaters. The arm bent into the correct position, and then hardened.
I suspect that it's a myth, but damn, it ought to be true.
Incidentally, did you know that the Pioneer computers were so simple they didn't have any jump instructions? They just executed all the instructions in memory one sweep after another. Conditionals were done by masking out blocks of code using condition codes. Slow, yes; but the processor could be implemented in a handful of radiation hardened transistors, and if the computer ever reset spontaneously due to, e.g., passing through Jupiter's magnetosphere, noone cared.
In fact, get an old PDA. New ones tend to eat batteries with things like colour screens and backlights. Also, they use built-in rechargable batteries that require special equipment to charge.
The ideal travel computer is something like an early Palm Pilot. Mono screen, no backlight. No moving parts. Runs off AAA batteries; you can get new disposables practically everywhere, and you can take some rechargables and a solar panel for those times when you can't get them. You can even get keyboards for them. One of these plus a cell phone or sat phone will give you (slow, expensive) 'net access everywhere. Also, they're cheap; drop one in a river and you're not watching a thousand currency units of your choice bubble gently.
I'm rather tempted by the AlphaSmart Dana. Palm device with a real keyboard, 560x160 screen, ports, runs of AA batteries (which are even easier to come by than AAA)... Disclaimer: I've never seen one. They might be crap.
"Beni Cherniavsky mentioned a very intriguing counterpart to bitrate peeling. If you have a peeler that saves the bits it chopped off, you could reconstitute the higher quality files by adding the missing bits to the lower quality file. This idea could lead to a music download service where you can download a low quality preview version of a song, and if you are interested, download the missing bits to make it a high quality version."
Or, even more interesting: peel a Vorbis file all the way down to the minimum quality. Concatenate the bits together in order. Now you have a file that you can play back, in its entirety, when it's only 10% downloaded. All you have to do is wait for the minimum quality version to download; from then on, the entire file is playable. It's just that the longer you wait, the more peels get added, the higher the quality... holographic audio downloading.
Whaaaat? This is absolute BS. If it's 105, that means it was built around 1897. All but 5 states were admitted to the Union as of 1896. Check out this site [consultawebsurfer.com]
Oops --- mea culpa. It seems that in my caffeine-deprived state I was under the impression that 1889 > 1896. What can I say? It's not my country.
Okay, so it's older than five states. And bits of it are older than Utah. That's still pretty impressive...
I will second the Eiffel Tower. It is a whole lot bigger in person than it is in the pictures. When I visited Paris, I only got to the lower deck because it was too windy --- and that was a hell of a long way up.
The entire tower's only 324m, but it's such an open structure that it makes you painfully aware of every metre of it. It was quite the most vertiginous thing I've ever done. I regret that I never made it to the top.
By the way, I should point out that this year it's 105 years old. There are twelve states in the US that old...
Re:Not very well-explained nor convincing
on
Concept Programming
·
· Score: 2
The next example was discussing takinga derivative and how you can translate some incorrect Java syntax that takes a derivative into the Java equivalent. Why not write a method to do this? What is to be gained by using a non-standard syntax? It makes it harder to write (you have to learn something in addition to Java), and harder to read (same reason).
Because taking a derivative is not a function. It's a metafunction; it is applied to a function and returns another function. For example: the derivative of sin(x) is cos(x), for all x. This can't be implemented as a method and so using method syntax would be nonsensical.
I agree that I just didn't get the point of the Minimum example; but the meko thin-tool application (which is used for the above transformation) looked really cool. Program transformation is a hot topic in the functional programming world; I'd like to see what it can do in the procedural world.
Oh, yeah, and all programs can be modelled as a function; all Turing-complete languages are equivalent, so a program can be transformed from one language to any other, including the purely functional languages. This is the entire basis behind a lot of computational theory, and so their mathematical representation makes a lot of sense.
Making an embedded Linux is easy --- fundamentally it consists of three binaries: the kernel, ash, and busybox.
Unfortunately, when you add functionality it gets real complicated real quick. I tried putting together a turnkey Mozilla distro with an embedded Linux, KDrive X, and Mozilla. Did it work? Did it hell. Mozilla would just hang on startup, waiting for some service that wasn't there. I spent ages trying to find out why and eventually had to give up.
I haven't looked closely at QPlus, because it's in Korea and is Slashdotted out of its tiny mind. What I'd really like is a source-based system, where I can just type 'make' in the top level and it will rebuild everything I'm using, libc and all. Unfortunately the review talks about RPMs, so I suspect it isn't.
Has anyone here actually used QPlus and can comment on it?
Hrmph. 44.1kHz just doesn't cut it for me; I prefer my audio at 196kHz, 48 bits per sample, 6.1 channels, and I resent paying for anything less. But then again my computer is a Thinking Machine CM5, which doesn't even come with a sound card, so I guess it's all academic for me, too.
Oh, come on. If you've got a CM5, you don't need a sound card. It's fast enough so you can exploit the race conditions between the universal cellular calculation states to hack reality and cause the audio to spontaneously manifest via Brownian motion in your room.
(You can also use the same effect to revitalise flat cola. It's well worth looking into, although you do need to download a kernel patch.)
Lo, many years ago I had a lot of luck with EARS on my 66MHz 486. It's a very simple discrete trainable recogniser; you have to teach it every word before it would recognise it. But it was fast then, it should be really fast now, and was pretty decent for recognising simple commands.
Slashdot Mirror
It's a lot more economical. It means that they can launch a vehicle on top of a commerical GEO booster, and that the vehicle can make its own way out of geostationary orbit into lunar orbit.
Yes, the Apollo program managed to do the job in four days. But it did it by launching a huge, custom-made multistage vehicle on top of a Saturn 5, which was already a huge, custom-made multistage vehicle. The part of the Apollo spacecraft that actually entered lunar orbit consisted of the lander, the command module, and the service module. The service module was by far the largest component and was mostly fuel tank. It weighed twenty-five tonnes.
SMART I, OTOH, weighs less than half a tonne. Slight difference. It's actually sharing its Ariane 5 booster with a bunch of other satellites.
What I would like to know is what SMART I is going to use for station-keeping around the moon. Luna is lumpy; the part that faces the Earth sticks out, and that makes for really weird gravitation fields around it. There are, basically, no stable orbits around it. If you don't keep adjusting your course you'll either get slung off into space or plough into the ground. Is the ion drive powerful enough to do this usefully, or does it have conventional thrusters as well?
Frankly, I fail to see what the fuss is about. The Tube is public. You have no expectations of privacy. If you're worried about being tracked, buy your tickets with cash --- but remember your trenchcoat and false beard...
Well, once you have cheap access to space, a whole bunch of things suddenly become much more profitable.
Example: most near-Earth asteroids contain very high quantities of heavy metals. There are all sorts of things you can do with iridium, platinum or gold alloys. How would you like a car that ran off ordinary petrol but used a fuel cell instead of an IC engine? Quieter, lighter, cheaper, more reliable --- provided you can get the palladium catalysts required to make it work.
Example: it would be possible to start mass producing things in microgravity. Defect-free crystal growth would lead to much cheaper electronics among other things. If you can get the cost of access cheap enough, even mundane things like steel refining will change: vacuum foam steel girders would be cheaper, lighter and stronger than conventional rolled girders.
Example: Outside geostationary orbit is a great place to be if you want to do something hazardous. Want to build a really messy experimental nuclear power reactor? Now you can do it and it won't be in anyone's back yard.
Example: there's more you can do with a space elevator than get to orbit. They provide an ideal anchoring point for telecommunications systems, among other things: put a communications complex 500km up and you've got LEO-quality satellite communications while still able to use fixed position dishes. Plus it's repairable. Cheaper satellite TV, anyone?
Example: low gee hospitals .
Example: Tourism!
These are just a few examples I can think of off the top of my head --- I'm sure that given a few minutes thought I could come up with some more. The great thing about a space elevator is not that it's directly profitable, but that it's an enabler. It makes a whole bunch of other things become profitable, and opens up the possibility for a whole variety of other industries, currently unthought of, that would be even more profitable. It provides new wealth to the economy, which produces long-term gains in the same way that feeding starving children (although an admirable goal in itself) or building aircraft carriers just don't do. It's the old teach-a-starving-man-to-fish argument: invest, don't spend.
The Russian space industry is doing things right in a way that NASA have never managed. The Russians have focused on making spaceflight boring: so boring, in fact, that the last accident in a Soyuz capsule was in 1971. That's a safety record that makes the shuttle look a bit sick. It also helps that the cost is a tiny fraction of the shuttle; I worked out once that you for the price of a single shuttle launch, you could get the Russians to lift about four times the amount of cargo, plus people, in five seperate vehicles and still have change.
From an engineering point of view, the lesson is painfully obvious: generalisation means compromises. The shuttle is trying to be a heavylifter and a man-rated lifter and a space station and a reentry vehicle, so no wonder it sucks. Much better to focus on small, simple vehicles that do one thing very well.
The Russians have the best man-rated lifter in the world: the Soyuz. It doesn't do much, just takes people from the ground to LEO and back again, but it does it cheaply and reliably. They have the Progress, which I believe is the world's only orbital tug; it can launch, rendezvous with a vehicle, dock, undock and ditch safely, all by remote control. No-one else has anything like it. They have a whole selection of reliable heavylifters, although they are beginning to get competition in that area.
If the Russians with their, ah, mostly broken economy can do it, why are the Americans having so much trouble?
I just wish it were politically feasible for someone with money to just buy the entire Russian space industry, lock stock and barrel, and do some decent investment...
Also, preemption on some processors is an incredible pain: for example, the 6502 has a single, fixed address, 256-byte stack. In order to preempt a process on the 6502 you'd have to manually copy the entire stack somewhere else in memory, and copy another process' stack in. Not pleasant.
One of Linux' big problems is a lack of modularity. Building an entire Linux system, from scratch, is an incredible pain; you have to buy books to explain how. This is particularly annoying when, say, you're building a Linux appliance and want to tweak things. You want to compile with -m686? On BSD you just change one setting, run one makefile and everything rebuilds. On Linux you have to configure a zillion packages independantly.
OTOH, one of BSD's big problems is hardware. I have a Hauppauge Nova-T DVB card. Is it supported under BSD? Need you ask? (Although, surprisingly, BT8x8-based TV cards are supported by OpenBSD.) I have a long-term project to build a PVR. If I could use BSD, I'd go for it like a shot --- it's just so much easier to configure. As it is, I have to go for Linux, which is so much of a pain that I haven't been able to muster the energy to get started yet...
Actually, Minidiscs only store about 150MB. Standard MD ATRAC compression is around 300kb/s; MP3 and ogg can easily beat this --- just crank up the bit rate (although ATRAC does appear to be a technically superior compression algorithm).
And they're certainly not robust. I mean, come on, moving parts. An equivalent CF-based player will be smaller and tougher and consume much less power (no disk to keep spinning).
But that said, I really like Minidiscs. They're good enough for most purposes. They're cheap, as you say. I really like being able to losslessly edit on any device, even the smallest handheld player (even if the filesystem does suck and lose free space --- it can't coalesce disjoint blank areas shorter than about 12 seconds, it just loses them). I just wish that Sony had pushed data drives a bit harder; they'd make excellent floppy/ZIP disk replacements.
Lots and lots of technical information can be found on this FAQ page. Interesting reading.
BTW, how did you manage to fit a CD's worth of music onto a 32MB CF card? That must be some compression algorithm!
Um, you can't get 100% hydrogen peroxide. It exists in equilibrium with water; above a certain critical point it spontaneously (and slowly) decomposes to produce water and dissolved oxygen.
In fact, peroxide is a really great rocket fuel. It's cheap. It's easy to handle. It's environmentally friendly. It can be used in monoprop and biprop engines, depending on what you do with it. It's hypergolic, which means it's trivial to build restartable engines (the shuttle's engines aren't restartable; they can only start with assistance from the ground). It's safe, too --- much safer than hydrazine, the most common hypergolic fuel, which is horribly poisonous, carcinogenic and can be unstable, to boot.
Yes, hydrogen peroxide can be nasty. It's a rocket fuel, for gods' sake --- it's supposed to decompose violently. You just have to be careful, and it's a hell of a lot easier to manage than stuff like liquid oxygen. Now, that stuff really is painful to handle.
Peroxide isn't the best fuel; it's got a specific impulse of only about 160-190 seconds when used as a monoprop, but so does hydrazine. And, if you use it as a biprop with kerosene, it goes up to 200-230, which means your ship can have one small tank of kerosene for the main engines and one large tank of peroxide which runs the main engines plus the thrusters. Compare with the shuttle, which uses loads of different fuel types, each with their own storage and delivery systems.
(The best fuels on the referenced page are in the region of 300 to 385. Hydrogen and flourine. Ack!)
But hydrogen peroxide is the perfect choice for a small setup like Armadillo. All you need are a few simple safety precautions --- bleeder valves, non-reactive storage facilities, some basic technical expertise in handling the stuff --- and you're fine.
Actually, docking with the station was completely out of the question. Firstly, the shuttles have bugger all orbital manouevering capability; once the main engines are shut down, that's it. (They don't restart.) All you have is the OMS, which is strictly low power. Did you know that it takes the same amount of delta-vee to change your orbital plane by ninety degrees (that is, polar to equatorial or vice versa) as it does to launch in the first place? Columbia was nowhere near the station, and there was nowhere for it to go.
Secondly, SpaceLab was in the cargo hold connected to the airlock so they couldn't dock to anything anyway.
It may have been possible to launch another shuttle. The crash launch programme takes a bit over a week. Had they known for sure that there was a problem immediately after launch, this could have been done; Columbia's mission was 17 days, plenty of time.
But there wasn't really any way they could have known. Debris falls off the shuttles all the time, and it's been carefully investigated and --- up until now --- hasn't been a problem. The crew might have done a spacewalk and manually examined the bottom of the ship, but ST-107, being a SpaceLab mission, didn't carry jet packs; the bottom of the space shuttles is completely smooth. No hand holds. And I don't know if you can get off the shuttle at all when SpaceLab is installed.
No, Columbia's loss was just one of those things. You do your best, take every reasonable precaution, but sometimes... things go wrong.
Admittedly, things weren't helped by the shuttle's poor design. Wheel wells? Opening through the heat shield? Definitely an accident waiting to happen. Those tiles are a really bad idea, too. Ablative shielding is old tech and Just Works. The Russians use simple, dumb capsules and have never lost one on reentry due to heat shield failure. (They did lose a crew when a valve jammed open and the astronauts died in vacuum, but the capsule still landed perfectly.) For years their heat shields were made out of oak.
wier'd, verb, past tense (archaic). To turn someone into a small dam.
retarted, verb, past tense. Bill Gates after he has just been hit with another cream pie.
128MB Memory Stick: 58.73 UKP (not inc. VAT)
128MB Compact Flash: 30.27 UKP (not inc. VAT)
Both prices from Dabs. (Normally, I buy memory from Crucial, but they don't stock memory sticks.)
So: memory sticks are expensive, hard to get, and only work in Sony devices. That's a pretty fair reason not to use them in my book.
Good, isn't it? Alas, he hasn't updated his MP3.com web site since 2001. A web search doesn't show up much for him, either. I think he just doesn't use the 'net much.
If anyone finds anything more by him, do let us know...
Wanna bet?
Google search for 'C interpreter' (767000 hits)
A better solution than just blocking everything would be to set up a microcell inside the theatre/concert hall/etc. Any calls to a phone inside the microcell get routed to the theatre/auditorium/whatever's reception, where a message can be left. If it's a genuine emergency, the message can be forwarded on.
I gather this is actually possible, but I don't know why no-one's tried it.
I suspect that it's a myth, but damn, it ought to be true.
Incidentally, did you know that the Pioneer computers were so simple they didn't have any jump instructions? They just executed all the instructions in memory one sweep after another. Conditionals were done by masking out blocks of code using condition codes. Slow, yes; but the processor could be implemented in a handful of radiation hardened transistors, and if the computer ever reset spontaneously due to, e.g., passing through Jupiter's magnetosphere, noone cared.
And they're still going...
A laptop? Forget it. Get a PDA.
In fact, get an old PDA. New ones tend to eat batteries with things like colour screens and backlights. Also, they use built-in rechargable batteries that require special equipment to charge.
The ideal travel computer is something like an early Palm Pilot. Mono screen, no backlight. No moving parts. Runs off AAA batteries; you can get new disposables practically everywhere, and you can take some rechargables and a solar panel for those times when you can't get them. You can even get keyboards for them. One of these plus a cell phone or sat phone will give you (slow, expensive) 'net access everywhere. Also, they're cheap; drop one in a river and you're not watching a thousand currency units of your choice bubble gently.
I'm rather tempted by the AlphaSmart Dana. Palm device with a real keyboard, 560x160 screen, ports, runs of AA batteries (which are even easier to come by than AAA)... Disclaimer: I've never seen one. They might be crap.
Or, even more interesting: peel a Vorbis file all the way down to the minimum quality. Concatenate the bits together in order. Now you have a file that you can play back, in its entirety, when it's only 10% downloaded. All you have to do is wait for the minimum quality version to download; from then on, the entire file is playable. It's just that the longer you wait, the more peels get added, the higher the quality... holographic audio downloading.
Oops --- mea culpa. It seems that in my caffeine-deprived state I was under the impression that 1889 > 1896. What can I say? It's not my country.
Okay, so it's older than five states. And bits of it are older than Utah. That's still pretty impressive...
The entire tower's only 324m, but it's such an open structure that it makes you painfully aware of every metre of it. It was quite the most vertiginous thing I've ever done. I regret that I never made it to the top.
By the way, I should point out that this year it's 105 years old. There are twelve states in the US that old...
$ units
1991 units, 71 prefixes, 32 nonlinear units
You have: 15 rods / hogshead
You want: miles / gallon
* 0.00074404911
/ 1343.9973
There you go; 15 rods to the hogshead is 0.0007 US miles to the gallon.
...do exactly this.
Because taking a derivative is not a function. It's a metafunction; it is applied to a function and returns another function. For example: the derivative of sin(x) is cos(x), for all x. This can't be implemented as a method and so using method syntax would be nonsensical.
I agree that I just didn't get the point of the Minimum example; but the meko thin-tool application (which is used for the above transformation) looked really cool. Program transformation is a hot topic in the functional programming world; I'd like to see what it can do in the procedural world.
Oh, yeah, and all programs can be modelled as a function; all Turing-complete languages are equivalent, so a program can be transformed from one language to any other, including the purely functional languages. This is the entire basis behind a lot of computational theory, and so their mathematical representation makes a lot of sense.
Unfortunately, when you add functionality it gets real complicated real quick. I tried putting together a turnkey Mozilla distro with an embedded Linux, KDrive X, and Mozilla. Did it work? Did it hell. Mozilla would just hang on startup, waiting for some service that wasn't there. I spent ages trying to find out why and eventually had to give up.
I haven't looked closely at QPlus, because it's in Korea and is Slashdotted out of its tiny mind. What I'd really like is a source-based system, where I can just type 'make' in the top level and it will rebuild everything I'm using, libc and all. Unfortunately the review talks about RPMs, so I suspect it isn't.
Has anyone here actually used QPlus and can comment on it?
Oh, come on. If you've got a CM5, you don't need a sound card. It's fast enough so you can exploit the race conditions between the universal cellular calculation states to hack reality and cause the audio to spontaneously manifest via Brownian motion in your room.
(You can also use the same effect to revitalise flat cola. It's well worth looking into, although you do need to download a kernel patch.)
Lo, many years ago I had a lot of luck with EARS on my 66MHz 486. It's a very simple discrete trainable recogniser; you have to teach it every word before it would recognise it. But it was fast then, it should be really fast now, and was pretty decent for recognising simple commands.