It's a bit like what the late canoeing guru Bill Mason said about river portages. Carrying your canoe and gear for a half mile upstream to get around a rapids may not be fun, but it means you'll have stretches of river where there are no power boats. Now you don't have to hate powerboats or want them all to go away entirely to be able to enjoy being away from them.
If you're a hiker, you don't necessarily object to roads being built to some mountaintops, so people who don't or can't hike can enjoy them. But you wouldn't want them to be everywhere. If you hike the Appalachian Trail, you may enjoy an occasional night in a motel, but you don't want the entire trail lined with them. If you're a bikepacker, an occasional bike shop on your route is convenient, but having to prepare for crossing remote areas unsupported is part of the experience you're after.
The sacrifices a place demands from you can be part of its appeal, not the least because it keeps out people who aren't willing to make those sacrifices. But the absence of services creates experiences that are increasingly difficult to obtain in connected, consumer society: self-sufficiency. Isolation. Quiet.
Cell coverage is a good thing, but the existence of a few final places where even checking your phone would be pointless is also a good thing. It's maybe not the highest thing on the conservation priority list, because you can always shut your phone off and keep it in the bottom of your bag. But you'll still know it's there, and that makes a difference. Maybe not a big difference, but some.
There's an even simpler reason not to take the probiotic pills you see in the drug store or even the supermarket these days.
You wouldn't take a pill from an open bottle of "nutritional supplement" you found lying in the gutter because you wouldn't be sure what's actually in it. The thing is your chances aren't much better if you purchased the unopened bottle from a pharmacy. Thanks to a legal system which equates buying congressmen with speech, supplements are effectively completely unregulated.
When a state AG's office had samples of supplements being sold in the state tested, in nearly 80% of the cases the thing being sold couldn't be detected in the bottle, although plenty of contaminants were found.
NEJM reported a case of a police sergeant who lost his job due to a failed drug test: supplements he was taking were contaminated with amphetamines. If nobody's checking, why bother cleaning the pill machinery for that supplement batch?
Another study looked at supplements from 12 companies; only two companies sold what was in the bottle. Seven adulterated what was sold in the bottle with cheaper subtitutes; two used substitutes entirely. Four companies' products contained 20% or more contaminants by weight, and one company's products were 100% contaminants:basically random shit they swept off the floor. Many contaminants found were either toxic or could not be positively identified as benign.
Nobody should take any "nutritional supplements" for anything. People should use food instead. If you eat a banana, it might not be the best banana it could be, but at least you can be certain it is a banana. And it will feed your gut microbiome.
There really aren't a lot of obstacles to deal with. Just point the boat and go.
I imagine that's true, if you have an effectively unlimited energy, but the terms of this contest is to build a vehicle less than eight feet long that can cross the Atlantic.
That's not a lot of room to stuff with batteries or diesel fuel, or to cover with solar panels. And that's the whole point. If you could make the vehicle a hundred feet long, this challenge would be expensive, but easy. By making the boat tiny, you make the challenge affordable, but tough.
The race has two divisions, sail powered and unlimited, but it's hard to see what they had in mind for propulsion by having an unlimited division. Possibly some seawater-replenished fuel cell.
You are 22 years behind the times when it comes to Federal law. The Communications Decency Act prevents the government from treating service providers who attempts to police user-provided content as a publishers of said content. The idea was to remove a disincentive for service providers from making good faith efforts to extirpate naughty material, but it also protects providers who make a good faith effort to extirpate what they believe to be bullshit.
So, in the names of "hand-off, anything-goes" you want the government to regulate private sector use of the Internet when a site owner has an editorial position you don't like.
I'm not sure that cars were ever really the focus. I don't think Musk wants to be Henry Ford. He wants to be John D. Rockefeller.
Anyhow, no giants have been sleeping. The time hasn't been right for them, and now they're probably moving a little faster because the right time would be too late.
Your employer isn't doing anything special for you by paying you. It's not a perq.
But then a healthy workspace isn't a perq either. Have we really got that twisted in our thinking? Do we really believe an employer is doing the employee some kind of *favor* by paying him?
Frankly, that attitude is just plain dumb... on the part of the employers. Sure you're not in business to make your employees like you, but your not in business making your employees miserable either. When I ran a development team I used to make people go home when I thought they'd been working too long, because their sacrifice meant nothing to me.
But you'd still have URLs under the covers. Or you'd end up re-inventing them.
Some years ago I spent time pondering all the possible ways things could be identified in a system. I came up with three broad, non-mutually exclusive strategies:
(1) Analytically: identify a thing by a set of properties which are unique to it (e.g. relational primary keys); (2) Algorithmically: use an algorithm that is guaranteed to issue identifiers that are unique in the required scope (e.g. UUIDs for global scope, or within restricted, local contexts simple serial numbers); and, (3) By authority: put somebody in charge of naming things; they can use any method they like (e.g. scientific journal editors control species taxonomies).
All of these approaches have drawbacks. For example relational primary keys are not robust when data is shared across applications, because different applications obey different rules. Algorithmically assigned identifiers are either not globally unique (serial numbers) or are long and arbitrary therefore hard to key in accurately (UUIDs).
URLs are identifiers by authority. Somebody says that "https://tech.slashdot.org/story/18/09/04/1722244/google-wants-to-kill-the-url" identifies this discussion, and so it does. But you can see two very useful features of URLs in operation here. First, they allow whoever runs the tech.slashdot.org to delegate authority for stories to whomever is in charge of the "../stories" URLs. Second, it allows whomever is in charge of stories to transform his local unique identifier ("18/09/04/1722244/google-wants-to-kill-the-url") into a globally unique address.
If you include these two features, delegation of authority and transformation of locally unique IDs to globally unique locators, you pretty much end creating something like a URL. Since the idea of mixing protocols like http and ftp isn't really a big thing anymore, you could ditch the protocol identifier and make it a URI, e.g. ("web:tech.slashdot.org/story/18/09/04/1722244/google-wants-to-kill-the-url") and the browser would automatically request the resource using https, which is what everyone should be using now.
Another possibility is to use universal resource names (e.g., "urn:isbn:978-0471117094") with some kind of global directory. But that directory would have to return to the browser something that is an awful lot like a URL, even though the user doesn't see it.
The US is moving away from coal because of natural gas; if Poland would do the same, it'd have to turn to Russian, which they do not want to do.
One of the reasons the Obama administration was so fracking friendly was the goal of becoming a gas exporter, thus limiting the influence of Russian in Europe.
As for Billings, climate change projections of the mountain states like Montana is less winter snowpack, which translates to more water in the winter but less in the summer. This will have a patchwork of effects, many of which will be most pronounced states downstream in the Missouri River basin.
Well, this is different from normal distrust. If you think allies aren't spying on you, you're naive. But that's different from not *trusting* your allies.
Oh, of course that makes a difference. The world has two kinds of people, good people whose opinions you should trust without question, and bad people, whose opinions you should reject without consideration.
Asking whether the coral is more "ecologically valuable" than the starfish is premature, because we don't really understand the ecological relationship between the starfish and the reef. It is possible that the reef as we know it might not even be possible without starfish predation.
The crown-of-thorns starfish not some exotic species, it evolved with the corals it preys upon. Had humans never evolved, it would still be killing off sections of coral reef, only there wouldn't be anyone around to be upset at the pictures.
Now there are two possible justifications for humans intervening in the starfish/reef interaction. The first is if prior human actions artificially inflate the predator starfish population. The second is if prior human actions artificially weaken the prey coral population. In both those cases the sensible and indeed only feasible solution is to stop doing whatever it is we've been doing that creates the problem.
You can't eradicate an animal which lays twenty million viable eggs at a time by killing individuals, even on a massive scale. The very next year an area you to all visible appearances swept clean will be covered again. What you have to do is tweak the factors that alter the survival rate of those twenty million eggs.
I worked for many years in the control of vector-borne diseases. There has never been a successful case of mosquito eradication from a region. Never.
In fact pursing the goal of eradication actually makes it harder to limit human exposure harder in the long term. The reason is that to eradicate a mosquito population, you'd have to saturate the entire region with lethal doses of pesticide, which is physically and economically impossible.
Attempting total eradication only creates evolutionary pressure on the population to develop pesticide resistance and greater fecundity.
You know what really sets mature adults apart from wet-behind-the-ears kids? They accept that they're fallible.
Language safety got a bad reputation back in the 70s when I learned to program because language designers didn't know how to do it without getting in the way of ordinary work. So back then given a choice of K&R C or the Pascal dialects available, I'd have chosen K&R C because Pascal's strict type checking made common use-cases awkward. But there's absolutely no question that modern ANSI C's stricter type-safety makes it far better than K&R C.
You can't rely on a language to do your job for you, but you can certainly rely on a language to make your job easier.
The Russian project uses two naval nuclear reactors, the same ones used in their nuclear ice breakers.
Engineering decisions often result in some things getting harder while other things get easier. Putting a nuclear reactor on a ship certainly simplifies the problem of obtaining cooling water, but you run into the problem of space. To keep naval reactors physically compact, they run on highly enriched uranium. At least American and Russian ones do. France uses low enriched uranium in its submarines, but their reactors as installed are much less powerful than American or Russian.
The KLT-40 reactor used in the Russian project is designed to run on 40% to 90% enriched uranium. At the upper end of that scale the fuel could easily be used in a crude gun type weapon as was dropped on Hiroshima. The fuel at the lower end could be used in a more sophisticated design, or as part of a multi-stage weapon, or used as a shortcut to obtaining higher enrichment levels for a less sophisticated design.
And that fuel is already on a conveniently mobile platform.
It's not that the floating reactor idea is inherently impractical, it's just that the Russian project doesn't really demonstrate that idea is feasible for widespread deployment, because you wouldn't want hundreds of these things all over the place. And it doesn't really tell you anything about the economics of potential civilian designs, because you wouldn't use HEU in those and HEU simplifies everything else.
Well, I think as long as we are dealing with prototypes, there would on the engineers' part be such a rebutable presumption of guilt, because that's how you do engineering.
From a legal perspective it makes no sense not to go with the more usual preponderance of evidence standard, especially as there's bound to be a lot more data. But if for some reason complete log and telemetry data aren't available for the robot driver, in *that* case there might reasonably be a presumption of guilt.
Well, why would you preferentially replace the *good* drivers first?
I don't think it follows form the 98% human fault rate that robotic drivers don't try to prevent accidents. Who would want to ride in a car which didn't drive defensively? But I suspect robots aren't quite as good as human with dealing with other humans behavioral flexibility, which is a nice way of saying "unpredictability".
That flexibility is sometimes good, sometimes bad. In a world of robotic drivers, no car would stop to honk at another car for entering the the highway the wrong way; but then that other car wouldn't be doing that.
In the interim things will be interesting. The very best drivers on their good days can do feats of inference about other drivers that I doubt the software is up to yet. So in the near future some human-caused accidents will be avoidable by humans because we know that late model BMW passing us at high speed on the right is going to cut us off.
Slashdot Mirror
It's a bit like what the late canoeing guru Bill Mason said about river portages. Carrying your canoe and gear for a half mile upstream to get around a rapids may not be fun, but it means you'll have stretches of river where there are no power boats. Now you don't have to hate powerboats or want them all to go away entirely to be able to enjoy being away from them.
If you're a hiker, you don't necessarily object to roads being built to some mountaintops, so people who don't or can't hike can enjoy them. But you wouldn't want them to be everywhere. If you hike the Appalachian Trail, you may enjoy an occasional night in a motel, but you don't want the entire trail lined with them. If you're a bikepacker, an occasional bike shop on your route is convenient, but having to prepare for crossing remote areas unsupported is part of the experience you're after.
The sacrifices a place demands from you can be part of its appeal, not the least because it keeps out people who aren't willing to make those sacrifices. But the absence of services creates experiences that are increasingly difficult to obtain in connected, consumer society: self-sufficiency. Isolation. Quiet.
Cell coverage is a good thing, but the existence of a few final places where even checking your phone would be pointless is also a good thing. It's maybe not the highest thing on the conservation priority list, because you can always shut your phone off and keep it in the bottom of your bag. But you'll still know it's there, and that makes a difference. Maybe not a big difference, but some.
There's an even simpler reason not to take the probiotic pills you see in the drug store or even the supermarket these days.
You wouldn't take a pill from an open bottle of "nutritional supplement" you found lying in the gutter because you wouldn't be sure what's actually in it. The thing is your chances aren't much better if you purchased the unopened bottle from a pharmacy. Thanks to a legal system which equates buying congressmen with speech, supplements are effectively completely unregulated.
When a state AG's office had samples of supplements being sold in the state tested, in nearly 80% of the cases the thing being sold couldn't be detected in the bottle, although plenty of contaminants were found.
NEJM reported a case of a police sergeant who lost his job due to a failed drug test: supplements he was taking were contaminated with amphetamines. If nobody's checking, why bother cleaning the pill machinery for that supplement batch?
Another study looked at supplements from 12 companies; only two companies sold what was in the bottle. Seven adulterated what was sold in the bottle with cheaper subtitutes; two used substitutes entirely. Four companies' products contained 20% or more contaminants by weight, and one company's products were 100% contaminants:basically random shit they swept off the floor. Many contaminants found were either toxic or could not be positively identified as benign.
Nobody should take any "nutritional supplements" for anything. People should use food instead. If you eat a banana, it might not be the best banana it could be, but at least you can be certain it is a banana. And it will feed your gut microbiome.
The study could in no way establish that microflora from food is bad, because it didn't test that.
There really aren't a lot of obstacles to deal with. Just point the boat and go.
I imagine that's true, if you have an effectively unlimited energy, but the terms of this contest is to build a vehicle less than eight feet long that can cross the Atlantic.
That's not a lot of room to stuff with batteries or diesel fuel, or to cover with solar panels. And that's the whole point. If you could make the vehicle a hundred feet long, this challenge would be expensive, but easy. By making the boat tiny, you make the challenge affordable, but tough.
The race has two divisions, sail powered and unlimited, but it's hard to see what they had in mind for propulsion by having an unlimited division. Possibly some seawater-replenished fuel cell.
"De facto common carrier status"?
You are 22 years behind the times when it comes to Federal law. The Communications Decency Act prevents the government from treating service providers who attempts to police user-provided content as a publishers of said content. The idea was to remove a disincentive for service providers from making good faith efforts to extirpate naughty material, but it also protects providers who make a good faith effort to extirpate what they believe to be bullshit.
It really doesn't matter how you feel about legal actions they take to maximize their profits.
So, in the names of "hand-off, anything-goes" you want the government to regulate private sector use of the Internet when a site owner has an editorial position you don't like.
I'm not sure that cars were ever really the focus. I don't think Musk wants to be Henry Ford. He wants to be John D. Rockefeller.
Anyhow, no giants have been sleeping. The time hasn't been right for them, and now they're probably moving a little faster because the right time would be too late.
You don't know what a perq is.
Your employer isn't doing anything special for you by paying you. It's not a perq.
But then a healthy workspace isn't a perq either. Have we really got that twisted in our thinking? Do we really believe an employer is doing the employee some kind of *favor* by paying him?
Frankly, that attitude is just plain dumb ... on the part of the employers. Sure you're not in business to make your employees like you, but your not in business making your employees miserable either. When I ran a development team I used to make people go home when I thought they'd been working too long, because their sacrifice meant nothing to me.
But you'd still have URLs under the covers. Or you'd end up re-inventing them.
Some years ago I spent time pondering all the possible ways things could be identified in a system. I came up with three broad, non-mutually exclusive strategies:
(1) Analytically: identify a thing by a set of properties which are unique to it (e.g. relational primary keys);
(2) Algorithmically: use an algorithm that is guaranteed to issue identifiers that are unique in the required scope (e.g. UUIDs for global scope, or within restricted, local contexts simple serial numbers);
and,
(3) By authority: put somebody in charge of naming things; they can use any method they like (e.g. scientific journal editors control species taxonomies).
All of these approaches have drawbacks. For example relational primary keys are not robust when data is shared across applications, because different applications obey different rules. Algorithmically assigned identifiers are either not globally unique (serial numbers) or are long and arbitrary therefore hard to key in accurately (UUIDs).
URLs are identifiers by authority. Somebody says that "https://tech.slashdot.org/story/18/09/04/1722244/google-wants-to-kill-the-url" identifies this discussion, and so it does. But you can see two very useful features of URLs in operation here. First, they allow whoever runs the tech.slashdot.org to delegate authority for stories to whomever is in charge of the "../stories" URLs. Second, it allows whomever is in charge of stories to transform his local unique identifier ("18/09/04/1722244/google-wants-to-kill-the-url") into a globally unique address.
If you include these two features, delegation of authority and transformation of locally unique IDs to globally unique locators, you pretty much end creating something like a URL. Since the idea of mixing protocols like http and ftp isn't really a big thing anymore, you could ditch the protocol identifier and make it a URI, e.g. ("web:tech.slashdot.org/story/18/09/04/1722244/google-wants-to-kill-the-url") and the browser would automatically request the resource using https, which is what everyone should be using now.
Another possibility is to use universal resource names (e.g., "urn:isbn:978-0471117094") with some kind of global directory. But that directory would have to return to the browser something that is an awful lot like a URL, even though the user doesn't see it.
You say that like it's a bad thing.
The US is moving away from coal because of natural gas; if Poland would do the same, it'd have to turn to Russian, which they do not want to do.
One of the reasons the Obama administration was so fracking friendly was the goal of becoming a gas exporter, thus limiting the influence of Russian in Europe.
As for Billings, climate change projections of the mountain states like Montana is less winter snowpack, which translates to more water in the winter but less in the summer. This will have a patchwork of effects, many of which will be most pronounced states downstream in the Missouri River basin.
Well, this is different from normal distrust. If you think allies aren't spying on you, you're naive. But that's different from not *trusting* your allies.
It's not the yanks they're worried about. It's the Russia, and Russia's proxies.
We may no longer have a choice about getting hosed, but we still have a choice about how *quickly* we get hosed.
Would you rather gethosed by a garden hose on mist or a firehose?
It's the ability of VCs to monetize open source software.
Not a problem for me, or for established businesses either.
Oh, of course that makes a difference. The world has two kinds of people, good people whose opinions you should trust without question, and bad people, whose opinions you should reject without consideration.
I take it Pao is one of the bad people.
A population *crash* isn't tantamount to eradication.
Asking whether the coral is more "ecologically valuable" than the starfish is premature, because we don't really understand the ecological relationship between the starfish and the reef. It is possible that the reef as we know it might not even be possible without starfish predation.
The crown-of-thorns starfish not some exotic species, it evolved with the corals it preys upon. Had humans never evolved, it would still be killing off sections of coral reef, only there wouldn't be anyone around to be upset at the pictures.
Now there are two possible justifications for humans intervening in the starfish/reef interaction. The first is if prior human actions artificially inflate the predator starfish population. The second is if prior human actions artificially weaken the prey coral population. In both those cases the sensible and indeed only feasible solution is to stop doing whatever it is we've been doing that creates the problem.
You can't eradicate an animal which lays twenty million viable eggs at a time by killing individuals, even on a massive scale. The very next year an area you to all visible appearances swept clean will be covered again. What you have to do is tweak the factors that alter the survival rate of those twenty million eggs.
I worked for many years in the control of vector-borne diseases. There has never been a successful case of mosquito eradication from a region. Never.
In fact pursing the goal of eradication actually makes it harder to limit human exposure harder in the long term. The reason is that to eradicate a mosquito population, you'd have to saturate the entire region with lethal doses of pesticide, which is physically and economically impossible.
Attempting total eradication only creates evolutionary pressure on the population to develop pesticide resistance and greater fecundity.
You know what really sets mature adults apart from wet-behind-the-ears kids? They accept that they're fallible.
Language safety got a bad reputation back in the 70s when I learned to program because language designers didn't know how to do it without getting in the way of ordinary work. So back then given a choice of K&R C or the Pascal dialects available, I'd have chosen K&R C because Pascal's strict type checking made common use-cases awkward. But there's absolutely no question that modern ANSI C's stricter type-safety makes it far better than K&R C.
You can't rely on a language to do your job for you, but you can certainly rely on a language to make your job easier.
The Russian project uses two naval nuclear reactors, the same ones used in their nuclear ice breakers.
Engineering decisions often result in some things getting harder while other things get easier. Putting a nuclear reactor on a ship certainly simplifies the problem of obtaining cooling water, but you run into the problem of space. To keep naval reactors physically compact, they run on highly enriched uranium. At least American and Russian ones do. France uses low enriched uranium in its submarines, but their reactors as installed are much less powerful than American or Russian.
The KLT-40 reactor used in the Russian project is designed to run on 40% to 90% enriched uranium. At the upper end of that scale the fuel could easily be used in a crude gun type weapon as was dropped on Hiroshima. The fuel at the lower end could be used in a more sophisticated design, or as part of a multi-stage weapon, or used as a shortcut to obtaining higher enrichment levels for a less sophisticated design.
And that fuel is already on a conveniently mobile platform.
It's not that the floating reactor idea is inherently impractical, it's just that the Russian project doesn't really demonstrate that idea is feasible for widespread deployment, because you wouldn't want hundreds of these things all over the place. And it doesn't really tell you anything about the economics of potential civilian designs, because you wouldn't use HEU in those and HEU simplifies everything else.
Well, I think as long as we are dealing with prototypes, there would on the engineers' part be such a rebutable presumption of guilt, because that's how you do engineering.
From a legal perspective it makes no sense not to go with the more usual preponderance of evidence standard, especially as there's bound to be a lot more data. But if for some reason complete log and telemetry data aren't available for the robot driver, in *that* case there might reasonably be a presumption of guilt.
Well, why would you preferentially replace the *good* drivers first?
I don't think it follows form the 98% human fault rate that robotic drivers don't try to prevent accidents. Who would want to ride in a car which didn't drive defensively? But I suspect robots aren't quite as good as human with dealing with other humans behavioral flexibility, which is a nice way of saying "unpredictability".
That flexibility is sometimes good, sometimes bad. In a world of robotic drivers, no car would stop to honk at another car for entering the the highway the wrong way; but then that other car wouldn't be doing that.
In the interim things will be interesting. The very best drivers on their good days can do feats of inference about other drivers that I doubt the software is up to yet. So in the near future some human-caused accidents will be avoidable by humans because we know that late model BMW passing us at high speed on the right is going to cut us off.