For closed source software to be the best tool for the job, at least one of the following must be true:
1. The best available open source software doesn't do what I need and can't be readily made to do what I need.
2. The closed source software has sufficient APIs to cleanly integrate in to my environment and does a substantially better job than the best available open source software.
3. I just want a throwaway: something cheap that does the job I need done now without requiring any effort on my part. If I throw it away next year, no big deal.
Anyway, those still fighting the closed source/open source divide have missed the boat. The modern threat to open source software is software as a service. When you don't have possession of the object code either and can't even choose to stay with the version you liked, you well and truly have no freedom.
"This is not to say that the NSA should have had all of the authorities it was given. [...] The NSA did its job -- it implemented the authorities it was given."
Just did its job. I've heard something like that before. If I can only remember where...
Exactly. The NSA reached spying on Americans one step at a time, each step thoughtful and with the best of intentions. Some paranoids may be shocked to learn there was neither nefarious intent nor careless disregard of people's rights. Just good intentions and thoughtful compromise.
Fascinating word, compromise. It's represents both the positive give and take of cooperation and the destructive loss of that which is important.
Ugandan commerce has a -severe- shrinkage problem.
Shrinkage is the difference between the number of items entering the supply chain and the number of items sold at retail. It's the total count of items lost, broken or stolen before they can be sold.
The $42 price reflects the two copies that vanish before the third reaches a paying customer. A -severe- shrinkage problem.
Agree. We already knew that knee-jerk reaction time (literally and figuratively) was too fast for congnition to play any real role. This study adds absolutely nothing to that debate. The reporter's characterization of the study's conclusions is entirely specious.
Zero grasp indeed. The last thing you want is uncontrolled server inlet temperatures, whether hot or cold. You want a stable inlet temperature so the boards don't expand, contract and break solder joints. They've also completely overlooked humidity control which should remain between 40% and 55% if you want your servers to keep serving.
And where are the wires? Wires go every which way in a data center and there are many tons of them. A data center without wires is a decoration.
There was a science fiction book called Realtime Interrupt in which wicked advertisers trapped a few people in a virtual reality where they interacted unwittingly with AIs. The point was to test products and advertisements. The first go-around the AIs didn't learn well from the humans so they went off in to weird desires. The second time the AIs took too much of a lead from the humans and started copycatting every weird or violent thing the humans did.
I guess the fine folks at Microsoft didn't read that book.
The Lightcrimp Plus kit is more than $1000 and the connectors are about $10 each. Compare to $60 and $0.25 for copper. A two orders of magnitude difference in price is a problem.
Lightcrimp Plus uses an index matching gel inside the connector so you don't need to glue or polish the end. Cleave the fiber, insert the end into the connector and use a tool to mechanically push the tip into the gel and hold it there. This means you can terminate (attach ends to) a fiber optic cable with about the same level of skill as attaching an RJ45 connector to a copper cable.
The other two approaches to terminating fiber optic cable require training and skill. They are:
1. Polish. You insert the end of the fiber through the center of the connector so that the end just reaches the tip of the connector. Hold it in place with an epoxy or mechanical crimp. Polish the end of the fiber so it's perfectly flat with the connector tip. Tools and supplies are cheap but the process is arduous and highly prone to failure.
2. Fusion splice. Buy preterminated connectors from a factory. Splice the pigtail cables from the connectors with your fiber using a machine which lines them up perfectly and then generates an electrical arc. $10k machine needed to splice cable. Good success rate after much training and experience.
The world won't make any serious move to fiber until the key Amphenol patents on Lightcrimp Plus expire making field terminations easy and cheap. (Lightcrimp Plus already makes them easy but not at all cheap)
Until then specific applications will use fiber but common networking will continue to use twisted pair.
Okay, then the government may be able to produce a software image for the iphone and then command Apple to sign it. Maybe. Reads like the whole deputizing thing is on shaky legal ground.
Engineering *is* all about math and science. Software developers are not engineers. Devs are artists with strong sub-calculus math skills with a knack for flawlessly drawing within the lines while imagining a result far beyond them. We're artists first.
No disrespect, but you don't seem to understand the implications of a "common law" legal system. The Constitution sits atop common law precedent. It does not provide the foundation for common law nor is it expected to detail every niggling little thing the government may do as it employs the powers clearly granted to it.
I don't believe the government can compel someone to produce a new product in support of their investigation. That appears to be prohibited by the Constitution's text against involuntary servitude.
The government may compel third parties to turn over relevant evidence. That's not in the constitution either but it's long settled common law.
Can the police lawfully commandeer a car that's not itself evidence or is that something we only see on TV? If they can commandeer a car then surely they can commandeer an encryption key. Right?
According to the article you referenced, the most damaging thing Milgram did to his data was going off script in an attempt to convince the "teacher" to continue administering shocks long after he had reason to believe the "learner" was seriously injured or dead. I fail to see how that comes anywhere close to discrediting the experiment.
The ethics of the experiment are a whole other matter and the article offers evidence to debunk Milgram's claim that the participants were all properly debriefed. But a failure in compassion does not discredit the experimental results.
If the researchers had designed a correct control for the experiment, they'd know that robots have nothing to do with it. Milgram's Obedience Experiment 50 years ago tells us exactly what happened: people deemed the robot to be an authority, thus followed it uncritically.
Who cares about scientific research and progress? I do. But this wasn't a story about science, it was a story about price. Nobody is worse at predicting price than researchers who discover science but don't industrially produce anything.
Wake me up when they have batteries actually built and selling at that price point. Until then it's just bluster: there's no way to know what industrial challenges will creep in and drive the price up.
Good. I guess he realized that reintroducing slavery and indenture was not something North Carolina wants to be connected with. Seriously, compelling civilian software developers to write software to the government's requirements?
Back when I worked for an ISP my boss had a line he used to describe some of the competitors: "It takes no particular talent to sell a dollar for fifty cents."
"Scientists" should understand the difference between the low earth orbit that an ICBM can almost achieve versus intercepting an object in deep space. Deflect an asteroid a fraction of a degree when it's still a month away and it's certain to miss the planet. When the same asteroid is in range of an ICBM, it's far far too late to do anything.
For closed source software to be the best tool for the job, at least one of the following must be true:
1. The best available open source software doesn't do what I need and can't be readily made to do what I need.
2. The closed source software has sufficient APIs to cleanly integrate in to my environment and does a substantially better job than the best available open source software.
3. I just want a throwaway: something cheap that does the job I need done now without requiring any effort on my part. If I throw it away next year, no big deal.
Anyway, those still fighting the closed source/open source divide have missed the boat. The modern threat to open source software is software as a service. When you don't have possession of the object code either and can't even choose to stay with the version you liked, you well and truly have no freedom.
"This is not to say that the NSA should have had all of the authorities it was given. [...] The NSA did its job -- it implemented the authorities it was given."
Just did its job. I've heard something like that before. If I can only remember where...
Only if you can tell the difference between "I disagree with that gentleman's opinion," and "That dude's just makin' trouble."
Exactly. The NSA reached spying on Americans one step at a time, each step thoughtful and with the best of intentions. Some paranoids may be shocked to learn there was neither nefarious intent nor careless disregard of people's rights. Just good intentions and thoughtful compromise.
Fascinating word, compromise. It's represents both the positive give and take of cooperation and the destructive loss of that which is important.
Not necessarily. They steal a bunch of packages hoping for valuables. The books end up in the garbage.
Ugandan commerce has a -severe- shrinkage problem.
Shrinkage is the difference between the number of items entering the supply chain and the number of items sold at retail. It's the total count of items lost, broken or stolen before they can be sold.
The $42 price reflects the two copies that vanish before the third reaches a paying customer. A -severe- shrinkage problem.
The headline is a lie. You can't project probabilities from a single known example. The math doesn't work that way.
Agree. We already knew that knee-jerk reaction time (literally and figuratively) was too fast for congnition to play any real role. This study adds absolutely nothing to that debate. The reporter's characterization of the study's conclusions is entirely specious.
But then that's what modern reporters do.
Zero grasp indeed. The last thing you want is uncontrolled server inlet temperatures, whether hot or cold. You want a stable inlet temperature so the boards don't expand, contract and break solder joints. They've also completely overlooked humidity control which should remain between 40% and 55% if you want your servers to keep serving.
And where are the wires? Wires go every which way in a data center and there are many tons of them. A data center without wires is a decoration.
There was a science fiction book called Realtime Interrupt in which wicked advertisers trapped a few people in a virtual reality where they interacted unwittingly with AIs. The point was to test products and advertisements. The first go-around the AIs didn't learn well from the humans so they went off in to weird desires. The second time the AIs took too much of a lead from the humans and started copycatting every weird or violent thing the humans did.
I guess the fine folks at Microsoft didn't read that book.
The Lightcrimp Plus kit is more than $1000 and the connectors are about $10 each. Compare to $60 and $0.25 for copper. A two orders of magnitude difference in price is a problem.
Lightcrimp Plus uses an index matching gel inside the connector so you don't need to glue or polish the end. Cleave the fiber, insert the end into the connector and use a tool to mechanically push the tip into the gel and hold it there. This means you can terminate (attach ends to) a fiber optic cable with about the same level of skill as attaching an RJ45 connector to a copper cable.
The other two approaches to terminating fiber optic cable require training and skill. They are:
1. Polish. You insert the end of the fiber through the center of the connector so that the end just reaches the tip of the connector. Hold it in place with an epoxy or mechanical crimp. Polish the end of the fiber so it's perfectly flat with the connector tip. Tools and supplies are cheap but the process is arduous and highly prone to failure.
2. Fusion splice. Buy preterminated connectors from a factory. Splice the pigtail cables from the connectors with your fiber using a machine which lines them up perfectly and then generates an electrical arc. $10k machine needed to splice cable. Good success rate after much training and experience.
The world is moving to fiber. Copper is so 1999.
The world won't make any serious move to fiber until the key Amphenol patents on Lightcrimp Plus expire making field terminations easy and cheap. (Lightcrimp Plus already makes them easy but not at all cheap)
Until then specific applications will use fiber but common networking will continue to use twisted pair.
Okay, then the government may be able to produce a software image for the iphone and then command Apple to sign it. Maybe. Reads like the whole deputizing thing is on shaky legal ground.
Engineering *is* all about math and science. Software developers are not engineers. Devs are artists with strong sub-calculus math skills with a knack for flawlessly drawing within the lines while imagining a result far beyond them. We're artists first.
No disrespect, but you don't seem to understand the implications of a "common law" legal system. The Constitution sits atop common law precedent. It does not provide the foundation for common law nor is it expected to detail every niggling little thing the government may do as it employs the powers clearly granted to it.
I don't believe the government can compel someone to produce a new product in support of their investigation. That appears to be prohibited by the Constitution's text against involuntary servitude.
The government may compel third parties to turn over relevant evidence. That's not in the constitution either but it's long settled common law.
Can the police lawfully commandeer a car that's not itself evidence or is that something we only see on TV? If they can commandeer a car then surely they can commandeer an encryption key. Right?
Does the government have the authority to commandeer Apple's code-signing key so that the phone will accept software built by the government?
According to the article you referenced, the most damaging thing Milgram did to his data was going off script in an attempt to convince the "teacher" to continue administering shocks long after he had reason to believe the "learner" was seriously injured or dead. I fail to see how that comes anywhere close to discrediting the experiment.
The ethics of the experiment are a whole other matter and the article offers evidence to debunk Milgram's claim that the participants were all properly debriefed. But a failure in compassion does not discredit the experimental results.
If the researchers had designed a correct control for the experiment, they'd know that robots have nothing to do with it. Milgram's Obedience Experiment 50 years ago tells us exactly what happened: people deemed the robot to be an authority, thus followed it uncritically.
Who cares about scientific research and progress? I do. But this wasn't a story about science, it was a story about price. Nobody is worse at predicting price than researchers who discover science but don't industrially produce anything.
Wake me up when they have batteries actually built and selling at that price point. Until then it's just bluster: there's no way to know what industrial challenges will creep in and drive the price up.
More to the point, he's asking to argue jury nullification which is, in fact, illegal to do.
Good. I guess he realized that reintroducing slavery and indenture was not something North Carolina wants to be connected with. Seriously, compelling civilian software developers to write software to the government's requirements?
Back when I worked for an ISP my boss had a line he used to describe some of the competitors: "It takes no particular talent to sell a dollar for fifty cents."
"Scientists" should understand the difference between the low earth orbit that an ICBM can almost achieve versus intercepting an object in deep space. Deflect an asteroid a fraction of a degree when it's still a month away and it's certain to miss the planet. When the same asteroid is in range of an ICBM, it's far far too late to do anything.