What about things like software keyloggers on the hardware? Otherwise, I'm all for it, especially in environments that do not by nature deal with sensitive data, where the people are somewhat tech savvy and trustworthy, you do not need powerful workstations for graphics/video, and you do not necessarily care whether they project a professional image with their equipment (sales). Yeah, that leaves out quite a lot unfortunately. But it absolutely can work, and work well in the right environment.
But biofuel research has no immediate profitability or practical necessity (since we have oil), while Twitter is a highly, um, used, uh, and even my refrigerator can use twitter if I, uh, and there's a lot of profit just waiting to be, uh, figured out how to, um, be made somehow [yeah, we're screwed]
I could make all kinds of witty remarks like "MCSE? $10 sounds about right." And I'm not sure what value an A+ actually holds (I expect anyone mildly interested in computers to know more than A+ requires). But really this state of affairs is just sad. I wasn't in on the wealth of 1999 because I was in a different field. I watched and saw the train wreck coming years in advance, like most sane people. Sane people also saw the current train wreck inevitably coming as early as 1998.
I've never made great money. Your numbers still don't gel with my experience. 1999 saw some inexperienced people making six figures (double what you suggested). $50k is not what it was ten years ago because the cost of living actually rose fairly dramatically over the last decade. I'm not in Silicon Valley or Manhattan or anywhere crazy like that, and I don't have expectations greater than a living wage.
There are solid businesses and they are willing to pay a living wage for good people. Many questionable businesses seem to be willing to pay close to a living wage. I see people with certification or a bachelor's degree and some experience making more than you suggest if they are merely competent. Cisco certifications in particular seem to command a premium wage. And I also see a few good people remain unemployed.
At the end of the day, I have to count myself blessed, since after changing careers into IT and going through two startups failing in fairly rapid succession (as an employee), I have once again found a great job.
Good luck with that $10/hour programmer. I feel for both of you.
Good grief, $50,000 won't get you a visionary artist guru! If you're lucky it would get you an aspiring visionary, aspiring artist, OR an aspiring guru. You say you need a programmer, and you're basically willing to pay for a burger flipper. Try the local high school maybe?
The really unfortunate thing is how many actual, established businesses seem to think that $10-$15/hour should get them even a competent support tech. It just goes to show that some people really are dreamers, and also that too many people (even some with talent) are really desperate.
To say that marketing people have emotional intelligence... Well, I suppose many of them do. It seems that society's idea of any type of "intelligence" involves the ability and willingness to be manipulative and perpetrate great evils.
There is a difference between a programmer who acts as an assembly line worker, and, as you put it, a "developer". A programmer is a code monkey. They tend to have training in code and little else. There are quite a lot of them these days.
What you are calling a developer is someone who codes but also functions as a consultant, designer, and sometimes engineer. In enterprises, these functions are usually discrete. Someone will design, someone will engineer and write a spec, and some coders who may or may not have any other skills will code it. In smaller or more flexible settings, there is a role for someone who can fulfill all of these roles, and that is and should be considered worth quite a bit more than a code monkey. The reality is probably that relatively few people are really good at it.
There seems to be a fairly broad consensus that too much separation of labor results in massive inefficiency and often in a poorly designed and implemented product, regardless of good intentions. It seems the more conservative engineers are the holdouts on this, but then many engineers seem to have problems with simple, elegant design. Go figure.
If I could only achieve a similar result. As it is, my wife won't drive even when I'm falling asleep (though she prods me and offers) and often won't alert me to turns I'm missing (she insists that I might actually know where I'm trying to go). Sometimes, she doesn't even gasp when I almost kill us! Any suggestions on how I can get her to share at least _some_ of the work?
While it's often said that a Lame Duck congress can't get much done, it is the perfect time for them to pass unpopular legislation that powerful lobbies want passed. It's one of the few times congress can get away with it while having very few political repercussions.
As far as I know, that's how "trusted computing" has always been implemented. The trusted modules that Apple put in their computers for a brief period of time were not used by the OS - they were there for developers or enterprises to use in precisely the way you suggest. Turns out they never got used for much of anything. You'll find trusted modules in some server hardware. It's once again there for developers or enterprises to use as you suggest.
The fundamental problem is that in order to really implement what you suggest, it has to be done one of two ways - 1) by developers, for a specific application on specific hardware. This means it's not for consumers. 2) at the OS level, meaning that code signing has to be done or at least approved by the OS maker. I suppose you could also do this with a third-party repository, but that would simply shift to a different third party, leaving it outside of your personal control once again.
Trusted computing could perhaps technically be put into the OS and under the control of consumers. The problem there is that since consumers wouldn't know what to trust, it would provide no better protection than the status quo and would perhaps make things worse. So you have to either hand control to OS makers, or leave it in the hands of enterprises or developers, where it currently resides.
In other words, everything is as it should be. If you want such personal control for yourself, buy the hardware and do the development or package maintenance for all packages all by yourself. Or do it in an enterprise. Or perhaps form a cooperative to do all this work. Oh wait, that would once again be essentially moving to a third-party trusted repository system. There is no practicable way for you personally to get all the control and security you are asking for.
I'm certain it takes a great deal of data analysis and expertise to predict that information attacks will increase. Anything that can be properly classified as a safe assumption makes for silly predictions.
More to the point, while the current level of the river may not affect a safe assumption about the rise, the prediction is pretty worthless without such supporting information. Also pretty basic stuff that you almost certainly know and recognize. Yet you try to gloss it over. Not only is this information important, but the entire conversation is pointless without such information. You can't have a clue as to where or how strong to build the levees if you don't even know the water level. Thus, as was originally suggested, there is a very real need for a baseline. This holds true on every data point.
Yes, you need a baseline on the current/usual rainfall. For example, assume it customarily rains 24 hours a day, one inch per hour. You do not know this information. You are told that it is going to rain for a straight week. You predict a rise in the river. Your prediction is wrong, because it is always raining and the river level has assumed equilibrium.
You don't need a baseline on the water level - you need a baseline on the rate of rainfall. To predict a rise in attacks, you would need to know the current frequency of attacks, or at least something about how often they have occurred. Pretty basic stuff here - think before you post.
The CEO would have to be involved in replacing the HR staff. He might do it, but the IT staff said he seems to spend most of his time browsing pr0n instead of running the company.
Being able to "skin" your actors in real time with costumes/makeup will have a profound impact on a lot of films, but also the film making industry itself. I'm sure there are a lot fewer model-makers/matte painters since the advent of CG, will this have the same effect on makeup/props/costumers?
I'm quite sure that the makeup people have plenty of work to do. Have you seen how bad people look in HD? If a camera adds ten pounds, HD adds twenty pounds and twenty years.
It would have been possible for you to respond much more directly to that statement. The iFixit article makes it very clear that the Kinect does pretty hefty onboard processing, resulting in sending a color image and a depth map over what is essentially USB. It's pretty clear that the depth map essentially includes recognition of object positions which is calibrated onboard the Kinect with the information from the microphones, so that the information send to the XBox includes object position along with pre-processed audio for the position of any object. It appears that the Kinect recognizes you as an object and will pan to center you better in the frame all by itself. Did you even look at the array of discrete processors on page 2 of the article? While it may be doubtful that the Kinect runs WinCE, it's quite clear that the XBox does NOT do most of the heavy lifting in processing the image and sound data. Quite the opposite - it looks as though the Kinect provides quite a lot of processed information along with a relatively small amount of raw data. There are no doubt some control commands from the XBox without which the Kinect will not function (i.e. it won't operate by itself with just power), but this most likely is fairly simple message passing which can be reverse engineered with relative ease.
If I were in the position to do so, I would get an in-line logic analyzer just to look at it myself.
This is exactly what I need to pull that phone to display trick Tony pulled in Iron Man 2.
You have to admit that's cool. Some of your friends are watching a movie and you point your phone at the screen to commandeer the display and show video of your recent surgery, or stupid cat tricks, or even live video surveillance of your empty bedroom...
This might help explain why people with dyslexia are as a whole frequently perceived to have somewhat greater intelligence than might be expected. They are forced to pay attention to their environment when they cannot read, and they are forced to pay more attention to what they are reading when they can read.
Indeed, I haven't written cursive in many years, and stopped using it as soon as it was allowed to do so. By high school I simply printed. What I have found also is that the "cursive" they teach is also dead wrong. You can write in a much more beautiful and legible script if you simply have good careful printing, and then simply don't lift your pen.
Additionally, all of the old, beautiful script was written for a purpose - one-off documents couldn't be done with a printing press even. And such documents were done for posterity. Beautiful script isn't written quickly, as the argument for cursive goes. Beautiful script was generally a slow thing, although of course scribes could get pretty quick at writing beautiful script. Another purpose of that beautiful script was to signify the formality and authenticity of documents, along with signatures and seals. Generally only a notary was capable of producing a document in the correct form with nicely-formed, legible script. This function has been entirely replaced.
I'm not necessarily convinced that what we've got now is any better, although it certainly is more accessible.
I take what you are saying, but the claim appears to be that this is not simply a proxy. A simple proxy would not even require client software. My assumption has been that it works a bit differently, in that the client does most of the work - opens and handles the normal connections and somehow obscures the encrypted proxy requests in this stream of normal traffic.
It is admittedly tailored to one country's current blocking strategy at the moment. Whether there is anything of value then depends on whether this is still at its heart a regular proxy that can be easily discovered and blocked, or whether it is rather some type of dynamically changing proxy which could be vastly more difficult to block entirely (similar to some botnet control networks). If the latter, there could be value in this combination of strategies, although it still would not be perfect.
The bits of the article you quote seem clear enough. It's two-way encrypted communication first. On top of that, the encrypted stream is
hidden in a stream of normal traffic, to obscure its existence. While I'm sure it could be possible to nevertheless discover the existence of the stream, it could be dramatically more difficult - if it's obfuscated to look as if it is a part of the normal traffic. Even if it's discovered, it would then have to be completely teased out and decrypted. It's also quite possible that while such an encrypted stream could possibly be discovered in any given instance, that it would be many of orders of magnitude more difficult to create an algorithm to recognize such encrypted streams among all normal traffic. Or it could be easier if statistical analysis is used. Either way, there is a possibility that something of real value has been created here.
While you got the definition of bootlegging correct, you definitely didn't get copyright law correct. Merely making a copy is illegal unless it falls under an exception. Of course, the exceptions used to be considered much broader than they are now considered, at least for practical purposes.
Yeah, if you allowed $160,000/year for that stuff (one well paid or two not so well paid guys, another identical system for redundancy, and backup, power, etc), you'd come up with $2.50/GB/year. $1,000,000/year for that stuff and you'd have a cost of $15/GB/year, which is $1.25/GB/month. Let's go crazy and say $4,000,000/year for salaries, power, network, backup - that's $5/GB/month, amortizing everything over a single year. Compares favorably with $30/GB/month. So your point was what exactly?
Whoosh! You were thinking $25/GB up-front cost, and trying to reduce it. Summary said they were paying $30/GB/month. Which, if you consider only hardware costs, quite literally means that they could buy all the hardware needed for a complete system every single month.
You still need to figure in salaries, maintenance, electricity, cooling - but all of that should not come to $30/GB/month. And yet according to the summary it does, since it would only take one month's allocation to buy all the hardware.
You do not know the history or understand the situation. AOL had a complete private network before internet access was even possible. It competed with the likes of Compuserve and Prodigy. All were private networks, and the internet did not exist outside of universities. Once the internet became available, the private networks gradually became irrelevant. At first, they offered internet access in addition to their private network. Then, nobody cared about them any longer.
Facebook is exactly like this. The type of open source system contemplated here is a distributed system - there is no need for you to run your own facebook - like site. It's a bit more like bittorrent, to use a very bad analogy. If it were possible to do everything you do on Facebook without setting up your own site and without having to do it using a private entity like Facebook, well, then why would you? Over time, people would start using the open, decentralized alternative. Facebook would get gradually less attractive, because it would end up having less of the really interesting people and features over time.
From a historical perspective, this is not only possible, but absolutely inevitable. Private systems always go down in favor of open markets, unless the open markets are forcibly limited. Even then, the inevitable always happens.
Slashdot Mirror
What about things like software keyloggers on the hardware? Otherwise, I'm all for it, especially in environments that do not by nature deal with sensitive data, where the people are somewhat tech savvy and trustworthy, you do not need powerful workstations for graphics/video, and you do not necessarily care whether they project a professional image with their equipment (sales). Yeah, that leaves out quite a lot unfortunately. But it absolutely can work, and work well in the right environment.
But biofuel research has no immediate profitability or practical necessity (since we have oil), while Twitter is a highly, um, used, uh, and even my refrigerator can use twitter if I, uh, and there's a lot of profit just waiting to be, uh, figured out how to, um, be made somehow [yeah, we're screwed]
I could make all kinds of witty remarks like "MCSE? $10 sounds about right." And I'm not sure what value an A+ actually holds (I expect anyone mildly interested in computers to know more than A+ requires). But really this state of affairs is just sad. I wasn't in on the wealth of 1999 because I was in a different field. I watched and saw the train wreck coming years in advance, like most sane people. Sane people also saw the current train wreck inevitably coming as early as 1998.
I've never made great money. Your numbers still don't gel with my experience. 1999 saw some inexperienced people making six figures (double what you suggested). $50k is not what it was ten years ago because the cost of living actually rose fairly dramatically over the last decade. I'm not in Silicon Valley or Manhattan or anywhere crazy like that, and I don't have expectations greater than a living wage.
There are solid businesses and they are willing to pay a living wage for good people. Many questionable businesses seem to be willing to pay close to a living wage. I see people with certification or a bachelor's degree and some experience making more than you suggest if they are merely competent. Cisco certifications in particular seem to command a premium wage. And I also see a few good people remain unemployed.
At the end of the day, I have to count myself blessed, since after changing careers into IT and going through two startups failing in fairly rapid succession (as an employee), I have once again found a great job.
Good luck with that $10/hour programmer. I feel for both of you.
Good grief, $50,000 won't get you a visionary artist guru! If you're lucky it would get you an aspiring visionary, aspiring artist, OR an aspiring guru. You say you need a programmer, and you're basically willing to pay for a burger flipper. Try the local high school maybe?
The really unfortunate thing is how many actual, established businesses seem to think that $10-$15/hour should get them even a competent support tech. It just goes to show that some people really are dreamers, and also that too many people (even some with talent) are really desperate.
To say that marketing people have emotional intelligence... Well, I suppose many of them do. It seems that society's idea of any type of "intelligence" involves the ability and willingness to be manipulative and perpetrate great evils.
There is a difference between a programmer who acts as an assembly line worker, and, as you put it, a "developer". A programmer is a code monkey. They tend to have training in code and little else. There are quite a lot of them these days.
What you are calling a developer is someone who codes but also functions as a consultant, designer, and sometimes engineer. In enterprises, these functions are usually discrete. Someone will design, someone will engineer and write a spec, and some coders who may or may not have any other skills will code it. In smaller or more flexible settings, there is a role for someone who can fulfill all of these roles, and that is and should be considered worth quite a bit more than a code monkey. The reality is probably that relatively few people are really good at it.
There seems to be a fairly broad consensus that too much separation of labor results in massive inefficiency and often in a poorly designed and implemented product, regardless of good intentions. It seems the more conservative engineers are the holdouts on this, but then many engineers seem to have problems with simple, elegant design. Go figure.
If I could only achieve a similar result. As it is, my wife won't drive even when I'm falling asleep (though she prods me and offers) and often won't alert me to turns I'm missing (she insists that I might actually know where I'm trying to go). Sometimes, she doesn't even gasp when I almost kill us! Any suggestions on how I can get her to share at least _some_ of the work?
While it's often said that a Lame Duck congress can't get much done, it is the perfect time for them to pass unpopular legislation that powerful lobbies want passed. It's one of the few times congress can get away with it while having very few political repercussions.
As far as I know, that's how "trusted computing" has always been implemented. The trusted modules that Apple put in their computers for a brief period of time were not used by the OS - they were there for developers or enterprises to use in precisely the way you suggest. Turns out they never got used for much of anything. You'll find trusted modules in some server hardware. It's once again there for developers or enterprises to use as you suggest.
The fundamental problem is that in order to really implement what you suggest, it has to be done one of two ways - 1) by developers, for a specific application on specific hardware. This means it's not for consumers. 2) at the OS level, meaning that code signing has to be done or at least approved by the OS maker. I suppose you could also do this with a third-party repository, but that would simply shift to a different third party, leaving it outside of your personal control once again.
Trusted computing could perhaps technically be put into the OS and under the control of consumers. The problem there is that since consumers wouldn't know what to trust, it would provide no better protection than the status quo and would perhaps make things worse. So you have to either hand control to OS makers, or leave it in the hands of enterprises or developers, where it currently resides.
In other words, everything is as it should be. If you want such personal control for yourself, buy the hardware and do the development or package maintenance for all packages all by yourself. Or do it in an enterprise. Or perhaps form a cooperative to do all this work. Oh wait, that would once again be essentially moving to a third-party trusted repository system. There is no practicable way for you personally to get all the control and security you are asking for.
I'm certain it takes a great deal of data analysis and expertise to predict that information attacks will increase. Anything that can be properly classified as a safe assumption makes for silly predictions.
More to the point, while the current level of the river may not affect a safe assumption about the rise, the prediction is pretty worthless without such supporting information. Also pretty basic stuff that you almost certainly know and recognize. Yet you try to gloss it over. Not only is this information important, but the entire conversation is pointless without such information. You can't have a clue as to where or how strong to build the levees if you don't even know the water level. Thus, as was originally suggested, there is a very real need for a baseline. This holds true on every data point.
Yes, you need a baseline on the current/usual rainfall. For example, assume it customarily rains 24 hours a day, one inch per hour. You do not know this information. You are told that it is going to rain for a straight week. You predict a rise in the river. Your prediction is wrong, because it is always raining and the river level has assumed equilibrium.
You don't need a baseline on the water level - you need a baseline on the rate of rainfall. To predict a rise in attacks, you would need to know the current frequency of attacks, or at least something about how often they have occurred. Pretty basic stuff here - think before you post.
Have they found the Black Lion, and is Princess Allura okay?
Oh, wait, that's Arus.
The CEO would have to be involved in replacing the HR staff. He might do it, but the IT staff said he seems to spend most of his time browsing pr0n instead of running the company.
Being able to "skin" your actors in real time with costumes/makeup will have a profound impact on a lot of films, but also the film making industry itself. I'm sure there are a lot fewer model-makers/matte painters since the advent of CG, will this have the same effect on makeup/props/costumers?
I'm quite sure that the makeup people have plenty of work to do. Have you seen how bad people look in HD? If a camera adds ten pounds, HD adds twenty pounds and twenty years.
It would have been possible for you to respond much more directly to that statement. The iFixit article makes it very clear that the Kinect does pretty hefty onboard processing, resulting in sending a color image and a depth map over what is essentially USB. It's pretty clear that the depth map essentially includes recognition of object positions which is calibrated onboard the Kinect with the information from the microphones, so that the information send to the XBox includes object position along with pre-processed audio for the position of any object. It appears that the Kinect recognizes you as an object and will pan to center you better in the frame all by itself. Did you even look at the array of discrete processors on page 2 of the article? While it may be doubtful that the Kinect runs WinCE, it's quite clear that the XBox does NOT do most of the heavy lifting in processing the image and sound data. Quite the opposite - it looks as though the Kinect provides quite a lot of processed information along with a relatively small amount of raw data. There are no doubt some control commands from the XBox without which the Kinect will not function (i.e. it won't operate by itself with just power), but this most likely is fairly simple message passing which can be reverse engineered with relative ease.
If I were in the position to do so, I would get an in-line logic analyzer just to look at it myself.
This is exactly what I need to pull that phone to display trick Tony pulled in Iron Man 2.
You have to admit that's cool. Some of your friends are watching a movie and you point your phone at the screen to commandeer the display and show video of your recent surgery, or stupid cat tricks, or even live video surveillance of your empty bedroom...
This might help explain why people with dyslexia are as a whole frequently perceived to have somewhat greater intelligence than might be expected. They are forced to pay attention to their environment when they cannot read, and they are forced to pay more attention to what they are reading when they can read.
Indeed, I haven't written cursive in many years, and stopped using it as soon as it was allowed to do so. By high school I simply printed. What I have found also is that the "cursive" they teach is also dead wrong. You can write in a much more beautiful and legible script if you simply have good careful printing, and then simply don't lift your pen.
Additionally, all of the old, beautiful script was written for a purpose - one-off documents couldn't be done with a printing press even. And such documents were done for posterity. Beautiful script isn't written quickly, as the argument for cursive goes. Beautiful script was generally a slow thing, although of course scribes could get pretty quick at writing beautiful script. Another purpose of that beautiful script was to signify the formality and authenticity of documents, along with signatures and seals. Generally only a notary was capable of producing a document in the correct form with nicely-formed, legible script. This function has been entirely replaced.
I'm not necessarily convinced that what we've got now is any better, although it certainly is more accessible.
I take what you are saying, but the claim appears to be that this is not simply a proxy. A simple proxy would not even require client software. My assumption has been that it works a bit differently, in that the client does most of the work - opens and handles the normal connections and somehow obscures the encrypted proxy requests in this stream of normal traffic.
It is admittedly tailored to one country's current blocking strategy at the moment. Whether there is anything of value then depends on whether this is still at its heart a regular proxy that can be easily discovered and blocked, or whether it is rather some type of dynamically changing proxy which could be vastly more difficult to block entirely (similar to some botnet control networks). If the latter, there could be value in this combination of strategies, although it still would not be perfect.
The bits of the article you quote seem clear enough. It's two-way encrypted communication first. On top of that, the encrypted stream is hidden in a stream of normal traffic, to obscure its existence. While I'm sure it could be possible to nevertheless discover the existence of the stream, it could be dramatically more difficult - if it's obfuscated to look as if it is a part of the normal traffic. Even if it's discovered, it would then have to be completely teased out and decrypted. It's also quite possible that while such an encrypted stream could possibly be discovered in any given instance, that it would be many of orders of magnitude more difficult to create an algorithm to recognize such encrypted streams among all normal traffic. Or it could be easier if statistical analysis is used. Either way, there is a possibility that something of real value has been created here.
While you got the definition of bootlegging correct, you definitely didn't get copyright law correct. Merely making a copy is illegal unless it falls under an exception. Of course, the exceptions used to be considered much broader than they are now considered, at least for practical purposes.
You're not doing it right. :)
Yeah, if you allowed $160,000/year for that stuff (one well paid or two not so well paid guys, another identical system for redundancy, and backup, power, etc), you'd come up with $2.50/GB/year. $1,000,000/year for that stuff and you'd have a cost of $15/GB/year, which is $1.25/GB/month. Let's go crazy and say $4,000,000/year for salaries, power, network, backup - that's $5/GB/month, amortizing everything over a single year. Compares favorably with $30/GB/month. So your point was what exactly?
Whoosh! You were thinking $25/GB up-front cost, and trying to reduce it. Summary said they were paying $30/GB/month. Which, if you consider only hardware costs, quite literally means that they could buy all the hardware needed for a complete system every single month.
You still need to figure in salaries, maintenance, electricity, cooling - but all of that should not come to $30/GB/month. And yet according to the summary it does, since it would only take one month's allocation to buy all the hardware.
You do not know the history or understand the situation. AOL had a complete private network before internet access was even possible. It competed with the likes of Compuserve and Prodigy. All were private networks, and the internet did not exist outside of universities. Once the internet became available, the private networks gradually became irrelevant. At first, they offered internet access in addition to their private network. Then, nobody cared about them any longer.
Facebook is exactly like this. The type of open source system contemplated here is a distributed system - there is no need for you to run your own facebook - like site. It's a bit more like bittorrent, to use a very bad analogy. If it were possible to do everything you do on Facebook without setting up your own site and without having to do it using a private entity like Facebook, well, then why would you? Over time, people would start using the open, decentralized alternative. Facebook would get gradually less attractive, because it would end up having less of the really interesting people and features over time.
From a historical perspective, this is not only possible, but absolutely inevitable. Private systems always go down in favor of open markets, unless the open markets are forcibly limited. Even then, the inevitable always happens.