"If a significant majority of the solutions show one phase instead of the other, it suggests that there is some predictability of the feature (in this case, the NAO) despite the uncertainty and errors from the initial conditions."
There is, however, some chance that the models share some common incorrect assumption. They don't agree because they are correct, but because they are similar to each other. The Literary Digest effect -- sort of.
As William Shakespeare might have said "Computers can predict weather a year ahead. So can I. So can any other man. But will the weather come when I summon it?" See Henry IV Part 1: Act 3, Scene 1, Page 3 nfs.sparknotes.com/henry4pt1/page_133.html for details.
Did Microsoft know these things might be used in a football stadium? Did they provide any guidance to the NFl on what sort of connectivity the stadiums must provide? Sounds awfully like yet another instance of the entirely too common practice blaming the user for being unable to use a crappy product.
"For example a file can be 'owned' by any number of groups in windows. But in POSIX it belongs to one group. In NTFS I can do things like user 1 can change the permissions but not read the file but user 2 can not even look at the file (will not see it in the listing) but can modify the permissions. Totally contrived but possible to do. I am not totally sure you could actually pull that off in a posix sort of system."
Well, yes, you probably could do that. But I've always wondered if the potential for the small subset of us who occasionally make mistakes creating truly baffling problems for the rest of you to solve, was worth the benefit.
Gee. You don't think that it could be possible that doing computer security even adequately is beyond what people are capable of actually doing? Golly, that might mean that e-commerce is doomed and that all computers are really good for is research and entertainment.
That might put a kink in some folks plans to promote the cloud into a vehicle that will enrich them beyond all belief.
"To my knowledge Google cars have only been driven in near perfect conditions. Ie no blizzards or blowing snow or ice of any kind."
My impression also. I don't have any problem with that. It's an intelligent way to run an R&D program. The only problem is that Google is probably only 15% or so of the way to having a viable autonomous vehicle that you can drop anywhere where there is an accurate digital streetmap in any sort of weather and not have it be a menace to living creatures and inanimate objects. That's OK also as long as management and the public understand that there is a VERY long way to go.
"also the technology on Google cars is way more unaffordable than a Tesla."
Price comes down when your production run is 20 million units?
"So at some point in the future we will see that computers achieve a higher safety level than any sample of human drivers, while remaining imperfect. At this point, it will probably become necessary to ban manual driving on highways, for the protection of the other people on the highway."
Simplistic. For one thing, automatic things like cars is surely going to turn out to be MUCH harder than most anyone expects. Ask folks who have experience applying computers to real world physical processes. The number and variety of things that can and will go wrong is astonishing. Especially once one ventures of the expressway and away from pre-plotted test circuits to a world of kids, pets, horsecarts, wildlife, poorly marked intersections, bicyclists, construction projects, etc, etc, etc. Moreover, even after most of the bugs are ironed out, there are going to be occasional situations -- a road covered with heavy snow (where's the edge?), a fireman or policeman directing traffic up over the sidewalk and up a one-way street in the wrong direction -- that will require human intervention.
(And removing the steering wheel and pedals is almost certainly a bad idea. They'll give us some sort of probably horribly designed joystick or touch screen interface instead?)
I would agree that there may well eventually come a time when overriding the computer -- at least in urban areas -- will be something that can require an explanation after the event. But that time is surely a LONG way in the future.
"The list is intentionally vague because they don't want to limit your designs, but they want you to think about issues that you may not realize are issues"
All well and good. Really.
But California is apparently treating this intentionally vague survey as if it were a concrete list of requirements. Toyota would seem to be correct. That's not a very good idea. Preposterous may not be the right word, but I reckon it'll do. What is Toyota supposed to do to avoid risk of lawsuits or fines or having their test program arbitrarily shut down years and millions of dollars downstream when one of their autonomous vehicles somehow becomes an issue?
Thanks for the link. But that doesn't look like a checklist so much as like a PowerPoint slide. Not that it's evil or stupid, but how does one check off items like "Human Machine Interface: Approaches for communicating information to the driver, occupant and other road users"?
I should think that there must be more detail somewhere.
"2.0 can start off as a project to supply secure connectivity to the military, government and critical infrastructure; internal non-public facing usage. Basically take the important hacking targets off today's internet."
As with Gandhi on Western Civilization, that would be a good idea. I think however there might be some implementation problems due to our complete inability to design and write the software to "supply secure connectivity" to anyone or anything connected to any public network.
"And a ULA employee *should* be able to know when to shoot, and what happens if LOX equipment is hit."
Hitting a target the size of a rocket isn't improbable, but hitting a sensitive area from a mile away without zeroing in the weapon would seem to require extraordinary skill or considerable luck. Perhaps marginally more probable would be a malfunction that somehow propelled a small component from the rocket or fueling gear toward the ULA building at many hundreds of meters per second -- fast enough to overcome drag and hit the building. Seems unlikely, but presumably has to be checked out.
This is entirely too sensible and therefore has no chance whatsoever of being generally adopted.
Reality is that passwords are a huge usability problem that is exacerbated by trying to treat the user as a programmable system component. Sadly, passwords of practical length/format don't, and probably can't, provide much security. And users, in general, are not reliably programmable.
What's my answer? Don't have one. But what folks are trying to do isn't working and probably can't work.
I think that the situation might be worth worrying about. But what do I know?
"Think a non-network engineer can do or wants to do any of that stuff?"
Hell, I don't think most folks who could do that stuff have any desire to actually do it for their household gear... and then deal with the inevitable breakdowns... especially if some clownshow in Redmond or Shanghai is perpetually sending out broken automatic "firmware" updates to enhance security or "user experience".
W95 OSR2 USB support pretty much didn't work at all, ever. It took about five years for USB support in any OS to reach the level of being a crapshoot -- some stuff working flawlessly and other stuff not working at all. After 2002 or so, USB usually worked in Windows. Persuading Unixes to work with USB was a challenge.for another few years after 2002. Not that it couldn't be made to work... eventually... if one was patient enough.
Personally, I thought Windows peaked along about Win95 OSR2 -- which was actually quite a good OS for 1997. Fast, compact, ran OK with next to no memory and was very reliable. Pretty much all downhill from there I think.
But what does one do for applications? And I should think the lack of USB support that works might be an issue.
If one has PCs in their care that have minimal/no exposure to the internet, is updating them at all advisable? It's clear that Microsoft can't QA their products adequately. And they are hardly alone in that. IMO, that probably makes updates a greater risk than malware.
Frankly the "cloud" is increasingly like an uncharted polar sea full of icebergs and rocks. Turning your navigation over to pilots who are questionably competent and quite possibly on drugs as well may not be a good idea. May be best to sail only close to home and only on days when the visibility is good and the seas calm.
"Some places don't even have street names. See Japan..."
Japan does have a building address system. See https://en.wikipedia.org/wiki/... But like a lot of things Japanese it's different than what we Westerners are used to. It works. Mail gets delivered. It's computer comprehensible... probably
"The other one is an unfounded and as-yet unsupported belief that autonomous vehicles will eliminate traffic deaths and accidents."
Not eliminate. Reduce.
It's a reasonable expectation I think. Long term, anyway. Autonomous cars probably aren't going to speed, run red lights, try to beat trains to level crossings, etc, etc, etc. Yes, there will be a large number of accidents -- some serious, some fatal -- while the cars learn to recognize open manhole covers, moose, hand lettered signs that say "BRIDGE OUT". But learn they will.
It'll be interesting to see how "Silicon Valley" folks handle a domain where they are successfully sued for the damages caused by their screw-ups. Because they ARE going to be sued. And they are going to lose a lot of those suits. My guess is that they are not going to like it.
A few months ago, on a whim, I followed my GPS's directions to the next town instead of just driving down the road I knew I should take. I wound up on a dirt road a couple of km back in the hills looking at the (closed) doors of someones' barn.
Well, I'd probably use csv. But few folks outside of IT have any idea what it is. And, in fairness xls does allow formulas and allow plots to be specified. If your journal publishes other kinds of papers besides those dealing with lengthy lists of gene codes, maybe xls is a suitable one-size-fits-all data archive format... maybe... I guess...
Slashdot Mirror
If nothing else, a good demonstration of the popularity of Python when C is overkill
#!/bin/env python
print "Weather prediction: Precipitation"
Or bash
echo "Weather prediction: Precipitation"
"If a significant majority of the solutions show one phase instead of the other, it suggests that there is some predictability of the feature (in this case, the NAO) despite the uncertainty and errors from the initial conditions."
There is, however, some chance that the models share some common incorrect assumption. They don't agree because they are correct, but because they are similar to each other. The Literary Digest effect -- sort of.
Or they may simply agree by chance.
As William Shakespeare might have said "Computers can predict weather a year ahead. So can I. So can any other man. But will the weather come when I summon it?" See Henry IV Part 1: Act 3, Scene 1, Page 3 nfs.sparknotes.com/henry4pt1/page_133.html for details.
Did Microsoft know these things might be used in a football stadium? Did they provide any guidance to the NFl on what sort of connectivity the stadiums must provide? Sounds awfully like yet another instance of the entirely too common practice blaming the user for being unable to use a crappy product.
"For example a file can be 'owned' by any number of groups in windows. But in POSIX it belongs to one group. In NTFS I can do things like user 1 can change the permissions but not read the file but user 2 can not even look at the file (will not see it in the listing) but can modify the permissions. Totally contrived but possible to do. I am not totally sure you could actually pull that off in a posix sort of system."
Well, yes, you probably could do that. But I've always wondered if the potential for the small subset of us who occasionally make mistakes creating truly baffling problems for the rest of you to solve, was worth the benefit.
Gee. You don't think that it could be possible that doing computer security even adequately is beyond what people are capable of actually doing? Golly, that might mean that e-commerce is doomed and that all computers are really good for is research and entertainment.
That might put a kink in some folks plans to promote the cloud into a vehicle that will enrich them beyond all belief.
"To my knowledge Google cars have only been driven in near perfect conditions. Ie no blizzards or blowing snow or ice of any kind."
My impression also. I don't have any problem with that. It's an intelligent way to run an R&D program. The only problem is that Google is probably only 15% or so of the way to having a viable autonomous vehicle that you can drop anywhere where there is an accurate digital streetmap in any sort of weather and not have it be a menace to living creatures and inanimate objects. That's OK also as long as management and the public understand that there is a VERY long way to go.
"also the technology on Google cars is way more unaffordable than a Tesla."
Price comes down when your production run is 20 million units?
"So at some point in the future we will see that computers achieve a higher safety level than any sample of human drivers, while remaining imperfect. At this point, it will probably become necessary to ban manual driving on highways, for the protection of the other people on the highway."
Simplistic. For one thing, automatic things like cars is surely going to turn out to be MUCH harder than most anyone expects. Ask folks who have experience applying computers to real world physical processes. The number and variety of things that can and will go wrong is astonishing. Especially once one ventures of the expressway and away from pre-plotted test circuits to a world of kids, pets, horsecarts, wildlife, poorly marked intersections, bicyclists, construction projects, etc, etc, etc. Moreover, even after most of the bugs are ironed out, there are going to be occasional situations -- a road covered with heavy snow (where's the edge?), a fireman or policeman directing traffic up over the sidewalk and up a one-way street in the wrong direction -- that will require human intervention.
(And removing the steering wheel and pedals is almost certainly a bad idea. They'll give us some sort of probably horribly designed joystick or touch screen interface instead?)
I would agree that there may well eventually come a time when overriding the computer -- at least in urban areas -- will be something that can require an explanation after the event. But that time is surely a LONG way in the future.
"The list is intentionally vague because they don't want to limit your designs, but they want you to think about issues that you may not realize are issues"
All well and good. Really.
But California is apparently treating this intentionally vague survey as if it were a concrete list of requirements. Toyota would seem to be correct. That's not a very good idea. Preposterous may not be the right word, but I reckon it'll do. What is Toyota supposed to do to avoid risk of lawsuits or fines or having their test program arbitrarily shut down years and millions of dollars downstream when one of their autonomous vehicles somehow becomes an issue?
Thanks for the link. But that doesn't look like a checklist so much as like a PowerPoint slide. Not that it's evil or stupid, but how does one check off items like "Human Machine Interface: Approaches for communicating information to the driver,
occupant and other road users"?
I should think that there must be more detail somewhere.
"2.0 can start off as a project to supply secure connectivity to the military, government and critical infrastructure; internal non-public facing usage. Basically take the important hacking targets off today's internet."
As with Gandhi on Western Civilization, that would be a good idea. I think however there might be some implementation problems due to our complete inability to design and write the software to "supply secure connectivity" to anyone or anything connected to any public network.
"I'd rather have a government with a constitution that protects free speech and free expression running parts of the internet like DNS"
And where do you propose to find such a government?
"And a ULA employee *should* be able to know when to shoot, and what happens if LOX equipment is hit."
Hitting a target the size of a rocket isn't improbable, but hitting a sensitive area from a mile away without zeroing in the weapon would seem to require extraordinary skill or considerable luck. Perhaps marginally more probable would be a malfunction that somehow propelled a small component from the rocket or fueling gear toward the ULA building at many hundreds of meters per second -- fast enough to overcome drag and hit the building. Seems unlikely, but presumably has to be checked out.
This is entirely too sensible and therefore has no chance whatsoever of being generally adopted.
Reality is that passwords are a huge usability problem that is exacerbated by trying to treat the user as a programmable system component. Sadly, passwords of practical length/format don't, and probably can't, provide much security. And users, in general, are not reliably programmable.
What's my answer? Don't have one. But what folks are trying to do isn't working and probably can't work.
I think that the situation might be worth worrying about. But what do I know?
"Think a non-network engineer can do or wants to do any of that stuff?"
Hell, I don't think most folks who could do that stuff have any desire to actually do it for their household gear ... and then deal with the inevitable breakdowns ... especially if some clownshow in Redmond or Shanghai is perpetually sending out broken automatic "firmware" updates to enhance security or "user experience".
Are you sure that one IP address per atom is sufficient? Just to be safe, throw in another address byte. So, 272 bit addresses it will be.
W95 OSR2 USB support pretty much didn't work at all, ever. It took about five years for USB support in any OS to reach the level of being a crapshoot -- some stuff working flawlessly and other stuff not working at all. After 2002 or so, USB usually worked in Windows. Persuading Unixes to work with USB was a challenge.for another few years after 2002. Not that it couldn't be made to work ... eventually ... if one was patient enough.
Personally, I thought Windows peaked along about Win95 OSR2 -- which was actually quite a good OS for 1997. Fast, compact, ran OK with next to no memory and was very reliable. Pretty much all downhill from there I think.
But what does one do for applications? And I should think the lack of USB support that works might be an issue.
If one has PCs in their care that have minimal/no exposure to the internet, is updating them at all advisable? It's clear that Microsoft can't QA their products adequately. And they are hardly alone in that. IMO, that probably makes updates a greater risk than malware.
Frankly the "cloud" is increasingly like an uncharted polar sea full of icebergs and rocks. Turning your navigation over to pilots who are questionably competent and quite possibly on drugs as well may not be a good idea. May be best to sail only close to home and only on days when the visibility is good and the seas calm.
"Does it depend on systemd yet?"
No, but it has a systemd minor mode activated by typing ^c,Meta-W,7&$
Just wondering if there is any way I can come by some of whatever the dudes at UoM are smoking. I won't smoke it while driving. I promise.
"Some places don't even have street names. See Japan ..."
Japan does have a building address system. See https://en.wikipedia.org/wiki/... But like a lot of things Japanese it's different than what we Westerners are used to. It works. Mail gets delivered. It's computer comprehensible ... probably
"The other one is an unfounded and as-yet unsupported belief that autonomous vehicles will eliminate traffic deaths and accidents."
Not eliminate. Reduce.
It's a reasonable expectation I think. Long term, anyway. Autonomous cars probably aren't going to speed, run red lights, try to beat trains to level crossings, etc, etc, etc. Yes, there will be a large number of accidents -- some serious, some fatal -- while the cars learn to recognize open manhole covers, moose, hand lettered signs that say "BRIDGE OUT". But learn they will.
It'll be interesting to see how "Silicon Valley" folks handle a domain where they are successfully sued for the damages caused by their screw-ups. Because they ARE going to be sued. And they are going to lose a lot of those suits. My guess is that they are not going to like it.
A few months ago, on a whim, I followed my GPS's directions to the next town instead of just driving down the road I knew I should take. I wound up on a dirt road a couple of km back in the hills looking at the (closed) doors of someones' barn.
Well, I'd probably use csv. But few folks outside of IT have any idea what it is. And, in fairness xls does allow formulas and allow plots to be specified. If your journal publishes other kinds of papers besides those dealing with lengthy lists of gene codes, maybe xls is a suitable one-size-fits-all data archive format ... maybe ... I guess ...