There are several things to realize about this prize. First, the rover is very roughly a third of the work. I'd break into getting to LEO, getting to the lunar surface, and all the stuff on the surface (rover, video, communication, etc.).
If you're trying to do this on a budget comparable to the prize, each of those is very challenging. If you buy your orbital launch, the cheapest option is probably a SpaceX Falcon 1, which starts at $7M -- a third of your budget already. That means you get *one* attempt. This prize won't be won on the first flight of the hardware, not with a budget even approaching the $20M purse.
Getting from (Earth) orbit to the surface is tricky, but probably the easiest piece. Carmack is very close to demonstrating a large fraction of that with Pixel at the Lunar Lander Challenge in October. Left to do would be nontrivial navigation and a nontrivial performance boost. Here, buying the hardware you need certainly isn't off the shelf, but most of the pieces might be available. I suspect you'd find yourself blowing another large fraction of your budget even before the requisite development on this part.
The lunar rover and communications presents another set of challenges, which it sounds like CMU may well have experience with.
But, I'd say hiring NASA engineers is the wrong way to win this on a budget. NASA couldn't even begin to touch this prize for $100M. If you hire engineers who are used to working with budgets on a NASA size, you'll get a solution that costs NASA price tags, or close to them. If you want to spend a couple hundred million winning the prize, just to prove you can, it'll work -- but I would say that's kind of silly. I don't think this prize will be won for less than $20M, but I think it will be won for not a huge amount more.
Personally, I think Carmack and the rest of the people at Armadillo Aerospace are much more interesting to watch. If he continues at his current pace, he'll have hardware in LEO long before this prize expires, and on a much smaller budget than anyone has done before. And he's already been talking about what would be needed to win this prize. If you want to watch the interesting show, don't look to the people that say they'll do it the old way -- look to the people that want to do it orders of magnitude cheaper than it's been done before, by turning every piece of conventional wisdom on its head, and are busy proving they can rather than trumpeting their barely formed plans to the press.
Oddly enough, that's not something I've had an issue with.
All the data I've needed to analyze usually comes in the form of nice simple CSV text files, or comes from a program that can optionally output as such. So the import is almost irrelevant.
But then actually doing something with that data is where Excel wins out (I won't say it's good -- it's sadly lacking in some respects -- but it beats OOo hands down). I most emphatically don't want online LAMP type applications; nothing I've done has been of the "plug numbers in here, read answers there" form. It's all engineering problems, which usually end up as one-off spreadsheets intended for a specific purpose. I've done everything from electronic circuit design to data analysis to finite element analysis to thermodynamic modeling in Excel, without any particular difficulty. Some of those could have been done in OOo, though they would have been harder. Some couldn't have (at least, not reasonably). And the major differences are in the goal seek / solver tools, functions, circular references, and performance.
I'm sure there are applications where OOo is comparable or better -- but none of them are the applications I work with, at least not when it comes to the "advanced" or complicated things. (There's a decent amount of quick and dirty scratch work that I do in a spreadsheet, and for that I don't really care which I'm working in, they're both fine.)
Can anyone out there make a convincing case that Calc or Gnumeric are just as good as Excel, even for advanced users?
(I haven't worked with Gnumeric much, so this mostly applies to Calc)
I would, except that they're not even close. I run Linux, I use OpenOffice.org, and at work I've done some serious Excel work. There are a couple features I like in Calc, but nothing major. However, there are a lot of things that Excel does better. Graphs are a big one. Circular reference handling and goal seek / solver are also important. And last but certainly not least... handling of giant spreadsheets. That even Excel could be better at, but it's less painful than Calc. Complicated math like statistics and Fourier work is also lacking, and VB scripting is important for function definitions (though I'd love to use something better).
I've tried to do complicated stuff in Calc, but never for too long -- it just gets painful.
$30M is at least enough to attract interest. John Carmack has already started talking about what would be required to do it. And given that he's the odds-on favorite to win the Lunar Lander Challenge (both levels) this year at the X-Prize Cup, I'd say that's rather interesting. It sounds like he thinks the toughest problem to solve is navigation.
On the other hand, a lot of what has been happening lately probably *is* unhelpful. For example, banning anything pointy (nail clippers?) from airplanes -- what prevents another 9/11 is a change in attitude among passengers, not a complete lack of sharp things.
That said, there are plenty of things I oppose because I find the tradeoff abhorrent -- like the eavesdropping and other privacy invasions, and the imprisonment without due process, and...
That excuses the fluff (somewhat, but I've never really minded the "random geek does something irrelevant but neat" type of fluff), but it most assuredly does not excuse the rampant FUD and trolling.
I want a $39.95 plan to actually cost $39.95. As in, that's the number at the bottom of the bill that I have to pay each month.
I don't want to pay "regulatory surcharges" or "cost recovery fees" or anything else that isn't included in the advertised price. And this goes for all these sorts of contracts, not just cell phones.
It's a radio telescope; the atmosphere is almost irrelevant. This gives a very large effective size for diffraction purposes, meaning the resulting images can be much more finely resolved.
The issue of whether or not to put some sort of dust-clearing device on the panels was examined critically and decided on early in this project. In short: they didn't know what dust storms would do to the panels; it turns out they tend to remove dust. Several options for dust clearing were considered -- wipers, electrostatic techniques, peel-away plastic, and probably others I've forgotten. All of them would have *probably* worked, and all of them would have taken up space and weight. Essentially it came down to choosing between dust removal and an instrument. Faced with that decision, they decided that better quality, more complete data was more interesting than having the rovers run longer.
Of course, they got lucky, and the dust storms seem to clear dust off the panels. So there was even less need for dust-clearing than they thought there might be.
The obvious ones: higher energy density (meaning smaller and lighter, or longer life between charges), higher efficiency (and therefore less heat generation on charge / discharge), no flammable components, higher charge / discharge rates (and therefore faster charge / higher power density). All but the first (energy density) are available now, and they only lag batteries by about a factor of 10 in energy density. But if the stuff people have in labs turns into real products, or the rumors in the article I cited are correct, we'll see that last part change in the relatively near future (1-5 years, depending who you listen to). Also, don't underestimate the value of instant charging -- wouldn't it be neat to drop your cell phone on the charger for 30 seconds whenever it's convenient and have it be fully charged? Or your laptop charge in 5 minutes instead of 2 hours? That convenience factor is at least interesting, if not earth-shattering.
I mean exactly what you think I mean -- used in some applications, but not the most performance-critical ones. In the same way that NiMH batteries are available and in use now, I expect some form of LiIon to be available and in use even once supercaps are widespread.
How are they solving a different problem? Obviously right now they're not suited to replace LiIon batteries, and they're getting used to solve other problems, but if the energy density catches up to LiIons, then won't ultracaps replace them as soon as they're cost effective?
Fortunately, we have supercapacitors. While they're not there yet for energy density (still about 10x too little) they're rapidly improving. 10x isn't much at the rate these things have been improving, and there are plenty of labs with pieces that are much better than currently available commercial offerings, but that still need development work. If I had to guess, I'd say it's 5 years until the first supercaps appear in serious commercial use, and less than 10 until LiIon has gone the way of NiMH.
Of course, if you believe the rumors then it might be even faster than that -- we might be seeing serious applications in a year or so.
I, for one, will be glad to give LiIon a proper burial. But until then, we work with what's available.
Why is the entire front page populated with stories by kdawson? Did the rest of the "editors" quit or something? It'd be nice to have more of a mix of stories on occasion.
And BTW, no on ever believed that Saddam plotted 9/11
You say that, but polls have shown that over 50% of Americans think he was connected to Al Qaeda. I don't know what fraction think he was involved in 9/11, but I bet it's far higher than 0%.
It strikes me that the problem is that most spell checkers try to check everything, and that a lot of code has things that really shouldn't be spell checked at all, mixed with things that should. I imagine that one way to start would be to only alert on those errors that are almost correct -- if it looks like garbage, ignore it, but if it's close, assume it should be right. Perhaps ignore prefixes / suffixes as well -- pSomething is fine, pSometihng isn't. Also, CamelCase ought to be easy enough to detect -- treat it as word boundaries, and spell check the individual words. Again, egregious misspellings probably aren't -- nextObjFoo is ok, even though Obj isn't a word -- it's so far from being a word that we assume the programmer meant it that way.
Similarly, there should probably be a set of words added that aren't "English" but are used often enough to be worth adding to the dictionary. Things like Obj, Int, and Ptr.
I think the reason such spell checkers don't exist already is fairly simple -- everyone just assumes they're impossible, and doesn't try. Couple that with the fact that a mediocre quality one would be so annoying as to be worse than useless, and you have a recipe for a program that won't get written. I don't think either of those would have to be the case if someone sufficiently clever decided to attack the problem, though.
You can't build a rover that tiny and have it last as long as Spirit and Opportunity have. The reason is the same one this latest round of dust storms was worrying -- heat. Batteries need to stay warm; if they freeze, they are severely damaged or destroyed. The current rovers use a combination of a warm lump of plutonium (even with the requisite shielding, the energy density is ridiculously high) and electrical heating, along with aerogel insulation. Big things are easier to keep warm (cube/square law on the insulation), so you would spend all your power budget keeping the mini-rover alive. Never mind the issue of making sure it doesn't fall off.
I imagine the glue solution would fall victim to the same analysis as every other such plan. It would *probably* work, at the cost of a few ounces, like you said. A few ounces means it replaces an instrument.
A lot of people seem to assume that guaranteeing that the rovers don't die from loss of power should have been a goal. The actual goals were to maximize the value of the information returned, and sacrificing instruments interferes with that goal. Also note that the rovers are slowly failing in other ways -- ie their drive trains. In a sense, the optimal design has all the parts failing at once, as opposed to trying to keep a few parts alive as long as possible.
The issue was complicated. First, they weren't sure what the effect of dust storms would be. They got lucky; the storms seem to clean the panels more than they add dust. Second, they evaluated a number of different options for panel cleaning -- wipers, peel-away plastic covers, electrostatic devices, etc. The conclusion was there were a number of options, any of which would probably work for a while. However, it was decided that any of the options would take the place in space / weight of approximately one instrument. They decided that they'd rather have better info for a shorter time, especially given that the dust storms might turn out to clean the panels and let the rovers keep going without any cleaning system at all. So, they opted for instrumentation over longevity, and lucked out and got both.
That's why you don't *do* that. Or, more precisely, why the SSD shouldn't *let* you do that. All it needs to do is keep some hidden spare space (10%? 5%? 1%? I don't know, but it's not huge) and dynamically remap sectors to balance writes. If you have GB of remapping room, even a "full" disk with heavy load would take a long time to wear out.
While that may be true in general, it is certainly not as black and white as you make it seem. I've sometimes been on prescription levels of amphetamine that are higher than what I've seen people take recreationally -- and they said they didn't *want* to take more, because the effects are too strong. Opiates are similar -- the high end of the normal medical dosage is far higher than the low end of the recreational dosage.
Part of the reason, at least in the case of ADD and stimulants, appears to be that ADD people and non-ADD people react differently to stimulants.
It's also worth noting that many users of illicit stimulants aren't using them recreationally -- they're using them as productivity boosters. They're not trying to get high, they're trying to stay awake or stay focused so they can study or do their job better. Use of Adderall and similar amphetamines is common on college campuses, largely for things like working on papers. Plenty of truckers use stimulants to stay awake. And *lots* of people use caffeine for productivity reasons. Before you say that doesn't count, let me point out that caffeine produces symptoms of dependence. Interestingly, I've been addicted to caffeine in the sense of having withdrawl symptoms, but never amphetamines. (I'm not trying to say it's more addictive -- I'm fairly sure it's not -- just that different people respond differently, and things like dose and usage pattern matter.)
Slashdot Mirror
There are several things to realize about this prize. First, the rover is very roughly a third of the work. I'd break into getting to LEO, getting to the lunar surface, and all the stuff on the surface (rover, video, communication, etc.).
If you're trying to do this on a budget comparable to the prize, each of those is very challenging. If you buy your orbital launch, the cheapest option is probably a SpaceX Falcon 1, which starts at $7M -- a third of your budget already. That means you get *one* attempt. This prize won't be won on the first flight of the hardware, not with a budget even approaching the $20M purse.
Getting from (Earth) orbit to the surface is tricky, but probably the easiest piece. Carmack is very close to demonstrating a large fraction of that with Pixel at the Lunar Lander Challenge in October. Left to do would be nontrivial navigation and a nontrivial performance boost. Here, buying the hardware you need certainly isn't off the shelf, but most of the pieces might be available. I suspect you'd find yourself blowing another large fraction of your budget even before the requisite development on this part.
The lunar rover and communications presents another set of challenges, which it sounds like CMU may well have experience with.
But, I'd say hiring NASA engineers is the wrong way to win this on a budget. NASA couldn't even begin to touch this prize for $100M. If you hire engineers who are used to working with budgets on a NASA size, you'll get a solution that costs NASA price tags, or close to them. If you want to spend a couple hundred million winning the prize, just to prove you can, it'll work -- but I would say that's kind of silly. I don't think this prize will be won for less than $20M, but I think it will be won for not a huge amount more.
Personally, I think Carmack and the rest of the people at Armadillo Aerospace are much more interesting to watch. If he continues at his current pace, he'll have hardware in LEO long before this prize expires, and on a much smaller budget than anyone has done before. And he's already been talking about what would be needed to win this prize. If you want to watch the interesting show, don't look to the people that say they'll do it the old way -- look to the people that want to do it orders of magnitude cheaper than it's been done before, by turning every piece of conventional wisdom on its head, and are busy proving they can rather than trumpeting their barely formed plans to the press.
Oddly enough, that's not something I've had an issue with.
All the data I've needed to analyze usually comes in the form of nice simple CSV text files, or comes from a program that can optionally output as such. So the import is almost irrelevant.
But then actually doing something with that data is where Excel wins out (I won't say it's good -- it's sadly lacking in some respects -- but it beats OOo hands down). I most emphatically don't want online LAMP type applications; nothing I've done has been of the "plug numbers in here, read answers there" form. It's all engineering problems, which usually end up as one-off spreadsheets intended for a specific purpose. I've done everything from electronic circuit design to data analysis to finite element analysis to thermodynamic modeling in Excel, without any particular difficulty. Some of those could have been done in OOo, though they would have been harder. Some couldn't have (at least, not reasonably). And the major differences are in the goal seek / solver tools, functions, circular references, and performance.
I'm sure there are applications where OOo is comparable or better -- but none of them are the applications I work with, at least not when it comes to the "advanced" or complicated things. (There's a decent amount of quick and dirty scratch work that I do in a spreadsheet, and for that I don't really care which I'm working in, they're both fine.)
Can anyone out there make a convincing case that Calc or Gnumeric are just as good as Excel, even for advanced users?
(I haven't worked with Gnumeric much, so this mostly applies to Calc)
I would, except that they're not even close. I run Linux, I use OpenOffice.org, and at work I've done some serious Excel work. There are a couple features I like in Calc, but nothing major. However, there are a lot of things that Excel does better. Graphs are a big one. Circular reference handling and goal seek / solver are also important. And last but certainly not least... handling of giant spreadsheets. That even Excel could be better at, but it's less painful than Calc. Complicated math like statistics and Fourier work is also lacking, and VB scripting is important for function definitions (though I'd love to use something better).
I've tried to do complicated stuff in Calc, but never for too long -- it just gets painful.
$30M is at least enough to attract interest. John Carmack has already started talking about what would be required to do it. And given that he's the odds-on favorite to win the Lunar Lander Challenge (both levels) this year at the X-Prize Cup, I'd say that's rather interesting. It sounds like he thinks the toughest problem to solve is navigation.
And you can get to LEO for $7M to $8.5M on a Falcon 1.
On the other hand, a lot of what has been happening lately probably *is* unhelpful. For example, banning anything pointy (nail clippers?) from airplanes -- what prevents another 9/11 is a change in attitude among passengers, not a complete lack of sharp things.
That said, there are plenty of things I oppose because I find the tradeoff abhorrent -- like the eavesdropping and other privacy invasions, and the imprisonment without due process, and...
That excuses the fluff (somewhat, but I've never really minded the "random geek does something irrelevant but neat" type of fluff), but it most assuredly does not excuse the rampant FUD and trolling.
I want a $39.95 plan to actually cost $39.95. As in, that's the number at the bottom of the bill that I have to pay each month.
I don't want to pay "regulatory surcharges" or "cost recovery fees" or anything else that isn't included in the advertised price. And this goes for all these sorts of contracts, not just cell phones.
It's a collective agreement to slowly disarm themselves
Exactly. The only way to win at Prisoner's Dilemma is to change the rules.
It's a radio telescope; the atmosphere is almost irrelevant. This gives a very large effective size for diffraction purposes, meaning the resulting images can be much more finely resolved.
Except that the guts of a supercapacitor are relatively inert chemically, and are non-flammable. That makes the problem much easier to handle.
The issue of whether or not to put some sort of dust-clearing device on the panels was examined critically and decided on early in this project. In short: they didn't know what dust storms would do to the panels; it turns out they tend to remove dust. Several options for dust clearing were considered -- wipers, electrostatic techniques, peel-away plastic, and probably others I've forgotten. All of them would have *probably* worked, and all of them would have taken up space and weight. Essentially it came down to choosing between dust removal and an instrument. Faced with that decision, they decided that better quality, more complete data was more interesting than having the rovers run longer.
Of course, they got lucky, and the dust storms seem to clear dust off the panels. So there was even less need for dust-clearing than they thought there might be.
The obvious ones: higher energy density (meaning smaller and lighter, or longer life between charges), higher efficiency (and therefore less heat generation on charge / discharge), no flammable components, higher charge / discharge rates (and therefore faster charge / higher power density). All but the first (energy density) are available now, and they only lag batteries by about a factor of 10 in energy density. But if the stuff people have in labs turns into real products, or the rumors in the article I cited are correct, we'll see that last part change in the relatively near future (1-5 years, depending who you listen to). Also, don't underestimate the value of instant charging -- wouldn't it be neat to drop your cell phone on the charger for 30 seconds whenever it's convenient and have it be fully charged? Or your laptop charge in 5 minutes instead of 2 hours? That convenience factor is at least interesting, if not earth-shattering.
I mean exactly what you think I mean -- used in some applications, but not the most performance-critical ones. In the same way that NiMH batteries are available and in use now, I expect some form of LiIon to be available and in use even once supercaps are widespread.
How are they solving a different problem? Obviously right now they're not suited to replace LiIon batteries, and they're getting used to solve other problems, but if the energy density catches up to LiIons, then won't ultracaps replace them as soon as they're cost effective?
Fortunately, we have supercapacitors. While they're not there yet for energy density (still about 10x too little) they're rapidly improving. 10x isn't much at the rate these things have been improving, and there are plenty of labs with pieces that are much better than currently available commercial offerings, but that still need development work. If I had to guess, I'd say it's 5 years until the first supercaps appear in serious commercial use, and less than 10 until LiIon has gone the way of NiMH.
Of course, if you believe the rumors then it might be even faster than that -- we might be seeing serious applications in a year or so.
I, for one, will be glad to give LiIon a proper burial. But until then, we work with what's available.
Why is the entire front page populated with stories by kdawson? Did the rest of the "editors" quit or something? It'd be nice to have more of a mix of stories on occasion.
And BTW, no on ever believed that Saddam plotted 9/11
You say that, but polls have shown that over 50% of Americans think he was connected to Al Qaeda. I don't know what fraction think he was involved in 9/11, but I bet it's far higher than 0%.
It strikes me that the problem is that most spell checkers try to check everything, and that a lot of code has things that really shouldn't be spell checked at all, mixed with things that should. I imagine that one way to start would be to only alert on those errors that are almost correct -- if it looks like garbage, ignore it, but if it's close, assume it should be right. Perhaps ignore prefixes / suffixes as well -- pSomething is fine, pSometihng isn't. Also, CamelCase ought to be easy enough to detect -- treat it as word boundaries, and spell check the individual words. Again, egregious misspellings probably aren't -- nextObjFoo is ok, even though Obj isn't a word -- it's so far from being a word that we assume the programmer meant it that way.
Similarly, there should probably be a set of words added that aren't "English" but are used often enough to be worth adding to the dictionary. Things like Obj, Int, and Ptr.
I think the reason such spell checkers don't exist already is fairly simple -- everyone just assumes they're impossible, and doesn't try. Couple that with the fact that a mediocre quality one would be so annoying as to be worse than useless, and you have a recipe for a program that won't get written. I don't think either of those would have to be the case if someone sufficiently clever decided to attack the problem, though.
You can't build a rover that tiny and have it last as long as Spirit and Opportunity have. The reason is the same one this latest round of dust storms was worrying -- heat. Batteries need to stay warm; if they freeze, they are severely damaged or destroyed. The current rovers use a combination of a warm lump of plutonium (even with the requisite shielding, the energy density is ridiculously high) and electrical heating, along with aerogel insulation. Big things are easier to keep warm (cube/square law on the insulation), so you would spend all your power budget keeping the mini-rover alive. Never mind the issue of making sure it doesn't fall off.
I imagine the glue solution would fall victim to the same analysis as every other such plan. It would *probably* work, at the cost of a few ounces, like you said. A few ounces means it replaces an instrument.
A lot of people seem to assume that guaranteeing that the rovers don't die from loss of power should have been a goal. The actual goals were to maximize the value of the information returned, and sacrificing instruments interferes with that goal. Also note that the rovers are slowly failing in other ways -- ie their drive trains. In a sense, the optimal design has all the parts failing at once, as opposed to trying to keep a few parts alive as long as possible.
The issue was complicated. First, they weren't sure what the effect of dust storms would be. They got lucky; the storms seem to clean the panels more than they add dust. Second, they evaluated a number of different options for panel cleaning -- wipers, peel-away plastic covers, electrostatic devices, etc. The conclusion was there were a number of options, any of which would probably work for a while. However, it was decided that any of the options would take the place in space / weight of approximately one instrument. They decided that they'd rather have better info for a shorter time, especially given that the dust storms might turn out to clean the panels and let the rovers keep going without any cleaning system at all. So, they opted for instrumentation over longevity, and lucked out and got both.
Someone just won at Internet trolling.
Hear that, Slashdot trolls? When you get written up by Reuters, that's when you've arrived.
You would think it would be embarrassing to have tags like that on the front page. But, somehow, the slashdot editors don't seem to care.
That's why you don't *do* that. Or, more precisely, why the SSD shouldn't *let* you do that. All it needs to do is keep some hidden spare space (10%? 5%? 1%? I don't know, but it's not huge) and dynamically remap sectors to balance writes. If you have GB of remapping room, even a "full" disk with heavy load would take a long time to wear out.
While that may be true in general, it is certainly not as black and white as you make it seem. I've sometimes been on prescription levels of amphetamine that are higher than what I've seen people take recreationally -- and they said they didn't *want* to take more, because the effects are too strong. Opiates are similar -- the high end of the normal medical dosage is far higher than the low end of the recreational dosage.
Part of the reason, at least in the case of ADD and stimulants, appears to be that ADD people and non-ADD people react differently to stimulants.
It's also worth noting that many users of illicit stimulants aren't using them recreationally -- they're using them as productivity boosters. They're not trying to get high, they're trying to stay awake or stay focused so they can study or do their job better. Use of Adderall and similar amphetamines is common on college campuses, largely for things like working on papers. Plenty of truckers use stimulants to stay awake. And *lots* of people use caffeine for productivity reasons. Before you say that doesn't count, let me point out that caffeine produces symptoms of dependence. Interestingly, I've been addicted to caffeine in the sense of having withdrawl symptoms, but never amphetamines. (I'm not trying to say it's more addictive -- I'm fairly sure it's not -- just that different people respond differently, and things like dose and usage pattern matter.)