Funny thing happened to me once, I was actually interested in a very short term contract position, and the client asked me an hourly rate to open negotiations. I was young and I thought "I'll open realy crazy to force them to coutner" and said "$80/hr" and they so quickly said yes without a counter offer and they expressed how relieved they were...
Oh well, was a nice bonus at a rough time in my life anyway.
Well, the model is likely incrooect, but people expecting an eternal increase in percentage... Well it can't go over 100%...
They are saying a person who lives to 122 basically flipped a coin 17 times in a row and came up 'heads' (after they had done the already statistically unlikely task of making it to 105). If the probability did indeed level off, that doesn't mean there is realistic chance of making it very far if the probability is really high.
However, that makes it MORE likely to have duplicates, not less likely.
In either case, if it is truly a private/public key, the chances of generating the same key are astronomically low, even across trillions of tries if doing it properly. For the external approach, there exists the possibility of process error to install the same key to multiple device.
If by chance two onboard devices generate the same key by chance, then the external system can request key regeneration. If this facility is triggered at all, it probably means there's something wrong with your RNG strategy.
Putting aside a debate on whether or not that is a realistic probability, if it is a concern, the device must obvious be outputting a public key and you can have that facility detect a duplicate and request the device regenerate.
Also, if you did build a Dyson Sphere, we probably wouldn't even notice the missing star. If we by some miracle noticed the anomalous gravity behavior, we would postulate a number of possible explanations that could not let us confirm a Dyson Sphere anyway.
The way it should work is for a device to check if it has a key, if it doesn't generates it internally. It should *not* be the case that a device have the private key injected by something externally generating the key. Moving private keys around is bad practice and everything that purports to be secure needs to generate the key on-device rather than accept an external key.
Obviously their process is to generate a key and *then* load that key onto the device. For something like MAC address, this makes sense as you only have 24 bits to play with and you need to coordinate it.
For a cryptographic key, the device should generate key if no key existing and they should be relying upon that to generate keys, rather than any external process. It shouldn't be possible for a private key to be duplicated because it should never be possible for a private key to be extracted or injected into a device.
The general distances and scale do, however, explain why we would be hard pressed to observe them even if they *have* been moving from star to star. We keep assuming that because we haven't seen it, it isn't happening, but it's highly likely we wouldn't even be able to tell.
Space is big and we just can't observe it well enough to proclaim we know a bunch of stuff is missing yet.
Assuming there are no aliens seems a bit strong word, it'd be more along the lines of 'for all practical purposes, it doesn't matter if aliens exist at the moment so long as we don't know'.
You don't proclaim an 'unknown' in science as confidently one way or another.
A space faring civilization going between stars light-months apart still has no bearing on *our* ability to observe them.
There's the question whether anyone would find those megastructures a practical thing to do. And even if they exist and they seem 'mega', we *still* don't have a good way to observe, *particularly* a dyson sphere would be very likely to go unnoticed if complete.
There are things we *guess* we could maybe sort of observe if we look at it the right way, and then there's a whole lot of stuff we have no clue how we would observe at our distance. Go out a few dozen lightyears and we'd probably be hard pressed to measure Sol's ability to support life if we didn't already know the answer.
In this case, as far as I can tell, it's hyper threading, so treating your 16-core cpu as 16 processors is ok (in this specific case) but not as 32 processors.
Of course, they have a lot of sharing within the processor package and it seems plausible for there to be issues like this suggesting more isolation/dedicated resources, but that still can be done in a single processor package too.
Intel Clover Trail based platforms got end-of-supported in 2017.
Such devices started life as Windows 8/8.1 devices. When the users were upgraded to Windows 10, they actually got their support *shortened* in the process (Windows 10 moved to 'lifetime of the device' rather than a period of time and MS decided those devices were 'dead', because Intel decided it wasn't worth supporting MS in doing so).
Keep in mind the context is expecting Intel's ambitions for x86 in mobile, and on that front mobile market is *very* mature.
To your points: > Battery life doesn't fill a day.
This is a design choice that the vendors make. They prioritized thin over battery life. The processor role in this is so limited now that Intel can't even in theory 'fix' it with an awesome new processor.
> Displays are too small
For the class of device, this is the nature of the beast. The only potentially acceptable alternative would involve AR, which is a new class of device. Either way I'm not seeing a processor changing this.
> It's too big to hold. It's too thin to hold
This is subjective, and currently the manufacturers think this opinion is the minority and do not think they are 'wrong' about form factor.
> It can't do anything more than one thing at a time
The platforms are capable and do allow more than one thing at once. The form factor is prohibitive, and a default behavior is to suspend applications, but developers can opt out of suspending if their application has some need.
I have removable microSD card, but even beyond that the platform is very capable of sharing direct wirelessly, but the companies with control direct things to their servers for lock-in, though from a usability perspective that works well too.
Actually, my phone can do video while doing other things. Also can overlay a GPS window. It sucks and I rarely bother to do it because the form factor is too small for this to be pleasant.
This was a feature in Oreo in general devices, some Android Tablets had split screen earlier.
There is a gigantic problem with intel for mobile, when it comes to Android, everyone tests and optimizes for ARM, not Intel.
It's the opposite problem from non-Intel on the desktop. Non-Intel designs *can* do desktop from a technical perspective, in practice the happenstance of current software is just done that way.
Incidentally, 'aside from power consumption' would be a *huge* problem. Intel is doing better than people give them credit for on this front, and it has less to do with the efficiency of running, and everything to do with more fine grained 'sleep' capability than Intel had on desktops previously (more C states, more shallow S states, etc).
I think it's not that it's "supposed" drop anything, it's just that he was injecting the data faster than any human could, and it *looking* like it was being accepted, but in reality it just fell on the floor.
There may be some incorrect technical behavior or just an expected limitation of the input, but either way it doesn't matter for normal use because it's way faster than a human would ever input data.
And there's nothing wrong in trying to be a good corporate citizen - or even to appear to be one.
It's the 'appear to be one' that feels like the big problem in the industry. Trying to 'look good' ends up with less qualified employees *and* being patronizing toward classes of people all at the same time.
Someone who is of a particular minority that is very skilled and has really worked hard finds the position filled by the first minority hire that came along before him, because the organization is biased, but knew it hired to hire 'a' minority and wasn't too picky. This is frankly insulting to that minority group because it presumes the only way to hire is to lower standards.
Similarly with how these companies portray who is responsible for big news items. I was neck deep in a particular project that was in the news and saw a news article proclaiming that 'here are the people who made this possible'. I click to see some people I never had heard of. The company had decided to declare some seemingly random people who had nothing to do with the project at all as responsible. Why? Because it was a better diversity story. The people who were responsible were all white men. This is certainly a problem, but having token minorities to change the optics does little to fix the structural problem, it facilitates denial, and does so in an incredibly expensive and inefficient way.
Of course, this may all be the appropriate strategy in the fullness of time, as otherwise it's a chicken and egg (disenfranchised minorities think it's hopeless due to lack of examples and never try, and it's impossible to get the examples while people are thus discouraged), but it is far from a satisfying situation to watch it first hand, and unclear whether the same biases that prevented any minority hires are being reshaped to treat minority hires as more than a token effort.
Note that I have two Intel based devices and they both can pretty much run all the ARM applications.
They both suck terribly at anything that is vaguely demanding, but in *theory* an x86 based future would be able to run today's software.
Practically speaking I don't see any way for Intel to have some promise of value for x86 architecture in mobile form factor that would overcome the current market situation, but they at least did do their homework and made it technically possible.
Also, they are enabling an arms race (pun unavoidable). TI bowed out of the market because there were just too many competitors that drove them to either leave the market or go negative cash flow to stay in, for example.
So the functional benefits are nice, but more critically they enabled super dirt cheap chip vendors.
I don't think the death of geocities had anything to do with WAP and everything to do with being replaced by MySpace.
The point about having to embrace dead-end transitional technologies is valid, but WAP just didn't matter (it was too crappy to deliver the value of websites and also restrained to the high end of the cellular phone users with the devices and the plans to even get those pathetic chunks of data).
Here I think it's a huge leap to consider use of ARM on handsets a WAP-like fad. WAP was just so niche and ARM handsets are in nearly every pocket in the world and in active use.
At some point we have to recognize that market as mature (it's been a decade now...). At this point, the chances for x86 to suddenly win over the mobile market are about the chances for ARM to suddenly win over the desktop market. The market has spoken and as cool as a unified platform for desktop and mobile *sounds* in practice no one cares/wants it and instead people want their ARM iOS/Android apps for their handsets and their Intel Windows apps for their laptops/desktops. Both MS and Intel have messed around with making their respective platforms work in other contexts, but it doesn't really pan out in the consumer space.
It's at the point where the handset processors are pretty much 'good enough' for the task at hand and the battery life is driven mostly by other factors (radio,screen). There's not much room for a processor to deliver value that will be recognized, even in theory, unless AR/VR applications do take off and we need 8kx4k graphics capability and other neat stuff, and even then it'd probably not have anything to do with the core and everything to do with GPUs and dedicated chip designs for the application.
I've worked with a number of companies acquired by IBM, and what you say is mostly true. This time was unusual as they *really* did seem to try to use that team for a few years before ultimately giving up. They seemed to genuinely think they were unable to win without those people, then the sentiment seemed to become they can't win even with those people, not that they were winning and could cut the team as a result.
Slashdot Mirror
Funny thing happened to me once, I was actually interested in a very short term contract position, and the client asked me an hourly rate to open negotiations. I was young and I thought "I'll open realy crazy to force them to coutner" and said "$80/hr" and they so quickly said yes without a counter offer and they expressed how relieved they were...
Oh well, was a nice bonus at a rough time in my life anyway.
I won't bother to reply to a recruiter who sends something out of the blue.
If I start a conversation however, I will make clear that I'm not interested explicitly once I figure that out.
I have never had a situation where the other party failed to notify me of the situation if they had previously actually replied to my message.
Well, the model is likely incrooect, but people expecting an eternal increase in percentage... Well it can't go over 100%...
They are saying a person who lives to 122 basically flipped a coin 17 times in a row and came up 'heads' (after they had done the already statistically unlikely task of making it to 105). If the probability did indeed level off, that doesn't mean there is realistic chance of making it very far if the probability is really high.
However, that makes it MORE likely to have duplicates, not less likely.
In either case, if it is truly a private/public key, the chances of generating the same key are astronomically low, even across trillions of tries if doing it properly. For the external approach, there exists the possibility of process error to install the same key to multiple device.
If by chance two onboard devices generate the same key by chance, then the external system can request key regeneration. If this facility is triggered at all, it probably means there's something wrong with your RNG strategy.
Putting aside a debate on whether or not that is a realistic probability, if it is a concern, the device must obvious be outputting a public key and you can have that facility detect a duplicate and request the device regenerate.
Also, if you did build a Dyson Sphere, we probably wouldn't even notice the missing star. If we by some miracle noticed the anomalous gravity behavior, we would postulate a number of possible explanations that could not let us confirm a Dyson Sphere anyway.
The way it should work is for a device to check if it has a key, if it doesn't generates it internally. It should *not* be the case that a device have the private key injected by something externally generating the key. Moving private keys around is bad practice and everything that purports to be secure needs to generate the key on-device rather than accept an external key.
It's actually worse than that.
Obviously their process is to generate a key and *then* load that key onto the device. For something like MAC address, this makes sense as you only have 24 bits to play with and you need to coordinate it.
For a cryptographic key, the device should generate key if no key existing and they should be relying upon that to generate keys, rather than any external process. It shouldn't be possible for a private key to be duplicated because it should never be possible for a private key to be extracted or injected into a device.
The general distances and scale do, however, explain why we would be hard pressed to observe them even if they *have* been moving from star to star. We keep assuming that because we haven't seen it, it isn't happening, but it's highly likely we wouldn't even be able to tell.
Space is big and we just can't observe it well enough to proclaim we know a bunch of stuff is missing yet.
Assuming there are no aliens seems a bit strong word, it'd be more along the lines of 'for all practical purposes, it doesn't matter if aliens exist at the moment so long as we don't know'.
You don't proclaim an 'unknown' in science as confidently one way or another.
Also, even if you could build such structures, would it be a practical thing to do?
A space faring civilization going between stars light-months apart still has no bearing on *our* ability to observe them.
There's the question whether anyone would find those megastructures a practical thing to do. And even if they exist and they seem 'mega', we *still* don't have a good way to observe, *particularly* a dyson sphere would be very likely to go unnoticed if complete.
There are things we *guess* we could maybe sort of observe if we look at it the right way, and then there's a whole lot of stuff we have no clue how we would observe at our distance. Go out a few dozen lightyears and we'd probably be hard pressed to measure Sol's ability to support life if we didn't already know the answer.
They wouldn't need to be separated by a socket.
In this case, as far as I can tell, it's hyper threading, so treating your 16-core cpu as 16 processors is ok (in this specific case) but not as 32 processors.
Of course, they have a lot of sharing within the processor package and it seems plausible for there to be issues like this suggesting more isolation/dedicated resources, but that still can be done in a single processor package too.
This already happened.
Intel Clover Trail based platforms got end-of-supported in 2017.
Such devices started life as Windows 8/8.1 devices. When the users were upgraded to Windows 10, they actually got their support *shortened* in the process (Windows 10 moved to 'lifetime of the device' rather than a period of time and MS decided those devices were 'dead', because Intel decided it wasn't worth supporting MS in doing so).
Keep in mind the context is expecting Intel's ambitions for x86 in mobile, and on that front mobile market is *very* mature.
To your points:
> Battery life doesn't fill a day.
This is a design choice that the vendors make. They prioritized thin over battery life. The processor role in this is so limited now that Intel can't even in theory 'fix' it with an awesome new processor.
> Displays are too small
For the class of device, this is the nature of the beast. The only potentially acceptable alternative would involve AR, which is a new class of device. Either way I'm not seeing a processor changing this.
> It's too big to hold. It's too thin to hold
This is subjective, and currently the manufacturers think this opinion is the minority and do not think they are 'wrong' about form factor.
> It can't do anything more than one thing at a time
The platforms are capable and do allow more than one thing at once. The form factor is prohibitive, and a default behavior is to suspend applications, but developers can opt out of suspending if their application has some need.
> It can't project.
Mine can: https://www.motorola.com/us/pr...
> It can't transfer peer-to-peer.
I have removable microSD card, but even beyond that the platform is very capable of sharing direct wirelessly, but the companies with control direct things to their servers for lock-in, though from a usability perspective that works well too.
> It breaks very easily.
Mine doesn't: https://www.youtube.com/watch?...
Actually, my phone can do video while doing other things. Also can overlay a GPS window. It sucks and I rarely bother to do it because the form factor is too small for this to be pleasant.
This was a feature in Oreo in general devices, some Android Tablets had split screen earlier.
There is a gigantic problem with intel for mobile, when it comes to Android, everyone tests and optimizes for ARM, not Intel.
It's the opposite problem from non-Intel on the desktop. Non-Intel designs *can* do desktop from a technical perspective, in practice the happenstance of current software is just done that way.
Incidentally, 'aside from power consumption' would be a *huge* problem. Intel is doing better than people give them credit for on this front, and it has less to do with the efficiency of running, and everything to do with more fine grained 'sleep' capability than Intel had on desktops previously (more C states, more shallow S states, etc).
I think it's not that it's "supposed" drop anything, it's just that he was injecting the data faster than any human could, and it *looking* like it was being accepted, but in reality it just fell on the floor.
There may be some incorrect technical behavior or just an expected limitation of the input, but either way it doesn't matter for normal use because it's way faster than a human would ever input data.
And there's nothing wrong in trying to be a good corporate citizen - or even to appear to be one.
It's the 'appear to be one' that feels like the big problem in the industry. Trying to 'look good' ends up with less qualified employees *and* being patronizing toward classes of people all at the same time.
Someone who is of a particular minority that is very skilled and has really worked hard finds the position filled by the first minority hire that came along before him, because the organization is biased, but knew it hired to hire 'a' minority and wasn't too picky. This is frankly insulting to that minority group because it presumes the only way to hire is to lower standards.
Similarly with how these companies portray who is responsible for big news items. I was neck deep in a particular project that was in the news and saw a news article proclaiming that 'here are the people who made this possible'. I click to see some people I never had heard of. The company had decided to declare some seemingly random people who had nothing to do with the project at all as responsible. Why? Because it was a better diversity story. The people who were responsible were all white men. This is certainly a problem, but having token minorities to change the optics does little to fix the structural problem, it facilitates denial, and does so in an incredibly expensive and inefficient way.
Of course, this may all be the appropriate strategy in the fullness of time, as otherwise it's a chicken and egg (disenfranchised minorities think it's hopeless due to lack of examples and never try, and it's impossible to get the examples while people are thus discouraged), but it is far from a satisfying situation to watch it first hand, and unclear whether the same biases that prevented any minority hires are being reshaped to treat minority hires as more than a token effort.
Note that I have two Intel based devices and they both can pretty much run all the ARM applications.
They both suck terribly at anything that is vaguely demanding, but in *theory* an x86 based future would be able to run today's software.
Practically speaking I don't see any way for Intel to have some promise of value for x86 architecture in mobile form factor that would overcome the current market situation, but they at least did do their homework and made it technically possible.
Also, they are enabling an arms race (pun unavoidable). TI bowed out of the market because there were just too many competitors that drove them to either leave the market or go negative cash flow to stay in, for example.
So the functional benefits are nice, but more critically they enabled super dirt cheap chip vendors.
I don't think the death of geocities had anything to do with WAP and everything to do with being replaced by MySpace.
The point about having to embrace dead-end transitional technologies is valid, but WAP just didn't matter (it was too crappy to deliver the value of websites and also restrained to the high end of the cellular phone users with the devices and the plans to even get those pathetic chunks of data).
Here I think it's a huge leap to consider use of ARM on handsets a WAP-like fad. WAP was just so niche and ARM handsets are in nearly every pocket in the world and in active use.
At some point we have to recognize that market as mature (it's been a decade now...). At this point, the chances for x86 to suddenly win over the mobile market are about the chances for ARM to suddenly win over the desktop market. The market has spoken and as cool as a unified platform for desktop and mobile *sounds* in practice no one cares/wants it and instead people want their ARM iOS/Android apps for their handsets and their Intel Windows apps for their laptops/desktops. Both MS and Intel have messed around with making their respective platforms work in other contexts, but it doesn't really pan out in the consumer space.
It's at the point where the handset processors are pretty much 'good enough' for the task at hand and the battery life is driven mostly by other factors (radio ,screen). There's not much room for a processor to deliver value that will be recognized, even in theory, unless AR/VR applications do take off and we need 8kx4k graphics capability and other neat stuff, and even then it'd probably not have anything to do with the core and everything to do with GPUs and dedicated chip designs for the application.
I've worked with a number of companies acquired by IBM, and what you say is mostly true. This time was unusual as they *really* did seem to try to use that team for a few years before ultimately giving up. They seemed to genuinely think they were unable to win without those people, then the sentiment seemed to become they can't win even with those people, not that they were winning and could cut the team as a result.
Because these guys are passionate about what they were doing and they want to share how they feel their hard work is getting screwed up.