The other problems are bad A-D converters, modulating the power supply lines, and the thought of a hundred passengers with their WiFi left on. Or worse. I'd love to go thru 100 cabins with a DC-6ghz spectrum analyzer just to get some samples. I'd bet the info would fry a lot of otherwise thoughtful engineers.
You talk about amateur radio, with a number of mostly AM, SSB, and CW modulated traffic, and not all at 5KW. And they don't affect your *AUDIO* system.
Now imagine air speed indicators, small aperture radar, and some highly sensitive receivers that get bombarded with harmonics from laptops, game controllers, tablet/pads, GSM/CDMA phones, and other personal devices, any or all of which might be blurting out naturals or harmonics from their screens, USB devices, and so on.
This isn't a residential/consumer playing field, it's an aluminum tube with a bunch of unknown stuff blurting and reflecting around a cabin with a bunch of mission critical electronics-- whose shielding might be SOA, or might be generations old. Think about it. Remember your training, and both the 80/20 rule and Murphy's.
Outputs are rendered. Editing might take place with GPUs in the future. With 12 cores, I'll admit that Apple's MacOS desktop hardware has lots of strength. Yet it's the graphics components, and not the CPU (although it has great memory management) that makes FPPro move. Will a 64-bit ARM come soon? Likely. Certainly the memory bounds of a 32-bit CPU are limiting.
But Apple seems poised to move towards ARM chips, even if Intel supplies them. Why? Keeping costs low, controlling the platform, and making battery life longer. Does MacOS do this in the future? Does iOS9? Will iPhone/iPad chips by 2017 be fast and furious enough that we don't care if it's iOS or MacOS?
Using Darwin was an interesting choice for many reasons at the time: Jobs knew NeXtStep was fairly solid, and it was "open-ish" enough to gain a lot of developer support. Now, developer support is pretty strong, and Windows is no longer the low-hanging-fruit picked by developers first. Indeed the App Store is the largest thing going. So there's a loyal cadre of developers getting paid, even for apps that do flatulence.
Business needs still have a commonality of Office-like applications, and especially document generation in many formats. Apple already does that well--- and you're not going to do a fat website on iOS. At the point when you can, you don't need MacOS anymore. That'll be a while-- iOS is only slowly moving towards strong business use cases. MacOS success isn't mystifying, it's just that Apple is making stock price and revenues from consumers right now. People want to take that, make it into a new hammer, the view everything as a nail. It's not.
For many reasons, including the announcements and implications made yesterday at WWDC, it would appear that MacOS is safe, and the point is moot. I think it's speculative at best to consider the demise of iOS at this point; it's not like Microsoft killing off Windows, but it's close in some ways.
In either case, NAT offers *some* protection but may not be viable in some IPv6 and other situations. My recommendation would be to use an appliance to both make stateful examinations of conversations in the firewall sense, use/etc/hosts instead of DNS, examine key vulnerable drivers for MD5, and use other methods to vet basic VMs that are used to clone for production activities. Among other steps.
In other words, from a security profile, KVM and Xen and other methods like LXC each have their own implications. You need to understand them so as to plan for vulnerability containment.
I think there's a Monty Python episode like that, called "Fade to Black". Data ought to have a half-life. Otherwise, it will conquer the world. No one throws stuff out. I wonder how many SAN "accidents" are just frustrated DBAs.
We agree almost categorically. The only difference might be that investigation of outsourced services really pays. I tour facilities, talk to system engineers, and traceroute every bit of gear between one place and another. I beacon systems in pilots, just to find out what else is possibly listening on reachable segments. Is it a layer 2/3 switch with VLANing turned on? Uh oh.
What was old is new again, but with a decided twist. I can spin up thousands of VMs if I need to, but just because they can be spun up doesn't imbue any security metrics. A little research goes a long way.
And for the college grads that short-circuit security, well, there's an app for that called Workforce Development Unemployment Insurance Application.
That's part of your due diligence. If you're using an organization that might change your TOS, give resources to someone else, and not tell you about it, then it's litigation time.
Your bloodthirsty tenaciousness is admirable, but it transports across IP just as well as other transports. If you're using sleazy vendors, you can get bad results inside your NOC as well as at dubious hosting vendors.
The other side of the coin says that Google is getting a bad rep (whether or not deserved) by not exerting the same control that Apple does over its hardware AND software base.
Microsoft is WAY behind the market, and needs to catchup. Using that good old Jobs zen approach of cutting away the distractions might help them catch up. This of course, means that all of the other leaden madness that ties them down is somehow assuaged. But comparing them to SCO is a non-starter. Microsoft probably helped finance SCO just to be a PITA to the FOSS movement. But SCO's market cap wasn't even statistically relevant to Microsoft's. Like them or hate them, Microsoft has unbelievably high cash resources (even after buying Skype) to digest or kill what it wants.
BTW, "emo" is only a fraction of what motivates people to buy Apple stuff.
And you think that the fat drive array in your data center is IN your control? The location doesn't matter if your discipline isn't up to snuff. You'll get eaten internally, or externally-- or not-- if you apply the same studiousness to both.
Cloud is more than storage, and I understand the need for security, regulatory compliance, and general safety. The cloud is just as safe as your internal network, and they require equal effort to make them so. If you choose a reliable cloud provider, then impose your policy discipline and regimen, there is no difference between 'cloud' and your data center.
Those that make excuses for sites that go bad do us all an injustice, just as when your data center gets cracked or you leak data, you deserve a new job flipping burgers.
VisiCalc/SuperCalc/Lotus 123 all won because they could get real work done, rather than having an app built to do repetitive relational math. Because those worked so well, people tried to turn them into word processors because Wang and Lanier and IBM Displaywriters were so expensive. Then they started to sort stuff, and little dbs took off. It wasn't a nexus of storage control, it was impatience that drove the populist computing revolution..... and games.....and pr0n.
No, not really. While MS DOS was whacked, the Win95/98/98SE/ME editions were kinda sorta merged into NT4. Eventually, there was kinda maybe two API sets, one for desktops, and one for "servers".
As we both know, the results were "something for everyone!". I've tracked 61 separate server and desktop versions *currently*. Windows everywhere, or something silly like that.
For instance: Microsoft developed the MultiMedia II spec that hardware systems builders were supposed to adhere to. It contained a basic set of specs that included a CD-ROM, minimum screen and color resolution, and it imposed a discipline on systems design.
Another instance: they actually listened to developers and united the code bases for Windows into Windows 2000 from the DOS branch, and the Win32 branch.
No, not always. They did a few cool and interesting things. Their business practices then became like the people they were battling: proprietary, herding, FUD-driven, and obfuscatory to the max.
Then they pitted divisions against each other, used darwinian project management and became slaves to Wall Street, this ending any morality that was left.
Gates as a leader? No. Visionary? A bit of one, but that's lost to the mythos of others these days.
Thanks for the answers, seemingly as unsatisfying as they are. A side communication reveals that a friend that's read this thread now imagines reincarnation through quantum spiritual movement might be possible (if you'll pardon the pun: in his mind). He may be disappointed.
Parallel computing, and spawning "CPU threads" is a little like what the brain does. Today, there are heavy threads (cost in terms of CPU strokes used, memory displaced through the 'life cycle' of the thread, and so on). Light threads use less overhead, and per given set of launched resources, you can pack more light threads into a period of time as tasks.
Boiling it, there are threads, CPU/core capacity for them within a delta of thread size, and latency per time domain. Some CPUs have interesting instruction sets, big word sizes, pre-fetch, and other enhancing technologies. Let me stroke my chin and offer a response that perhaps clarifies.
Now I'm going to extrapolate, and QM is your field, as I'm just a shade-tree QM mechanic at best. The best entangled pair I know of is a retired couple down the street. My field for background purposes of this response, are very large systems. I'm no neurologist, but have an amateur's understanding of the brain. I'll start from there.
States in transition are what computers and quantum mechanics are about. States are boolean anded together in inventive ways through a transition. We observe the transition and obtain results. When people realize that binary math is moshing transistor states together to obtain results, it's a let down for them. Yes, there are special math CPUs and CPU/functions of FP math that can whack complexity down in their own clever ways.
The brain has multiple concurrent conscious activities going on all the time we're alive. There are fewer when we sleep. The activities draw on multiple concurrent, possibly iterative data sources. Summary so far for the brain: lots going on all the time in terms of concurrency. Kind of like parallelism, with fat and skinny threads being spawned and killed all the time. With age, like mine, there becomes latency for many geriatric reasons.
In QM, particle states are related to each other across time and space. Binding of particles is transient; they are, and are then are not. Within the context of inside the brain, are states bound to either other either through the quantum effect or other synaptic-like connection? I don't believe there's a sufficient answer for this question today.
If the quantum effect, either for memory or state transition influencing is possible inside the brain, then how about across brains, across wide distances. It would explain a lot of phenomena where people seem to know facts instantly about others that are identified as various descriptions of precognition, "psychic" events, even those related to spiritual observations.
I don't know the answers, but there seems to be compelling cases to fit QM into the model for inexplicable events. In terms of computing, predictability is king. Systems sentience hasn't been proven at this point, although there are fascinating applications that mime this, as psychopaths and sociopaths mimic their absent emotive capabilities. We get closer to a programmed sentience, or (I use this hesitatingly) AI that can't be discerned from humanity. This is differentiated from what may be QM effects in the human mind, in its consciousness. I'm deeply interested in the science behind all this, as it's a "meaning of life" sort of question for me, personally.
"Subject" and "Object" is a bad context. It suggests both a timeline and a state of transitiveness. It's a very state-machine way of looking at things.
With QM, you get to have multiple states, and multiple dependencies, each with multiple states. Asking computers, which are state machines, is a silly thing to do. They think like computers. We store memories and have cognition concurrently in multiple areas. To exacerbate the problem, we have two brains in the same head-- mostly.
IMHO, it's an interdisciplinary answer to a tough question. Conscious thought is the brain turned on to at least a sentient state. That's the floor. What goes above the floor are a bunch of things, all transient, until at the end of the day, we go somewhat unconscious at sleep. We know, or should, the difference between conscious reality and sleep/dreaming. Knowing the difference between reality and imagination is a very important sanity check for human brains.
We try to measure capacity for cognition, memory, and other characteristics in lots of ways that all agree are imperfect (think everything from IQ and memory tests, etc.). Some individuals, savants and autistics, have different cognition, and their brains process differently-- or so has been shown in various studies. Does QM play a part? I'm undecided.
Skype uses some nodes as super-nodes, but it's otherwise mostly peer-to-peer. There are landline gateways across the planet, but they're not easily termed as servers.
The Microsoft haters in TFA need some education. Skype hasn't yet been officially acquired by Microsoft, and has had many kinds of outages. A global carrier can't help but to have them occasionally for many different reasons, but Skype wasn't out across the planet. Somebody couldn't connect and thought that Skype was down when some others couldn't logon. Hardly scientific.
I'll admit that it seems like sheer heresy. Soon there'll be people burned at the stake. Even the saints of Motif would castigate those pesky Unity people.
I hear they even have the apostasy of having primitives for *tablets*.
I don't see any stinking Unity in the program. In fact, I'll imagine that if you wore a Unity t-shirt to that conference, you'd be taken out back and spanked, involuntarily....
The whole statistic is misleading and self-serving. Amazon WANTS you to get your reading materials delivered electronically because they make the most profit margin when you buy an e-something.
They're leading the world in terms of complete electronic distribution of materials that favor THEM. So, of course they're starting to sell a majority of e-stuff rather than paper equivalents. The number doesn't mean that the rest of the world is doing this.
If used materials were contributing to the number, they would have mentioned those-- but used e-stuff cuts their margin per SKU delivered to nearly nothing. Selling used e-stuff is their worst nightmare, and publishers aren't happy about it as they make no money, either. Library, sharing, and other methods/myths of content 'ownership' are to be disparaged.
Slashdot Mirror
The other problems are bad A-D converters, modulating the power supply lines, and the thought of a hundred passengers with their WiFi left on. Or worse. I'd love to go thru 100 cabins with a DC-6ghz spectrum analyzer just to get some samples. I'd bet the info would fry a lot of otherwise thoughtful engineers.
You talk about amateur radio, with a number of mostly AM, SSB, and CW modulated traffic, and not all at 5KW. And they don't affect your *AUDIO* system.
Now imagine air speed indicators, small aperture radar, and some highly sensitive receivers that get bombarded with harmonics from laptops, game controllers, tablet/pads, GSM/CDMA phones, and other personal devices, any or all of which might be blurting out naturals or harmonics from their screens, USB devices, and so on.
This isn't a residential/consumer playing field, it's an aluminum tube with a bunch of unknown stuff blurting and reflecting around a cabin with a bunch of mission critical electronics-- whose shielding might be SOA, or might be generations old. Think about it. Remember your training, and both the 80/20 rule and Murphy's.
Outputs are rendered. Editing might take place with GPUs in the future. With 12 cores, I'll admit that Apple's MacOS desktop hardware has lots of strength. Yet it's the graphics components, and not the CPU (although it has great memory management) that makes FPPro move. Will a 64-bit ARM come soon? Likely. Certainly the memory bounds of a 32-bit CPU are limiting.
FCPro can run on a few extra cores, or in the cloud. Rendering ought to be done in the cloud anyway.
But Apple seems poised to move towards ARM chips, even if Intel supplies them. Why? Keeping costs low, controlling the platform, and making battery life longer. Does MacOS do this in the future? Does iOS9? Will iPhone/iPad chips by 2017 be fast and furious enough that we don't care if it's iOS or MacOS?
Using Darwin was an interesting choice for many reasons at the time: Jobs knew NeXtStep was fairly solid, and it was "open-ish" enough to gain a lot of developer support. Now, developer support is pretty strong, and Windows is no longer the low-hanging-fruit picked by developers first. Indeed the App Store is the largest thing going. So there's a loyal cadre of developers getting paid, even for apps that do flatulence.
Business needs still have a commonality of Office-like applications, and especially document generation in many formats. Apple already does that well--- and you're not going to do a fat website on iOS. At the point when you can, you don't need MacOS anymore. That'll be a while-- iOS is only slowly moving towards strong business use cases. MacOS success isn't mystifying, it's just that Apple is making stock price and revenues from consumers right now. People want to take that, make it into a new hammer, the view everything as a nail. It's not.
For many reasons, including the announcements and implications made yesterday at WWDC, it would appear that MacOS is safe, and the point is moot. I think it's speculative at best to consider the demise of iOS at this point; it's not like Microsoft killing off Windows, but it's close in some ways.
In either case, NAT offers *some* protection but may not be viable in some IPv6 and other situations. My recommendation would be to use an appliance to both make stateful examinations of conversations in the firewall sense, use /etc/hosts instead of DNS, examine key vulnerable drivers for MD5, and use other methods to vet basic VMs that are used to clone for production activities. Among other steps.
In other words, from a security profile, KVM and Xen and other methods like LXC each have their own implications. You need to understand them so as to plan for vulnerability containment.
I think there's a Monty Python episode like that, called "Fade to Black". Data ought to have a half-life. Otherwise, it will conquer the world. No one throws stuff out. I wonder how many SAN "accidents" are just frustrated DBAs.
We agree almost categorically. The only difference might be that investigation of outsourced services really pays. I tour facilities, talk to system engineers, and traceroute every bit of gear between one place and another. I beacon systems in pilots, just to find out what else is possibly listening on reachable segments. Is it a layer 2/3 switch with VLANing turned on? Uh oh.
What was old is new again, but with a decided twist. I can spin up thousands of VMs if I need to, but just because they can be spun up doesn't imbue any security metrics. A little research goes a long way.
And for the college grads that short-circuit security, well, there's an app for that called Workforce Development Unemployment Insurance Application.
That's part of your due diligence. If you're using an organization that might change your TOS, give resources to someone else, and not tell you about it, then it's litigation time.
Your bloodthirsty tenaciousness is admirable, but it transports across IP just as well as other transports. If you're using sleazy vendors, you can get bad results inside your NOC as well as at dubious hosting vendors.
The other side of the coin says that Google is getting a bad rep (whether or not deserved) by not exerting the same control that Apple does over its hardware AND software base.
Microsoft is WAY behind the market, and needs to catchup. Using that good old Jobs zen approach of cutting away the distractions might help them catch up. This of course, means that all of the other leaden madness that ties them down is somehow assuaged. But comparing them to SCO is a non-starter. Microsoft probably helped finance SCO just to be a PITA to the FOSS movement. But SCO's market cap wasn't even statistically relevant to Microsoft's. Like them or hate them, Microsoft has unbelievably high cash resources (even after buying Skype) to digest or kill what it wants.
BTW, "emo" is only a fraction of what motivates people to buy Apple stuff.
And you think that the fat drive array in your data center is IN your control? The location doesn't matter if your discipline isn't up to snuff. You'll get eaten internally, or externally-- or not-- if you apply the same studiousness to both.
Cloud is more than storage, and I understand the need for security, regulatory compliance, and general safety. The cloud is just as safe as your internal network, and they require equal effort to make them so. If you choose a reliable cloud provider, then impose your policy discipline and regimen, there is no difference between 'cloud' and your data center.
Those that make excuses for sites that go bad do us all an injustice, just as when your data center gets cracked or you leak data, you deserve a new job flipping burgers.
VisiCalc/SuperCalc/Lotus 123 all won because they could get real work done, rather than having an app built to do repetitive relational math. Because those worked so well, people tried to turn them into word processors because Wang and Lanier and IBM Displaywriters were so expensive. Then they started to sort stuff, and little dbs took off. It wasn't a nexus of storage control, it was impatience that drove the populist computing revolution..... and games.....and pr0n.
No, not really. While MS DOS was whacked, the Win95/98/98SE/ME editions were kinda sorta merged into NT4. Eventually, there was kinda maybe two API sets, one for desktops, and one for "servers".
As we both know, the results were "something for everyone!". I've tracked 61 separate server and desktop versions *currently*. Windows everywhere, or something silly like that.
For instance: Microsoft developed the MultiMedia II spec that hardware systems builders were supposed to adhere to. It contained a basic set of specs that included a CD-ROM, minimum screen and color resolution, and it imposed a discipline on systems design.
Another instance: they actually listened to developers and united the code bases for Windows into Windows 2000 from the DOS branch, and the Win32 branch.
What they do wrong? Out of Memory Error in 29
No, not always. They did a few cool and interesting things. Their business practices then became like the people they were battling: proprietary, herding, FUD-driven, and obfuscatory to the max.
Then they pitted divisions against each other, used darwinian project management and became slaves to Wall Street, this ending any morality that was left.
Gates as a leader? No. Visionary? A bit of one, but that's lost to the mythos of others these days.
Thanks for the answers, seemingly as unsatisfying as they are. A side communication reveals that a friend that's read this thread now imagines reincarnation through quantum spiritual movement might be possible (if you'll pardon the pun: in his mind). He may be disappointed.
Parallel computing, and spawning "CPU threads" is a little like what the brain does. Today, there are heavy threads (cost in terms of CPU strokes used, memory displaced through the 'life cycle' of the thread, and so on). Light threads use less overhead, and per given set of launched resources, you can pack more light threads into a period of time as tasks.
Boiling it, there are threads, CPU/core capacity for them within a delta of thread size, and latency per time domain. Some CPUs have interesting instruction sets, big word sizes, pre-fetch, and other enhancing technologies. Let me stroke my chin and offer a response that perhaps clarifies.
Now I'm going to extrapolate, and QM is your field, as I'm just a shade-tree QM mechanic at best. The best entangled pair I know of is a retired couple down the street. My field for background purposes of this response, are very large systems. I'm no neurologist, but have an amateur's understanding of the brain. I'll start from there.
States in transition are what computers and quantum mechanics are about. States are boolean anded together in inventive ways through a transition. We observe the transition and obtain results. When people realize that binary math is moshing transistor states together to obtain results, it's a let down for them. Yes, there are special math CPUs and CPU/functions of FP math that can whack complexity down in their own clever ways.
The brain has multiple concurrent conscious activities going on all the time we're alive. There are fewer when we sleep. The activities draw on multiple concurrent, possibly iterative data sources. Summary so far for the brain: lots going on all the time in terms of concurrency. Kind of like parallelism, with fat and skinny threads being spawned and killed all the time. With age, like mine, there becomes latency for many geriatric reasons.
In QM, particle states are related to each other across time and space. Binding of particles is transient; they are, and are then are not. Within the context of inside the brain, are states bound to either other either through the quantum effect or other synaptic-like connection? I don't believe there's a sufficient answer for this question today.
If the quantum effect, either for memory or state transition influencing is possible inside the brain, then how about across brains, across wide distances. It would explain a lot of phenomena where people seem to know facts instantly about others that are identified as various descriptions of precognition, "psychic" events, even those related to spiritual observations.
I don't know the answers, but there seems to be compelling cases to fit QM into the model for inexplicable events. In terms of computing, predictability is king. Systems sentience hasn't been proven at this point, although there are fascinating applications that mime this, as psychopaths and sociopaths mimic their absent emotive capabilities. We get closer to a programmed sentience, or (I use this hesitatingly) AI that can't be discerned from humanity. This is differentiated from what may be QM effects in the human mind, in its consciousness. I'm deeply interested in the science behind all this, as it's a "meaning of life" sort of question for me, personally.
"Subject" and "Object" is a bad context. It suggests both a timeline and a state of transitiveness. It's a very state-machine way of looking at things.
With QM, you get to have multiple states, and multiple dependencies, each with multiple states. Asking computers, which are state machines, is a silly thing to do. They think like computers. We store memories and have cognition concurrently in multiple areas. To exacerbate the problem, we have two brains in the same head-- mostly.
IMHO, it's an interdisciplinary answer to a tough question. Conscious thought is the brain turned on to at least a sentient state. That's the floor. What goes above the floor are a bunch of things, all transient, until at the end of the day, we go somewhat unconscious at sleep. We know, or should, the difference between conscious reality and sleep/dreaming. Knowing the difference between reality and imagination is a very important sanity check for human brains.
We try to measure capacity for cognition, memory, and other characteristics in lots of ways that all agree are imperfect (think everything from IQ and memory tests, etc.). Some individuals, savants and autistics, have different cognition, and their brains process differently-- or so has been shown in various studies. Does QM play a part? I'm undecided.
Skype uses some nodes as super-nodes, but it's otherwise mostly peer-to-peer. There are landline gateways across the planet, but they're not easily termed as servers.
The Microsoft haters in TFA need some education. Skype hasn't yet been officially acquired by Microsoft, and has had many kinds of outages. A global carrier can't help but to have them occasionally for many different reasons, but Skype wasn't out across the planet. Somebody couldn't connect and thought that Skype was down when some others couldn't logon. Hardly scientific.
This would be juicy anti-trust and DoJ fodder, something the Obama Administration would get lots of points for. Oh, wait.....
I'll admit that it seems like sheer heresy. Soon there'll be people burned at the stake. Even the saints of Motif would castigate those pesky Unity people.
I hear they even have the apostasy of having primitives for *tablets*.
You say: Unity.
I don't see any stinking Unity in the program. In fact, I'll imagine that if you wore a Unity t-shirt to that conference, you'd be taken out back and spanked, involuntarily....
Which is one of the things that Cyanogen-Mod doesn't do on the original T-Mobile G1: WiFi.
Wanna tether? Have a nice day. Doesn't work. Wifi doesn't work. Bah.
The whole statistic is misleading and self-serving. Amazon WANTS you to get your reading materials delivered electronically because they make the most profit margin when you buy an e-something.
They're leading the world in terms of complete electronic distribution of materials that favor THEM. So, of course they're starting to sell a majority of e-stuff rather than paper equivalents. The number doesn't mean that the rest of the world is doing this.
If used materials were contributing to the number, they would have mentioned those-- but used e-stuff cuts their margin per SKU delivered to nearly nothing. Selling used e-stuff is their worst nightmare, and publishers aren't happy about it as they make no money, either. Library, sharing, and other methods/myths of content 'ownership' are to be disparaged.
Thank you for 1-Clicking Today!