Speaking of potential Radio Shack employees... changing the user string is a perfectly plausible fix.
Let's say you have a bug that creates an expensive UI watch thread. When you change your user agent the UI library will deliver the wrong version of the javascript that either is in a different commit that doesn't have the bug or the script fails to execute on the 'wrong' platform, raises an error to the console and dies (and no longer wasting resources). Sometimes a javascript thread crashing and being killed speeds up a website. You lose some piece of functionality you didn't realize the website was trying to provide and your experience greatly improves. That's the entire concept behind adblockers: trim superfluous javascripts to improve privacy and performance.
BLS 2014: 3.9 million H-1B Visas (New and renewed) 2014: 315,857
315,857 / 3,900,000 = 8%
I'm also incredibly curious how companies are hiring "11% fewer employees" because of H1B visas when H1B employees only make up 8% of the CS workforce (And that's even assuming 100% of all H1B visa applicants go into CS). That would imply that H1B employees are so efficient that the total workforce can be scaled down by nearly 33% thanks to their presence. Which would mean H1B employees are incredible!
Also, decay of plant material under hyrdo reseviors and active aquatic microbial digestion is a source of added methane emmissions. Studies show these emissions may be quite high.
No headlines make these emissions sound "quite high" relative to every other energy source they're in-line with solar and wind. They aren't perfect, but they aren't worse than fossil fuels by a factor of 10.
The faster you launch, the sooner the time between failures. So an increasingly large percent of your time becomes time down due to failures.
I think at some point you just take the 'damn the torpedoes' approach.Insurance should be happy to pay-out and most launches aren't NASA grade one-off missions. So maybe delay a New-Horizons class launch or manned mission until you figure out what went wrong, but meanwhile keep launching GPS satellites which have a very small incremental unit cost. If you lose one... that's why you built spares. I imagine a company like iridium would be more than happy to jump-the-line if you've established a reliability record that is within their acceptable losses rate.
They need the next version of Falcon 9, block 5 before they are properly re-usable.
That's a misinterpretation of what was stated. What Elon Musk said was that they want to consolidate their entire fleet onto one design. That means they will only be flying the current generation once or twice before they phase the 'legacy' design out of their fleet.
There have been no statements about fitness of the design itself, just the business case for maintaining multiple builds. A problem every software developer should be familiar with. If you can, you want to update everybody to a single release and maintain that one branch instead of backporting all of your updates to past releases.
I really wish Battle.net would embrace cross play in general. It's so frustrating to not be able to play with my Xbox Friends or vice versa even if they want to play at a *disadvantage*. B.net claims they don't allow Xbox and PC cross play because it wouldn't be fair to the console players but that should be a choice the console players are willing to make.
I have an Xbox One Controller adapter for my PC and sometimes play Overwatch with it. It's not ideal but I still have fun. They should tune the game for the input device and then have signed drivers which recognize verified controllers and enable auto-aim just like the console controls if you use an official or licensed Xbox One controller on PC. And they should let you decide if you want to play. Considering that match making seems like 95% of the determination of whether I win or not, I'll still enjoy the game with a controller.
Humans may not be able to "take over" but accidents were reduced according to the NHTSA investigation. So... that would mean humans being able to take over aren't necessary.
That being said, there are plenty of videos online of humans taking over so you're claim is patently false. Maybe they can't take over while watching Harry Potter but that's not the intended use.
Except that Android is in practice as locked down as Windows is since nearly all apps require Google services and you can't re-compile Android with Google services including the store.
even if Microsoft allows side loading, they can revoke it at any time and there's nothing anybody can do about it.
Even if Microsoft allows Win32 apps to be installed on Windows XP, they can revoke that ability via an update at any time and there's nothing anybody can do about it.
The logic here is completely false. Microsoft can do *ANYTHING* to *ANY* version of Windows and there is "nothing anybody can do about it". That's as true of Win32 as it is of UWP.
A lot of people do care about Netflix and Netflix on Chrome is only 720p (probably because of previous policies). Now that they're hardening the DRM we'll probably see Edge's exclusive 1080p/Dolby Digital capability go away.
It's only a bad metric if you don't have an anti-commercial software agenda.
I have an office365 subscription as well so I'm not apparently a customer either. But the reason I have a subscription is not so much "Office" as it is just for the 1TB/year of OneDrive.
Google Docs doesn't have all the features of MS Office, but it is "good enough" for most people. Instead of $7 per user per month, it is $0 per month.
Apples to Oranges. Google Docs isn't necessarily free. They also offer a $5/mo and $10/month plan. And Office365 doesn't necessarily cost money. You can have a free Outlook/OneDrive/Office Online account. The difference is that Office365 lets you use the full featured offline apps in addition to the online web apps which are comparable or better than Google Docs.
Also, a proper backup policy could completely eliminate this failure mode. Ironically you could *more easily* secure this with *more* internet integration. Have the backups be incremental and off-site. Setup the off-site service to keep backups for 7 days no matter what. If at any point someone hacks your system, physically insert a "RESET" DVD. Format *everything* back to factory defaults. Load the latest good database and you should be back in business in half an hour.
Who thought it was a good idea for essential systems like this to be online in the first place?!
Someone who understands their most profitable customers: business customers. If your business customers can check-in online through the app and be assigned a room which they can unlock from their phone without ever interacting with the front desk.
"Thank you Samantha for picking Great Hotel again. Your room number is 352. Click here to unlock the door. If you have any problems or questions please dial ## or stop by the front desk."
Obviously the devil is in the details but NFC keycards aren't going anywhere (no changing locks and lost keys) and internet aware locks are the obvious next step of convenience and cost cutting.
That is not true according to current cases dealing specifically with Tesla.
Considering it triggered a federal investigation that cleared Tesla I would say the exact opposite, the current case established that the non-open-source vendor was potentially liable depending on the error.
In the Tesla case the only reason they were let-off was because Tesla's legal team required numerous liability warnings that are enforced by the software and the situation was sufficiently challenging that it would be unreasonable to expect the software to handle the edge case.
All of this though is pointless. The question is never "who is liable" the question is "who can you sue?" and the answer is "everybody, always" regardless of true liability.
sounds like this is just another corrupt individual caught up taking bribes from US government.
To be clear "taking bribes from US government" in this case would almost certainly mean "providing intelligence on Russia's hacking and espionage activities.
No the end game is these people have no jobs because the car drives itself. And the end game is less than a decade away. Same with burger flippers. Same with nearly every factory position.
Musk said it best, "It's not Tesla vs Uber it's Uber vs Everybody." Because the question becomes do you use a personal vehicle or do you use Uber's vehicle?
We can plausibly blame China and Mexico for maybe 5-6 years and then it's going to be plainly obvious when the factories return but not the jobs that we ain't in Kansas anymore and we need a new economic system.
How on earth do all of these separate companies have the same problems where ANY breakage of the system mean planes with schedules pre-determined ages ago cannot fly?
Because all of them depend on software to handle their flight scheduling. Imagine this scenario: a plane lands 30 minutes late into Denver with 200 passengers. 50 of them will now miss their flight. Those 50 passengers are going to set off a cascade of modifications automatically to hundreds of dependencies. No human being can keep thousands of flights and millions of passengers in their head at once. There used to be a lot more slop in the system and margin for error. Now a plane delayed landing is almost certainly a plane delayed departing. The entire system has to minimize the damage by deciding whether it makes sense to delay a handful of flights to ensure they make it or attempt to accommodate them on later flights. And if they delay those later flights how will that impact all of the passengers on those flights, etc etc etc. What about that storm in Chicago which will undoubtedly delay any aircraft who don't leave *now*? There are millions of variables and millions of dependencies and if the satellite tracking system goes down suddenly the system won't know if a plane is going to be on-time or if it'll be an hour late. The only safe conclusion is to just stop all traffic until everything is sorted.
When you stop and think about it an Airline's IT system *IS* the company's day-to-day/minute-to-minute management. So if part of the system goes down in one region that will affect the whole system since a passenger currently in Tokyo very well may be flying to Singapore on a plane arriving from Denver. It's a global network of millions of interdependent pieces.
Many airlines are taking measures to minimize these impacts. It used to be that a plane would leave New York, arrive in Chicago, leave Chicago, arrive in Phoenix, leave Phoenix arrive in Seattle, stay for the night and work its way backwards. If the plane was delayed along any stop everything would get delayed. Airlines are trying to reroute their networks so that a plane just circulates back and forth between 1-2 places in a day.
This isn't so much about modeling thought processes as it is about illustrating how even in a simplified model one of our debugging approaches fails.
The logic that they're arguing appears to be:
"If we can't even properly reverse engineer an extremely simple deterministic computer chip using fault modeling, it's extremely unlikely that we can infer the mechanisms of an extremely complex non-deterministic processor like the brain."
Slashdot Mirror
Go back to your job at Radio Shack.
Speaking of potential Radio Shack employees... changing the user string is a perfectly plausible fix.
Let's say you have a bug that creates an expensive UI watch thread. When you change your user agent the UI library will deliver the wrong version of the javascript that either is in a different commit that doesn't have the bug or the script fails to execute on the 'wrong' platform, raises an error to the console and dies (and no longer wasting resources). Sometimes a javascript thread crashing and being killed speeds up a website. You lose some piece of functionality you didn't realize the website was trying to provide and your experience greatly improves. That's the entire concept behind adblockers: trim superfluous javascripts to improve privacy and performance.
im_thatoneguy and I don't feel like I've ever really even needed to mandate anything I just generally respond to the last incoming message format.
Strangers (American) : Email or Facebook Messenger.
Friends and Family: SMS (for throw away communication) or Facebook Messenger (If I want a log).
Strangers (Foreign): WhatsApp
Intraoffice: Slack
Yeah, but we all know the "Mainstream Media" is dominated by Jews. So when he criticizes the mainstream media he's just proving his point.
BLS 2014: 3.9 million
H-1B Visas (New and renewed) 2014: 315,857
315,857 / 3,900,000 = 8%
I'm also incredibly curious how companies are hiring "11% fewer employees" because of H1B visas when H1B employees only make up 8% of the CS workforce (And that's even assuming 100% of all H1B visa applicants go into CS). That would imply that H1B employees are so efficient that the total workforce can be scaled down by nearly 33% thanks to their presence. Which would mean H1B employees are incredible!
Also, decay of plant material under hyrdo reseviors and active aquatic microbial digestion is a source of added methane emmissions. Studies show these emissions may be quite high.
No headlines make these emissions sound "quite high" relative to every other energy source they're in-line with solar and wind. They aren't perfect, but they aren't worse than fossil fuels by a factor of 10.
Why couldn't they solve crimes with this? It might not be fool proof but I see no reason why it couldn't narrow the search.
The faster you launch, the sooner the time between failures. So an increasingly large percent of your time becomes time down due to failures.
I think at some point you just take the 'damn the torpedoes' approach.Insurance should be happy to pay-out and most launches aren't NASA grade one-off missions. So maybe delay a New-Horizons class launch or manned mission until you figure out what went wrong, but meanwhile keep launching GPS satellites which have a very small incremental unit cost. If you lose one... that's why you built spares. I imagine a company like iridium would be more than happy to jump-the-line if you've established a reliability record that is within their acceptable losses rate.
They need the next version of Falcon 9, block 5 before they are properly re-usable.
That's a misinterpretation of what was stated. What Elon Musk said was that they want to consolidate their entire fleet onto one design. That means they will only be flying the current generation once or twice before they phase the 'legacy' design out of their fleet.
There have been no statements about fitness of the design itself, just the business case for maintaining multiple builds. A problem every software developer should be familiar with. If you can, you want to update everybody to a single release and maintain that one branch instead of backporting all of your updates to past releases.
I really wish Battle.net would embrace cross play in general. It's so frustrating to not be able to play with my Xbox Friends or vice versa even if they want to play at a *disadvantage*. B.net claims they don't allow Xbox and PC cross play because it wouldn't be fair to the console players but that should be a choice the console players are willing to make.
I have an Xbox One Controller adapter for my PC and sometimes play Overwatch with it. It's not ideal but I still have fun. They should tune the game for the input device and then have signed drivers which recognize verified controllers and enable auto-aim just like the console controls if you use an official or licensed Xbox One controller on PC. And they should let you decide if you want to play. Considering that match making seems like 95% of the determination of whether I win or not, I'll still enjoy the game with a controller.
Humans may not be able to "take over" but accidents were reduced according to the NHTSA investigation. So... that would mean humans being able to take over aren't necessary.
That being said, there are plenty of videos online of humans taking over so you're claim is patently false. Maybe they can't take over while watching Harry Potter but that's not the intended use.
I really wish they'd release steam for Android.
Except that Android is in practice as locked down as Windows is since nearly all apps require Google services and you can't re-compile Android with Google services including the store.
even if Microsoft allows side loading, they can revoke it at any time and there's nothing anybody can do about it.
Even if Microsoft allows Win32 apps to be installed on Windows XP, they can revoke that ability via an update at any time and there's nothing anybody can do about it.
The logic here is completely false. Microsoft can do *ANYTHING* to *ANY* version of Windows and there is "nothing anybody can do about it". That's as true of Win32 as it is of UWP.
A lot of people do care about Netflix and Netflix on Chrome is only 720p (probably because of previous policies). Now that they're hardening the DRM we'll probably see Edge's exclusive 1080p/Dolby Digital capability go away.
It's only a bad metric if you don't have an anti-commercial software agenda.
I have an office365 subscription as well so I'm not apparently a customer either. But the reason I have a subscription is not so much "Office" as it is just for the 1TB/year of OneDrive.
In the US, the expense is taken over three years.
That hasn't been true since Obama's stimulus package which lets you take the deduction in full immediately on all capex.
Google Docs doesn't have all the features of MS Office, but it is "good enough" for most people. Instead of $7 per user per month, it is $0 per month.
Apples to Oranges. Google Docs isn't necessarily free. They also offer a $5/mo and $10/month plan. And Office365 doesn't necessarily cost money. You can have a free Outlook/OneDrive/Office Online account. The difference is that Office365 lets you use the full featured offline apps in addition to the online web apps which are comparable or better than Google Docs.
Also, a proper backup policy could completely eliminate this failure mode. Ironically you could *more easily* secure this with *more* internet integration. Have the backups be incremental and off-site. Setup the off-site service to keep backups for 7 days no matter what. If at any point someone hacks your system, physically insert a "RESET" DVD. Format *everything* back to factory defaults. Load the latest good database and you should be back in business in half an hour.
Who thought it was a good idea for essential systems like this to be online in the first place?!
Someone who understands their most profitable customers: business customers. If your business customers can check-in online through the app and be assigned a room which they can unlock from their phone without ever interacting with the front desk.
"Thank you Samantha for picking Great Hotel again. Your room number is 352. Click here to unlock the door. If you have any problems or questions please dial ## or stop by the front desk."
Obviously the devil is in the details but NFC keycards aren't going anywhere (no changing locks and lost keys) and internet aware locks are the obvious next step of convenience and cost cutting.
That is not true according to current cases dealing specifically with Tesla.
Considering it triggered a federal investigation that cleared Tesla I would say the exact opposite, the current case established that the non-open-source vendor was potentially liable depending on the error.
In the Tesla case the only reason they were let-off was because Tesla's legal team required numerous liability warnings that are enforced by the software and the situation was sufficiently challenging that it would be unreasonable to expect the software to handle the edge case.
All of this though is pointless. The question is never "who is liable" the question is "who can you sue?" and the answer is "everybody, always" regardless of true liability.
sounds like this is just another corrupt individual caught up taking bribes from US government.
To be clear "taking bribes from US government" in this case would almost certainly mean "providing intelligence on Russia's hacking and espionage activities.
No the end game is these people have no jobs because the car drives itself. And the end game is less than a decade away. Same with burger flippers. Same with nearly every factory position.
Musk said it best, "It's not Tesla vs Uber it's Uber vs Everybody." Because the question becomes do you use a personal vehicle or do you use Uber's vehicle?
We can plausibly blame China and Mexico for maybe 5-6 years and then it's going to be plainly obvious when the factories return but not the jobs that we ain't in Kansas anymore and we need a new economic system.
How on earth do all of these separate companies have the same problems where ANY breakage of the system mean planes with schedules pre-determined ages ago cannot fly?
Because all of them depend on software to handle their flight scheduling. Imagine this scenario: a plane lands 30 minutes late into Denver with 200 passengers. 50 of them will now miss their flight. Those 50 passengers are going to set off a cascade of modifications automatically to hundreds of dependencies. No human being can keep thousands of flights and millions of passengers in their head at once. There used to be a lot more slop in the system and margin for error. Now a plane delayed landing is almost certainly a plane delayed departing. The entire system has to minimize the damage by deciding whether it makes sense to delay a handful of flights to ensure they make it or attempt to accommodate them on later flights. And if they delay those later flights how will that impact all of the passengers on those flights, etc etc etc. What about that storm in Chicago which will undoubtedly delay any aircraft who don't leave *now*? There are millions of variables and millions of dependencies and if the satellite tracking system goes down suddenly the system won't know if a plane is going to be on-time or if it'll be an hour late. The only safe conclusion is to just stop all traffic until everything is sorted.
When you stop and think about it an Airline's IT system *IS* the company's day-to-day/minute-to-minute management. So if part of the system goes down in one region that will affect the whole system since a passenger currently in Tokyo very well may be flying to Singapore on a plane arriving from Denver. It's a global network of millions of interdependent pieces.
Many airlines are taking measures to minimize these impacts. It used to be that a plane would leave New York, arrive in Chicago, leave Chicago, arrive in Phoenix, leave Phoenix arrive in Seattle, stay for the night and work its way backwards. If the plane was delayed along any stop everything would get delayed. Airlines are trying to reroute their networks so that a plane just circulates back and forth between 1-2 places in a day.
Don't trust
Fixed that for you.
This isn't so much about modeling thought processes as it is about illustrating how even in a simplified model one of our debugging approaches fails.
The logic that they're arguing appears to be:
"If we can't even properly reverse engineer an extremely simple deterministic computer chip using fault modeling, it's extremely unlikely that we can infer the mechanisms of an extremely complex non-deterministic processor like the brain."
Neither LAX nor SFO airports are on the outskirts.
LAX is very close to central LA but SFO is at least 30 minutes from downtown by car and about 45 minutes if I remember correctly by BART.