"Crowdsourcing" search is a very bad idea, because crowds can be sourced.
Google shot themselves in the foot this way last October. They used to count reviews on Yelp and Citysearch in determining placement in Google Places, which, until last October, only affected the search engine for Google Maps. Since few people used the search engine for Google Maps to look for businesses by category, it wasn't spammed much.
Then, in October 2010, Google merged Places results into web search. Within two months, Places spam had overwhelmed search. The SEO efforts had become blatant. (Over-the top promotional video). The mainstream press picked up on the issue, and Google looked really stupid. Around November 2010, Google started de-emphasiziing Places results in web search, and stopped counting Citysearch and Yelp results. That brought the problem down to a merely annoying level.
From looking at Verizon forums, the problem seems to be that, unlike everybody else in web hosting, they discontinued FTP without supporting SFTP as a replacement. Most hosting services now require you to use SFTP instead of FTP, and SSH instead of Telnet.
Dreamweaver can use FTP or SFTP, so people with sites big enough to need Dreamweaver have no problem with that.
Currently, root certificates are wildcards, usable for any TLD. They need to be restricted to a single TLD, or a short list.
Single-nation CAs and government-operated CAs should be restricted to their TLD. For the generic TLDs, ("com", ".net", etc,) the CA/Browser Forum should require the CAs to post a large bond, from which a penalty is forfeited if any improperly issued cert is found. That should get the problem under control.
Is this push for more features, more releases, and lower quality coming from the people Google has working on Mozilla? Google has an incentive to migrate people to Chrome, where they define and control the platform.
My mail is on one of my hosting accounts, filtered by SpamAssassin. My videos are on blip.tv, although there's some legacy stuff on YouTube from before Google acquired it. Open source code is on SourceForge. Still pictures go on my web sites. I have a Facebook account, with platform applications turned off and all Zynga content blocked, which makes that tolerable.
I usually search with Google, but I'm never logged into Google. Anything medical-related or in a heavily spammed area, I search on Blekko.
Google has to be very careful for the next two years, because of the terms of their non-prosecution deal with the Justice Department over the drug ads issue. This is the one where Google management had to admit criminal guilt and pay $500 million dollars. For the next two years, if Google does anything out of line in the drug-ad area, DOJ can, at their sole discretion, bring felony criminal charges for Google's past actions. Read that agreement between Google and DOJ. Nobody signs something like that unless going to trial would be much worse.
Peter Neronha, the U. S. attorney who headed the prosecution, issued a statement yesterday. He says that "Larry Page knew what was going on. We know it from the investigation. We simply know it from the documents we reviewed, witnesses that we interviewed, that Larry Page knew what was going on". He went on to say that "this is not two or three rogue employees at the customer service level doing this on their own. This was a corporate decision to engage in this conduct.", and called Google's attempts to control the problem "window dressing".
Google now has to clean up their act. It's not voluntary any more.
The sub-$200 Linux netbook market seems to have completely disappeared, killed by Microsoft. There's some MeeGo crap, but that's tethered to an "app store", so it's like buying a subsidized phone. ("Creates a direct connection between your wallet and our bank account.")
I do enough input that I want a keyboard. Tablets are for passive consumers; you know, TV watchers.
The Pennsylvania Railroad tunnels to New Jersey already had flood gates (the PRR built to last), but they'd been neglected and weren't working. Amtrak has since fixed them.
The CIA has few undercover "spies", in the classic sense. Rarely does an intelligence agency have someone on the inside of an enemy. Usually, they have their people ("case workers") on the outside, who recruit people ("assets") on the inside.
During the Cold War, the KGB and CIA station chiefs in many cities knew who their counterpart was.
Yet the New York City Mafia families were broken partly because the FBI and the NYPD were able to get their people into positions of trust within the Mafia.
The paper is about wireless pairing, which is a special case.
MITM attacks in general are not entirely invisible. Because the MITM is decrypting and reencrypting the message with a different key, the crypt bits received are different than those which were sent. If you ask "what were the crypto bits you received from bit N to M?", the MITM has to be prepared to intercept that query and formulate a lie. This can be made difficult for the MITM. The early STU-3 encrypted phone sets had a little 2-digit display, and the parties could verify over the voice link that both parties saw the same number. Faking that would require splicing words into a verbal conversation in real time.
It's thus possible to design protocols which require that a MITM tamper with the plaintext merely to listen in. This idea doesn't seem to have been developed enough, at least not in the unclassified community.
The huge sales at $99 as HP liquidated their inventory of tablets indicates that the only problem is price. Generic Android tablets are now available in quantity for $50 to $109., FOB Shenzen, China. (Yes, some people who look at that will whine about the processor and memory specs. So?) Price won't be a problem for much longer.
Lambda expressions are not closures, and neither enable parallelization.
I noticed that too. Somehow, in the Java community, closures somehow became connected to lambda expressions. Whether you have a syntax for anonymous functions is completely separate from whether you have closures.
In a reasonably static language like Java, adding closures makes the concept of the "stack" a lot more complex. A closure can contain a binding to a local variable outside the closure. That binding can outlive the scope of that local variable. So local variables now have to be handled in a more general (and slower) way. The semantics of local variables gets a lot more complicated, too.
Perl, Python, and Javascript have closures, although most programmers who use those languages don't know it. In those languages, any nested function can potentially be a closure. (Well, in Perl, there's a warning if you do that.) The Java people don't seem to have taken that route, possibly because they don't want to incur the overhead of a closure unless the programmer really wants one.
Is an ISP or Google a carrier, content provider or both?
In the US, there's what's called Section 230 immunity for ISPs, which is part of the Communications Decency Act. However, Google's offenses relate to their advertising activities, which is a separate issue.
"Further, from 2003 through 2009, Google provided customer support to some of these Canadian online pharmacy advertisers to assist them in placing and optimizing their AdWords advertisements, and in improving the effectiveness of their websites."
"The investigation of Google had its origins in a separate, multimillion dollar financial fraud investigation unrelated to Google, the main target of which fled to Mexico. While a fugitive, he began to advertise the unlawful sale of drugs through Googleâ(TM)s AdWords program. After being apprehended in Mexico and returned to the United States by the U.S. Secret Service, he began cooperating with law enforcement and provided information about his use of the AdWords program. During the ensuing investigation of Google, the government established a number of undercover websites for the purpose of advertising the unlawful sale of controlled and non-controlled substances through Googleâ(TM)s AdWords program."
Tomorrow there will be a press release from the prosecutor. Some previous stories indicate that the drug ad business went beyond accidentally running such ads.
Additive machining is cute, but not a miracle.
It's a slow process. Building up objects one layer at a time takes forever. The consumables are rather expensive. Injection molding and casting are probably 100x cheaper in quantity.
High-endadditive machining system are getting to be quite good. The low-end machines, though, are not yet very useful. The precision is too low, the surface quality is poor, and the material options are too limited. TechShop has both a high-end commercial machine, which is usually busy, and a machine at the MakerBot level, which is almost never used. If you're making tiny parts, you need high precision.
The big advantage of many of the additive processes is that they don't have work-holding problems. The big limitation of CNC machining is that you have to clamp down the workpiece, and the clamps get in the way of what you're doing. Some part of the workpiece will be inaccessible. So most work requires multiple setups, each of which has to be aligned with the previous setup to 0.001in or better. Designs have to be planned to be clampable.
The more interesting processes can work metals. But they need 500W to 6KW lasers. If you're going to work in steel, you need enough power to melt steel.
For comparison, here's a high speed stamping press. This is how most of the small metal parts in the world are made. Once you get the tooling set up, parts come out at machine-gun speeds.
Russia is planning to extend the Trans-Siberian Railway to the Pacific Ocean by 2030. Only then is a tunnel worth considering. Also, there's no rail connection on the US side.
Why the hell should an API, the computer equivalent of a phone number, qualify for copyright protection?
An API is not the "equivalent of a phone number". Not even close. It's clearly a creative work. The question is whether it is too abstract to be protected by copyright. "Copyright protection is not available for ideas, program logic, algorithms, systems, methods, concepts, or layouts.", says the Copyright Office. It might be patentable, if sufficiently original, but that's not an issue here.
Copyright in fictional characters has been tried, occasionally with success. That's one of the broadest forms of copyright protection. Trademarking the name of the character provides more protection, which is why we see Darth Vader(tm) markings. That doesn't apply to an API.
So, why are both ICMP and IP considered to be in layer 3? ICMP is built on top of IP.
The real answer to that is that it's a Berkeley UNIXism. Some early TCP/IP implementations, including the one I worked on, had ICMP at a layer above IP, in the same layer with TCP and UDP. The Berkeley UNIX kernel, like other UNIX versions of the period, had real trouble communicating upward within the kernel, because this was before threads, let alone kernel threads.
To get around that kernel limitation, ICMP was crammed in with IP. This had some downsides, including the demise of ICMP Source Quench for congestion control, which didn't fit well into the mode of ICMP as an error-reporting mechanism for IP.
Slashdot Mirror
"Submit a patch" is open source's way of telling you to fuck off.
Submitting a bug report usually gets some response like expecting the user to build the thing from the source repository and repeat the bug.
"Crowdsourcing" search is a very bad idea, because crowds can be sourced.
Google shot themselves in the foot this way last October. They used to count reviews on Yelp and Citysearch in determining placement in Google Places, which, until last October, only affected the search engine for Google Maps. Since few people used the search engine for Google Maps to look for businesses by category, it wasn't spammed much.
Then, in October 2010, Google merged Places results into web search. Within two months, Places spam had overwhelmed search. The SEO efforts had become blatant. (Over-the top promotional video). The mainstream press picked up on the issue, and Google looked really stupid. Around November 2010, Google started de-emphasiziing Places results in web search, and stopped counting Citysearch and Yelp results. That brought the problem down to a merely annoying level.
Then in 2011, Page announced that Google employee bonuses would be dependent on Google getting into "social". So now "social" had to be built into everything. Hence "+1", followed, inevitably, by a market in "+1" boosts. See any black hat SEO forum to buy.
So, to fix that problem, Google has to have more information about Google account holders. This, as Eric Schmidt points out, can be obtained by insisting on real names and mining the user's social data.
Maybe they're doing it wrong.
From looking at Verizon forums, the problem seems to be that, unlike everybody else in web hosting, they discontinued FTP without supporting SFTP as a replacement. Most hosting services now require you to use SFTP instead of FTP, and SSH instead of Telnet.
Dreamweaver can use FTP or SFTP, so people with sites big enough to need Dreamweaver have no problem with that.
Currently, root certificates are wildcards, usable for any TLD. They need to be restricted to a single TLD, or a short list.
Single-nation CAs and government-operated CAs should be restricted to their TLD. For the generic TLDs, ("com", ".net", etc,) the CA/Browser Forum should require the CAs to post a large bond, from which a penalty is forfeited if any improperly issued cert is found. That should get the problem under control.
Is this push for more features, more releases, and lower quality coming from the people Google has working on Mozilla? Google has an incentive to migrate people to Chrome, where they define and control the platform.
My mail is on one of my hosting accounts, filtered by SpamAssassin. My videos are on blip.tv, although there's some legacy stuff on YouTube from before Google acquired it. Open source code is on SourceForge. Still pictures go on my web sites. I have a Facebook account, with platform applications turned off and all Zynga content blocked, which makes that tolerable. I usually search with Google, but I'm never logged into Google. Anything medical-related or in a heavily spammed area, I search on Blekko.
What Google services?
Mr. Schmidt gets to explain that to the Senate Judiciary Committee on September 21st. That's on top of Google's other legal problems.
Google has to be very careful for the next two years, because of the terms of their non-prosecution deal with the Justice Department over the drug ads issue. This is the one where Google management had to admit criminal guilt and pay $500 million dollars. For the next two years, if Google does anything out of line in the drug-ad area, DOJ can, at their sole discretion, bring felony criminal charges for Google's past actions. Read that agreement between Google and DOJ. Nobody signs something like that unless going to trial would be much worse.
Peter Neronha, the U. S. attorney who headed the prosecution, issued a statement yesterday. He says that "Larry Page knew what was going on. We know it from the investigation. We simply know it from the documents we reviewed, witnesses that we interviewed, that Larry Page knew what was going on". He went on to say that "this is not two or three rogue employees at the customer service level doing this on their own. This was a corporate decision to engage in this conduct.", and called Google's attempts to control the problem "window dressing".
Google now has to clean up their act. It's not voluntary any more.
The average selling price for subnotebooks rose to $521 last july from $343 in July of 2010." The industry is desperately trying to stop generic $200 machines from taking over the industry.
The sub-$200 Linux netbook market seems to have completely disappeared, killed by Microsoft. There's some MeeGo crap, but that's tethered to an "app store", so it's like buying a subsidized phone. ("Creates a direct connection between your wallet and our bank account.")
I do enough input that I want a keyboard. Tablets are for passive consumers; you know, TV watchers.
After 9/11, emergency flood gates were supposed to be installed in the NYC subway system. Water from fire hoses alone was enough to eventually completely flood the PATH tunnels to New Jersey. If the cement box that kept the Hudson River from pouring into the site had cracked open, the subway system would have flooded up to midtown. As of late 2010, some flood gates were being installed.
The Pennsylvania Railroad tunnels to New Jersey already had flood gates (the PRR built to last), but they'd been neglected and weren't working. Amtrak has since fixed them.
They'll have to go back to informants.
The CIA has few undercover "spies", in the classic sense. Rarely does an intelligence agency have someone on the inside of an enemy. Usually, they have their people ("case workers") on the outside, who recruit people ("assets") on the inside.
During the Cold War, the KGB and CIA station chiefs in many cities knew who their counterpart was.
Yet the New York City Mafia families were broken partly because the FBI and the NYPD were able to get their people into positions of trust within the Mafia.
The paper is about wireless pairing, which is a special case.
MITM attacks in general are not entirely invisible. Because the MITM is decrypting and reencrypting the message with a different key, the crypt bits received are different than those which were sent. If you ask "what were the crypto bits you received from bit N to M?", the MITM has to be prepared to intercept that query and formulate a lie. This can be made difficult for the MITM. The early STU-3 encrypted phone sets had a little 2-digit display, and the parties could verify over the voice link that both parties saw the same number. Faking that would require splicing words into a verbal conversation in real time.
It's thus possible to design protocols which require that a MITM tamper with the plaintext merely to listen in. This idea doesn't seem to have been developed enough, at least not in the unclassified community.
The huge sales at $99 as HP liquidated their inventory of tablets indicates that the only problem is price. Generic Android tablets are now available in quantity for $50 to $109., FOB Shenzen, China. (Yes, some people who look at that will whine about the processor and memory specs. So?) Price won't be a problem for much longer.
Lambda expressions are not closures, and neither enable parallelization.
I noticed that too. Somehow, in the Java community, closures somehow became connected to lambda expressions. Whether you have a syntax for anonymous functions is completely separate from whether you have closures.
In a reasonably static language like Java, adding closures makes the concept of the "stack" a lot more complex. A closure can contain a binding to a local variable outside the closure. That binding can outlive the scope of that local variable. So local variables now have to be handled in a more general (and slower) way. The semantics of local variables gets a lot more complicated, too.
Perl, Python, and Javascript have closures, although most programmers who use those languages don't know it. In those languages, any nested function can potentially be a closure. (Well, in Perl, there's a warning if you do that.) The Java people don't seem to have taken that route, possibly because they don't want to incur the overhead of a closure unless the programmer really wants one.
Which Techshop? San Jose doesn't have any additive manufacturing equipment at the moment.
Menlo Park has a pro machine, and San Francisco has something slightly above the MakerBot level.
It seems like you are trying to imply that anyone has suggested these systems for mass production of parts. No one has.
Yes, people have. Look up "personal fabrication revolution" for some of the looser talk on the subject.
Is an ISP or Google a carrier, content provider or both?
In the US, there's what's called Section 230 immunity for ISPs, which is part of the Communications Decency Act. However, Google's offenses relate to their advertising activities, which is a separate issue.
From the DOJ:
"Further, from 2003 through 2009, Google provided customer support to some of these Canadian online pharmacy advertisers to assist them in placing and optimizing their AdWords advertisements, and in improving the effectiveness of their websites."
"The investigation of Google had its origins in a separate, multimillion dollar financial fraud investigation unrelated to Google, the main target of which fled to Mexico. While a fugitive, he began to advertise the unlawful sale of drugs through Googleâ(TM)s AdWords program. After being apprehended in Mexico and returned to the United States by the U.S. Secret Service, he began cooperating with law enforcement and provided information about his use of the AdWords program. During the ensuing investigation of Google, the government established a number of undercover websites for the purpose of advertising the unlawful sale of controlled and non-controlled substances through Googleâ(TM)s AdWords program."
Tomorrow there will be a press release from the prosecutor. Some previous stories indicate that the drug ad business went beyond accidentally running such ads.
Additive machining is cute, but not a miracle. It's a slow process. Building up objects one layer at a time takes forever. The consumables are rather expensive. Injection molding and casting are probably 100x cheaper in quantity.
High-end additive machining system are getting to be quite good. The low-end machines, though, are not yet very useful. The precision is too low, the surface quality is poor, and the material options are too limited. TechShop has both a high-end commercial machine, which is usually busy, and a machine at the MakerBot level, which is almost never used. If you're making tiny parts, you need high precision.
The big advantage of many of the additive processes is that they don't have work-holding problems. The big limitation of CNC machining is that you have to clamp down the workpiece, and the clamps get in the way of what you're doing. Some part of the workpiece will be inaccessible. So most work requires multiple setups, each of which has to be aligned with the previous setup to 0.001in or better. Designs have to be planned to be clampable.
The more interesting processes can work metals. But they need 500W to 6KW lasers. If you're going to work in steel, you need enough power to melt steel.
For comparison, here's a high speed stamping press. This is how most of the small metal parts in the world are made. Once you get the tooling set up, parts come out at machine-gun speeds.
ITunes? AppStore? Apple has preloaded marketing-oriented apps; they're just all Apple's.
Russia is planning to extend the Trans-Siberian Railway to the Pacific Ocean by 2030. Only then is a tunnel worth considering. Also, there's no rail connection on the US side.
Why the hell should an API, the computer equivalent of a phone number, qualify for copyright protection?
An API is not the "equivalent of a phone number". Not even close. It's clearly a creative work. The question is whether it is too abstract to be protected by copyright. "Copyright protection is not available for ideas, program logic, algorithms, systems, methods, concepts, or layouts.", says the Copyright Office. It might be patentable, if sufficiently original, but that's not an issue here.
Copyright in fictional characters has been tried, occasionally with success. That's one of the broadest forms of copyright protection. Trademarking the name of the character provides more protection, which is why we see Darth Vader(tm) markings. That doesn't apply to an API.
So, why are both ICMP and IP considered to be in layer 3? ICMP is built on top of IP.
The real answer to that is that it's a Berkeley UNIXism. Some early TCP/IP implementations, including the one I worked on, had ICMP at a layer above IP, in the same layer with TCP and UDP. The Berkeley UNIX kernel, like other UNIX versions of the period, had real trouble communicating upward within the kernel, because this was before threads, let alone kernel threads.
To get around that kernel limitation, ICMP was crammed in with IP. This had some downsides, including the demise of ICMP Source Quench for congestion control, which didn't fit well into the mode of ICMP as an error-reporting mechanism for IP.
If you have a giant build, your design is not modular enough. Above some size, it's time to go to multiple intercommunicating programs.