The posts here I'm referring to are mindlessly dumping on the whole idea or possibility of there either being anything novel here or that it would actually work. Similar techniques are quite common in signal processing and audio, and we've been approaching all-digital radio technology incrementally for about 20 years now. The biggest novelty here is that they're claiming to be effectively all digital at 5GHz. While "Microprocessor Tech" may be an annoying marketing buzzword or mangling of terminology, the digital techniques and circuitry are valid, some of them are probably novel, and there are indeed many similarities to digital processing circuits found in microprocessors, as the original press release states. There's hype in their release about "no traditional radio parts," where there's likely to be at least an antenna match, but that's not the level of detail these folks are writing about here. Disappointing for a technology oriented site.
The antenna and matching network form a bandpass filter. For a data converter we'd call them a type of reconstruction filter. (It's funny to see people say things like "all digital circuitry is noisy," and yet they probably listen to MP3 or CD audio...)
Most of the people commenting on this story have no clue about signal processing or radios. It is quite possible to feed a "stream of bits" to an analog filter and create a clean analog signal. This is effectively what 1-bit delta-sigma data converters do, and it is close to what Class-D audio amplifiers do. The trick is indeed doing this with wide bandwidth signals and sufficient oversampling to have good signal quality. To get wide bandwidth at 5GHz, they probably are running the sampling rate in the GHz range to get a few 10's of MHz bandwidth and picking off (filtering/selecting) a harmonic at 5GHz. An antenna and matching network are a type of filter network. There's a lot of innovation in these areas, and it's annoying to see uniformed/.'ers dumping on an area they don't understand.
The folks that discovered the AT&T flaw downloaded information on something like 40,000 people and forwarded it to a journalist. Calling the institution itself after a spot check is pretty tame and seems well intentioned on its face compared to the AT&T situation.
You don't know he designed the replacement. The only documentation is that he signed off on validation of a new part revision from the supplier. This is why we need to let the investigation take it's course before deciding who's guilty and trashing people we don't even know based on news articles and assumptions ranted by lawmakers.
Like everything in engineering, there are numerous tradeoffs - cost, land availability, and safety tradeoffs between different construction methods. Apparently there have been studies of accidents with wide grass medians versus guardrails and decisions are made based available land and conditions. IIRC, the wide grass medians can actually be safer because cars will run off and have a chance to slow down without reaching the other side with oncoming traffic. Notice how many of these wide medians are lower in elevation in the center, so it's uphill to get to the other side. Close guard rails are guaranteed to push the out of control cars back into traffic. In some areas they have both - a wide median of mostly grass with a guardrail right down the middle.
The basic declarations of the IPCC have been consistent for years, namely increased severity of storms with more water content in the atmosphere, and more variation in global weather patterns.
Sony also did not license the format to as many content providers as the VHS companies did, leading to much less content available on betamax than on VHS. I recall this being one of the main issues at the time but I'm having trouble finding an article noting this.
This shouldn't be a question of auditing the quality of the number generator; the research shows you might be fooled. Whatever the actual end design is, production tests are constructed to verify the chip is manufactured to match the design. There are some posts further down discussing production test.
They probably wouldn't just skip scan testing altogether. Too many bad chips would go through and the customer would see a high(er) failure rate of bad chips being received.
The manufacturer could alter the test to match their circuit level change, though. This is easy enough to do.
This attack will succeed if the end customer is relying on the manufacturer to verify the chip electrically and if the customer only performs an optical inspection. The end customer has to run the full electrical tests, as well. Optical measurements can verify the masks are correct and electrical (scan and otherwise) verify the design behavior. This why there are 'trusted foundries' in the U.S....
Optical verification at chip level is quite difficult, and often destructive. There would have to be a sampling scheme in place to hope to catch every die site on the reticle... (A reticle is an array of IC die that is stepped across the wafer for lithographic exposure of resist layers for patterning the material on the wafer.)
We have a network-attached storage drive, a Linux server and several developers using Windows laptops and a few fixed PCs. While it's possible to run Subversion (SVN) directly on the NAS, we run it on the Linux box for which we had already set up a secure tunnel method of access from outside. The Linux SVN repository is simply a link to the real repository on the NAS (which is RAIDed and also backed up more frequently). We run TortiseSVN on the PCs, both fixed and laptops. As others noted, developers only need to pull in the projects they need to their PCs. This is working really well, no hiccups at all and free to use. We've used CVS, Git and a few others and found this gave the best tradeoff of features and complexity for what we do. YMMV.
Can you elaborate: "...weren't able to consider older candidates". Is that because you never found any with even relevant skills or because you were told to age discriminate??
That sounds awful. I know many others that have long stable careers in chunks of about 8 to 10 years at a time with each employer. It probably depends on what part of the industry you're in.
Exactly. I would urge skilled people to stop whining and start their own companies and do it right. After watching two of my big-company employers go downhill due to clueless management that's what I did. Go for it. Those right out of school might not have the same depth/breadth that more experienced engineers have, but they also should have fewer obligations (mortgages etc.). It's not for everyone, but it's also not considered as a possible path by enough people, in my opinion.
The problem with H1-B is that it is employer-based, not that it lets more skilled immigrants in. It's not a zero-sum game. The deeper the talent pool in the US, the more companies will be here and the more new companies will start. I didn't see anything in the recent immigration debate about changing the nature of H1-B though, but they did talk about more visas for students with skills to stay in the US after graduation.
Their argument is that the wire tap does not occur until they read it, apparently that the wiretap order can be applied retroactively.
In the Frontline documentary they talked a lot about how they partially traced communications of the 9/11 hijackers etc. The capability to archive everything would let them find patterns of communication much more quickly than just starting when the wire tap order is issued...
Slashdot Mirror
The posts here I'm referring to are mindlessly dumping on the whole idea or possibility of there either being anything novel here or that it would actually work. Similar techniques are quite common in signal processing and audio, and we've been approaching all-digital radio technology incrementally for about 20 years now. The biggest novelty here is that they're claiming to be effectively all digital at 5GHz. While "Microprocessor Tech" may be an annoying marketing buzzword or mangling of terminology, the digital techniques and circuitry are valid, some of them are probably novel, and there are indeed many similarities to digital processing circuits found in microprocessors, as the original press release states. There's hype in their release about "no traditional radio parts," where there's likely to be at least an antenna match, but that's not the level of detail these folks are writing about here. Disappointing for a technology oriented site.
The antenna and matching network form a bandpass filter. For a data converter we'd call them a type of reconstruction filter. (It's funny to see people say things like "all digital circuitry is noisy," and yet they probably listen to MP3 or CD audio...)
Most of the people commenting on this story have no clue about signal processing or radios. It is quite possible to feed a "stream of bits" to an analog filter and create a clean analog signal. This is effectively what 1-bit delta-sigma data converters do, and it is close to what Class-D audio amplifiers do. The trick is indeed doing this with wide bandwidth signals and sufficient oversampling to have good signal quality. To get wide bandwidth at 5GHz, they probably are running the sampling rate in the GHz range to get a few 10's of MHz bandwidth and picking off (filtering/selecting) a harmonic at 5GHz. An antenna and matching network are a type of filter network. There's a lot of innovation in these areas, and it's annoying to see uniformed /.'ers dumping on an area they don't understand.
The folks that discovered the AT&T flaw downloaded information on something like 40,000 people and forwarded it to a journalist. Calling the institution itself after a spot check is pretty tame and seems well intentioned on its face compared to the AT&T situation.
Link from autonews article here: http://docs.house.gov/meetings...
Link from autonews article here: http://docs.house.gov/meetings...
Link from autonews article here: http://docs.house.gov/meetings...
No, they just swapped in a different Eaton plunger part, according the document linked from the article.
The revision number did change according to the documentation linked in the article.
You don't know he designed the replacement. The only documentation is that he signed off on validation of a new part revision from the supplier. This is why we need to let the investigation take it's course before deciding who's guilty and trashing people we don't even know based on news articles and assumptions ranted by lawmakers.
The document linked in the article shows several revisions to the switch assembly without changing GM p/n (revs A, B, C1, C2, C3 and D).
Like everything in engineering, there are numerous tradeoffs - cost, land availability, and safety tradeoffs between different construction methods. Apparently there have been studies of accidents with wide grass medians versus guardrails and decisions are made based available land and conditions. IIRC, the wide grass medians can actually be safer because cars will run off and have a chance to slow down without reaching the other side with oncoming traffic. Notice how many of these wide medians are lower in elevation in the center, so it's uphill to get to the other side. Close guard rails are guaranteed to push the out of control cars back into traffic. In some areas they have both - a wide median of mostly grass with a guardrail right down the middle.
Citations?
And which of these was on either a per reviewed paper or IPCC report and which were in some other media?
The "hockey stick" graph has been found to be fairly accurate... See summary on Wikipedia, for example:
http://en.m.wikipedia.org/wiki/Hockey_stick_controversy
The basic declarations of the IPCC have been consistent for years, namely increased severity of storms with more water content in the atmosphere, and more variation in global weather patterns.
Sony also did not license the format to as many content providers as the VHS companies did, leading to much less content available on betamax than on VHS. I recall this being one of the main issues at the time but I'm having trouble finding an article noting this.
This shouldn't be a question of auditing the quality of the number generator; the research shows you might be fooled. Whatever the actual end design is, production tests are constructed to verify the chip is manufactured to match the design. There are some posts further down discussing production test.
They probably wouldn't just skip scan testing altogether. Too many bad chips would go through and the customer would see a high(er) failure rate of bad chips being received.
The manufacturer could alter the test to match their circuit level change, though. This is easy enough to do.
This attack will succeed if the end customer is relying on the manufacturer to verify the chip electrically and if the customer only performs an optical inspection. The end customer has to run the full electrical tests, as well. Optical measurements can verify the masks are correct and electrical (scan and otherwise) verify the design behavior. This why there are 'trusted foundries' in the U.S. ...
Optical verification at chip level is quite difficult, and often destructive. There would have to be a sampling scheme in place to hope to catch every die site on the reticle... (A reticle is an array of IC die that is stepped across the wafer for lithographic exposure of resist layers for patterning the material on the wafer.)
Should have said 'tested' not treated... Using swipe on a tablet...
Digital ICs are treated production with scan tests guaranteed to cover around 95 to 99% of possible faults.
We have a network-attached storage drive, a Linux server and several developers using Windows laptops and a few fixed PCs. While it's possible to run Subversion (SVN) directly on the NAS, we run it on the Linux box for which we had already set up a secure tunnel method of access from outside. The Linux SVN repository is simply a link to the real repository on the NAS (which is RAIDed and also backed up more frequently). We run TortiseSVN on the PCs, both fixed and laptops. As others noted, developers only need to pull in the projects they need to their PCs. This is working really well, no hiccups at all and free to use. We've used CVS, Git and a few others and found this gave the best tradeoff of features and complexity for what we do. YMMV.
Can you elaborate: "...weren't able to consider older candidates". Is that because you never found any with even relevant skills or because you were told to age discriminate??
That sounds awful. I know many others that have long stable careers in chunks of about 8 to 10 years at a time with each employer. It probably depends on what part of the industry you're in.
Exactly. I would urge skilled people to stop whining and start their own companies and do it right. After watching two of my big-company employers go downhill due to clueless management that's what I did. Go for it. Those right out of school might not have the same depth/breadth that more experienced engineers have, but they also should have fewer obligations (mortgages etc.). It's not for everyone, but it's also not considered as a possible path by enough people, in my opinion.
The problem with H1-B is that it is employer-based, not that it lets more skilled immigrants in. It's not a zero-sum game. The deeper the talent pool in the US, the more companies will be here and the more new companies will start. I didn't see anything in the recent immigration debate about changing the nature of H1-B though, but they did talk about more visas for students with skills to stay in the US after graduation.
Virginia Tech has had a similar machine up and running in a lobby of one of the engineering buildings since at least Aug 2012...similar name, too. http://www.vt.edu/spotlight/innovation/2012-08-13-3d/dreams.html
Their argument is that the wire tap does not occur until they read it, apparently that the wiretap order can be applied retroactively.
In the Frontline documentary they talked a lot about how they partially traced communications of the 9/11 hijackers etc. The capability to archive everything would let them find patterns of communication much more quickly than just starting when the wire tap order is issued...