Frankly, the same people who think you need a professional security assessment to upgrade from RedHat 7.2, and who insist on transferring confidential files via FTP. I'm not kidding: this came up with a remote business partner last week.
It's called AFS, and has quite good open implementations and Kerberos integration. Getting it installed by default on Windows clients is a huge burden, however.
Excuse me? Iraq is already in a state of civil war. Thousands of civilians being slaughtered every year, armed rebellion in the streets, the official military can't even step outside of their controlled zones safely. I don't know what you think a civil war is, but this certainly counts.
For those of you who haven't watched Fox News, they don't present both sides. They present their side, and then they bring in some poor caricature of an expert from a different group to mock and make fun of and never let them finish a sentence, and always cut them off when they start citing facts. It's embarassing to watch.
Oh, given that they had already found two fake pipe bombs, which no one here on Slashdot seems to know about (http://news.bostonherald.com/localRegional/view.b g?articleid=180349), the paranoia of the Boston authorities seems more understandable. It's also a really good idea sometimes to give your staff practice treating a small emergency as a real one, so that when they have the real one they're familiar with the procedures.
I'd much rather the Boston police take it seriously, and be wrong, than take it casually and be wrong.
That $6K is an easy figure to reach. One child, with dental care, vaccines, and an occasional emergency room visit for stitches can do a few thousand dollars a year. So can adults with prostate exams, mammograms, eyeglasses, and other basic medical costs. Then take one out of 10 people, and give them something chronic: diabetes, congestive heart failure,or the simple needs of old age. Then take one in 20 and give them something drastically expensive: child birth with complications, kidney failure, or a traffic accident. A single such event can easily cost $20,000. Then add in one in 20 in a nursing home, or needing other chronic hands-on medical support: those easily cost $100/day, or $30,000/year.
That's already in progress. Take a good look at the "Trusted Computing" features, and plans to use it to authenticate hardware such as hard drives and DVD drives to prevent their use by any software, OS, or tools that have not paid the money for the right software keys.
Why not? It worked for Robert Morris, who is now a computer science professor at MIT after writing the most destructive worm in UNIX history. Of course, Robert's father was head of the NSA, which helps you get a "stay out of jail free" card when you go to court. Look for details at http://en.wikipedia.org/wiki/Robert_Tappan_Morris.
And it's all open source, including the BIOS. The LinuxBIOS and OpenBIOS are very, very useful tools that easily outperform the commercial versions: children's laptops booting up in seconds are going to make the owners of more expensive, "feature-laden" laptops quite jealous.
OLPC is a very useful tool to education: being able to Google or Wikipedia for farming information, getting legal information and market information for poor farmers threatened by their landlords or lied to about crop prices, and simply getting detailed weather information locally are all amazingly useful. OLPC is about communications as much as any other grand purpose. And being able to shop around for better selling prices for their goods, or buying prices for food, fuel, and fertilizer may easily pay for the laptop within a year for a poor family.
And look for exactly such Microsoft patent or copyright based extensions for mail handlers to support Microsoft's SenderID product, and other XML based tools for which Microsoft has been grabbing patents any way it can for the last few years. And oh, my goodness, look for it with Trusted Computing based technologies, and with related DRM tools.
No, GPL has little to do with patents. Patents are a very separate set of intellectual property law, and are not universal. GPL is a copyright agreement: as such, if Novell "enhances" their tools with Microsoft tools or under a Microsoft license agreement is not allowed to publish its SuSE modifications to GPL software, it can't publish the GPL software.
That's very serious, and is likely to happen with hardware drivers, XML, file-sharing tools, and cross-platform software tools based on Microsoft patents.
Also note: Samba is very much a replacement for the older Novell Netware in its basic features. Novell's support for Samba just fell apart, because Jermey Allison (one of the key Samba developers) just resigned from Novell because he thinks this agreement violates the GPL. His resignation letter has been published, and is viewable at http://www.groklaw.net/article.php?story=200612210 81000710 .
Samba is pretty obviously one of the things Microsoft wanted to manipulate through this deal. Jermey Allison just took away one of their big hooks to manipulate the protocol and the standard implementations, at serious personal cost, by leaving Novell. Google grabbed him: expect to see some fascinating shared and large-scale storage work going on at Google.
I'm afraid it's not that easy. I've run into no less than 5 Windows apps, all of them touted by their creators as being available to Linux because you could run them on Citrix. All of them worked extremely poorly: poor available resolution, random parts of their displays showing up the wrong size due to poor font handling, random buttons not being availablee, failures to store configurations between active sessions, and other problems that did not happen with the applications installed locally on Windows.
They also did not display properly on Windows clients of the Citrix servers, so it wasn't the Linux client. All of these applications were themselves "steaming piles of shit", and unfortunately were lauded as "enterprise solutions". 2 of the companies using them went under financially: 2 of them wound up switching to open source products for stability and usability without buying very expensive individual client licenses for the software. The last is in the process of switching away, but hasn't chosen the replacement yet.
Also note that VNC worked imperfectly, but considerably better than Citrix at displaying the software. The need for one server per user, and the user access and authentication model for VNC were deemed inappropriate, so it was rejected.
This is especially true because the article ignored the basic definition of "spam". Many people think it is fraud or ripoff email, which is the largest and most irritating form of spam, and which causes "legitimate" advertisers to whine when they get marked spam.
But the more basic definition of spam is "UBC": Unsolicited Bulk Communications. If it's being sent out to a lot of people who didn't ask for it, it's spam. Those guidelines in the article are ways to protect so-called legitimate spam from being identified as fraudulent spam: they don't actually address the issue of spam itself. Because of this kind of failure, a failure encouraged by advertisers to protect their businesses, laws and policies are carefully written to protect them and not to simply block all spam. The result is foolishness like the US "CAN-SPAM" act.
There are other definitions of spam, but the UBC definition is clear and allows filtering or blocking of spam without getting involved in judging its content: it's much more workable.
How are you going to set your first pass at filters? Do you have to wait and collect several thousand spam messages yourself? And some of the spam is much less likely than others, either because your address is not yet known to that community of spammers or because it's less frequent. Why not filter against all the available types, rather than relying on your own efforts to calibrate or guild a new feature when you first get it?
This is especially true if you want to set up sitewide filters, rather than tuning it for individuals.
That is precisely what makes such a site useful: the variety of spam. If you've only tended to get Viagra spam, your filters will not be trained to detect phishing spam, or multi-level-marketing spam, or penny stock spam. A large and varied body of spam to test with is thus very useful, just as the world's most powerful immune system is often useless against a germ it's never been vaccinated for.
It's also useful to compare and train sitewide filters against a large body of spam, to make sure your filters are complete and aren't tuned to block legitimate mail as well. It's a very useful resource.
Try a variant on the third: the third supermarket can't get a license. So pressure builds, and the supermarket owner has to rely on its other markets to make money, and the local shoppers demand the rules be relaxed to allow the company to exist, but it can cost a year or two of lost business before the community standards relax and the company can open, and the store may never open there. But people will sneak over to the nearest town to do their shopping.
This happens all the time with drug stores that offer prophylactics, liquor stores, restaurants that want to open at late hours, and big companies like McDonald's and Wal-Mart that may have policies or products that are against the local standards in insular communities.
It's very often not the IT personnel blocking innovation: many of the IT people I know are fonts of innovation, and of the streamlining that helps reduce company costs in their love of making systems work. Rather, they face a middle management hurdle of immense difficulty. Having to spend 3 hours making a powerpoint slide because that's the approved format, doing it all over to add company logos, spending 3 distinct meetings with 3 managers at each, and having to present a complete cost/benefit analysis before being allowed to ask the questions of the other department's developers necessary to see if something can be done is a waste of time and money that can sap innovation dry.
And that was how I spent last week. My real work was done on my own time when I got frustrated about not being able to do what I was hired for.
This is hardly a Google problem. It's ICANN, and specifically Verisign for selling huge numbers of domains so cheaply, and making it so very difficult for a legitimate user to protect themselves or recover from a domain squatter. They also deliberately leave it very awkward to track the domains back to the squatters, making blacklists or automatic filters quite difficult to construct or keep up to date.
Draftiing the law is not the problem: Senate Bill 12 from Senator Bowen is not ideal, but at least attempts to tackle the issue by outlawing unsolicited commercial email and allowing consumers to sue for violations.
The existing legislation is targeted at spam the same way existing legislature is aimed at fraud, not spam itself. Most spam is fraudulent, true, because the penalties are so awkward to pursue. But lowering the threshold to prosecute by simply outlawing spam itself, rather than trying to establish fraud and relying on federal prosecutors instead of civil suits against the spammers and their ISP's is a big failure to address the problem.
No. It's really not. It's bulk, unsolicited communications. Closing the legal doors for fraudulent, or for commercial speech, leaves the door wide open for email worms, political speech, religious diatribes, and plain old irritating people by filling their mail boxes. It also leaves the door open for borderline uses, such as charitable solicitations and political advertising.
Classifying spam as commercial is also exactly why it's so hard to get good laws passed against it: trying to define what speech is commercial and what is not, in court or in law, is painful and confusing work. It's vastly easier to classify as bulk and unsolicited, which is easily measurable in a way that "commercial" speech will not be.
Usenet went through exactly those sorts of arguments, and settled correctly on not trying to interpret it as commercial or not, but simply blocking the bulk postings.
There are a lot of reasons. ISP's are scared of losiing their common carrier status, so they're very reluctant to set their own rules. They're also confused about spam: many people think of spam as only the fraudulent or sex email, so they try to make laws against only that. And it then gets challenged in court, or lobbied into tortured shapes that don't actually bind hte spammer's wrists.
Legitimate or semi-legitimate businesses also send bulk mail, often but not always spam, and lobby very, very heavily to protect it their ability to advertise this way. And there is an excuse for legislators to be cautious because they're concerned about interfering with anyone's free speech. Good state laws were also, very unforttunately, trumped by the CAN-SPAM act which frankly is designed to allow spam and only block certain types of fraud, and took away the ability for private citizens to do anything about it in court.
There have been effective laws agaiinst junk fax in the US and other nations, laws that have withstood constitutional challenges and really helped curb the problem. But until the advertising lobbyists are forced out of the way, there will be no good federal policies on this. These laws passed constitutional muster because they didn't refer to the content: they referred to the fact that the junk fax was unsolicited. Some US states have successfully passed and used such laws, but the CAN-SPAM act simply overrode them when it was passed.
Oh, the plans to manage content seem pretty obvious: control of formats, control of software, and the ability to block access to the hardware for non-MS-controlled software are all parts of Trusted Computing's capabilities. It's just not hte classic censorship that is easily detected as infringing on civil liberties and would gain quick challenges in court. Microsoft is engaging in a long campaign to get as much control as they can for obvious corporate reasons.
Slashdot Mirror
Frankly, the same people who think you need a professional security assessment to upgrade from RedHat 7.2, and who insist on transferring confidential files via FTP. I'm not kidding: this came up with a remote business partner last week.
It's called AFS, and has quite good open implementations and Kerberos integration. Getting it installed by default on Windows clients is a huge burden, however.
Excuse me? Iraq is already in a state of civil war. Thousands of civilians being slaughtered every year, armed rebellion in the streets, the official military can't even step outside of their controlled zones safely. I don't know what you think a civil war is, but this certainly counts.
For those of you who haven't watched Fox News, they don't present both sides. They present their side, and then they bring in some poor caricature of an expert from a different group to mock and make fun of and never let them finish a sentence, and always cut them off when they start citing facts. It's embarassing to watch.
Oh, given that they had already found two fake pipe bombs, which no one here on Slashdot seems to know about (http://news.bostonherald.com/localRegional/view.b g?articleid=180349), the paranoia of the Boston authorities seems more understandable. It's also a really good idea sometimes to give your staff practice treating a small emergency as a real one, so that when they have the real one they're familiar with the procedures.
I'd much rather the Boston police take it seriously, and be wrong, than take it casually and be wrong.
That $6K is an easy figure to reach. One child, with dental care, vaccines, and an occasional emergency room visit for stitches can do a few thousand dollars a year. So can adults with prostate exams, mammograms, eyeglasses, and other basic medical costs. Then take one out of 10 people, and give them something chronic: diabetes, congestive heart failure,or the simple needs of old age. Then take one in 20 and give them something drastically expensive: child birth with complications, kidney failure, or a traffic accident. A single such event can easily cost $20,000. Then add in one in 20 in a nursing home, or needing other chronic hands-on medical support: those easily cost $100/day, or $30,000/year.
It adds up pretty fast.
That's already in progress. Take a good look at the "Trusted Computing" features, and plans to use it to authenticate hardware such as hard drives and DVD drives to prevent their use by any software, OS, or tools that have not paid the money for the right software keys.
Why not? It worked for Robert Morris, who is now a computer science professor at MIT after writing the most destructive worm in UNIX history. Of course, Robert's father was head of the NSA, which helps you get a "stay out of jail free" card when you go to court. Look for details at http://en.wikipedia.org/wiki/Robert_Tappan_Morris.
And it's all open source, including the BIOS. The LinuxBIOS and OpenBIOS are very, very useful tools that easily outperform the commercial versions: children's laptops booting up in seconds are going to make the owners of more expensive, "feature-laden" laptops quite jealous.
OLPC is a very useful tool to education: being able to Google or Wikipedia for farming information, getting legal information and market information for poor farmers threatened by their landlords or lied to about crop prices, and simply getting detailed weather information locally are all amazingly useful. OLPC is about communications as much as any other grand purpose. And being able to shop around for better selling prices for their goods, or buying prices for food, fuel, and fertilizer may easily pay for the laptop within a year for a poor family.
That would be too bazaar.
And look for exactly such Microsoft patent or copyright based extensions for mail handlers to support Microsoft's SenderID product, and other XML based tools for which Microsoft has been grabbing patents any way it can for the last few years. And oh, my goodness, look for it with Trusted Computing based technologies, and with related DRM tools.
No, GPL has little to do with patents. Patents are a very separate set of intellectual property law, and are not universal. GPL is a copyright agreement: as such, if Novell "enhances" their tools with Microsoft tools or under a Microsoft license agreement is not allowed to publish its SuSE modifications to GPL software, it can't publish the GPL software.
That's very serious, and is likely to happen with hardware drivers, XML, file-sharing tools, and cross-platform software tools based on Microsoft patents.
Also note: Samba is very much a replacement for the older Novell Netware in its basic features. Novell's support for Samba just fell apart, because Jermey Allison (one of the key Samba developers) just resigned from Novell because he thinks this agreement violates the GPL. His resignation letter has been published, and is viewable at http://www.groklaw.net/article.php?story=200612210 81000710 .
Samba is pretty obviously one of the things Microsoft wanted to manipulate through this deal. Jermey Allison just took away one of their big hooks to manipulate the protocol and the standard implementations, at serious personal cost, by leaving Novell. Google grabbed him: expect to see some fascinating shared and large-scale storage work going on at Google.
I'm afraid it's not that easy. I've run into no less than 5 Windows apps, all of them touted by their creators as being available to Linux because you could run them on Citrix. All of them worked extremely poorly: poor available resolution, random parts of their displays showing up the wrong size due to poor font handling, random buttons not being availablee, failures to store configurations between active sessions, and other problems that did not happen with the applications installed locally on Windows.
They also did not display properly on Windows clients of the Citrix servers, so it wasn't the Linux client. All of these applications were themselves "steaming piles of shit", and unfortunately were lauded as "enterprise solutions". 2 of the companies using them went under financially: 2 of them wound up switching to open source products for stability and usability without buying very expensive individual client licenses for the software. The last is in the process of switching away, but hasn't chosen the replacement yet.
Also note that VNC worked imperfectly, but considerably better than Citrix at displaying the software. The need for one server per user, and the user access and authentication model for VNC were deemed inappropriate, so it was rejected.
This is especially true because the article ignored the basic definition of "spam". Many people think it is fraud or ripoff email, which is the largest and most irritating form of spam, and which causes "legitimate" advertisers to whine when they get marked spam.
But the more basic definition of spam is "UBC": Unsolicited Bulk Communications. If it's being sent out to a lot of people who didn't ask for it, it's spam. Those guidelines in the article are ways to protect so-called legitimate spam from being identified as fraudulent spam: they don't actually address the issue of spam itself. Because of this kind of failure, a failure encouraged by advertisers to protect their businesses, laws and policies are carefully written to protect them and not to simply block all spam. The result is foolishness like the US "CAN-SPAM" act.
There are other definitions of spam, but the UBC definition is clear and allows filtering or blocking of spam without getting involved in judging its content: it's much more workable.
How are you going to set your first pass at filters? Do you have to wait and collect several thousand spam messages yourself? And some of the spam is much less likely than others, either because your address is not yet known to that community of spammers or because it's less frequent. Why not filter against all the available types, rather than relying on your own efforts to calibrate or guild a new feature when you first get it?
This is especially true if you want to set up sitewide filters, rather than tuning it for individuals.
That is precisely what makes such a site useful: the variety of spam. If you've only tended to get Viagra spam, your filters will not be trained to detect phishing spam, or multi-level-marketing spam, or penny stock spam. A large and varied body of spam to test with is thus very useful, just as the world's most powerful immune system is often useless against a germ it's never been vaccinated for.
It's also useful to compare and train sitewide filters against a large body of spam, to make sure your filters are complete and aren't tuned to block legitimate mail as well. It's a very useful resource.
Try a variant on the third: the third supermarket can't get a license. So pressure builds, and the supermarket owner has to rely on its other markets to make money, and the local shoppers demand the rules be relaxed to allow the company to exist, but it can cost a year or two of lost business before the community standards relax and the company can open, and the store may never open there. But people will sneak over to the nearest town to do their shopping.
This happens all the time with drug stores that offer prophylactics, liquor stores, restaurants that want to open at late hours, and big companies like McDonald's and Wal-Mart that may have policies or products that are against the local standards in insular communities.
I take it you've never bought stock or insurance?
It's very often not the IT personnel blocking innovation: many of the IT people I know are fonts of innovation, and of the streamlining that helps reduce company costs in their love of making systems work. Rather, they face a middle management hurdle of immense difficulty. Having to spend 3 hours making a powerpoint slide because that's the approved format, doing it all over to add company logos, spending 3 distinct meetings with 3 managers at each, and having to present a complete cost/benefit analysis before being allowed to ask the questions of the other department's developers necessary to see if something can be done is a waste of time and money that can sap innovation dry.
And that was how I spent last week. My real work was done on my own time when I got frustrated about not being able to do what I was hired for.
This is hardly a Google problem. It's ICANN, and specifically Verisign for selling huge numbers of domains so cheaply, and making it so very difficult for a legitimate user to protect themselves or recover from a domain squatter. They also deliberately leave it very awkward to track the domains back to the squatters, making blacklists or automatic filters quite difficult to construct or keep up to date.
Draftiing the law is not the problem: Senate Bill 12 from Senator Bowen is not ideal, but at least attempts to tackle the issue by outlawing unsolicited commercial email and allowing consumers to sue for violations.
The existing legislation is targeted at spam the same way existing legislature is aimed at fraud, not spam itself. Most spam is fraudulent, true, because the penalties are so awkward to pursue. But lowering the threshold to prosecute by simply outlawing spam itself, rather than trying to establish fraud and relying on federal prosecutors instead of civil suits against the spammers and their ISP's is a big failure to address the problem.
No. It's really not. It's bulk, unsolicited communications. Closing the legal doors for fraudulent, or for commercial speech, leaves the door wide open for email worms, political speech, religious diatribes, and plain old irritating people by filling their mail boxes. It also leaves the door open for borderline uses, such as charitable solicitations and political advertising.
Classifying spam as commercial is also exactly why it's so hard to get good laws passed against it: trying to define what speech is commercial and what is not, in court or in law, is painful and confusing work. It's vastly easier to classify as bulk and unsolicited, which is easily measurable in a way that "commercial" speech will not be.
Usenet went through exactly those sorts of arguments, and settled correctly on not trying to interpret it as commercial or not, but simply blocking the bulk postings.
There are a lot of reasons. ISP's are scared of losiing their common carrier status, so they're very reluctant to set their own rules. They're also confused about spam: many people think of spam as only the fraudulent or sex email, so they try to make laws against only that. And it then gets challenged in court, or lobbied into tortured shapes that don't actually bind hte spammer's wrists.
Legitimate or semi-legitimate businesses also send bulk mail, often but not always spam, and lobby very, very heavily to protect it their ability to advertise this way. And there is an excuse for legislators to be cautious because they're concerned about interfering with anyone's free speech. Good state laws were also, very unforttunately, trumped by the CAN-SPAM act which frankly is designed to allow spam and only block certain types of fraud, and took away the ability for private citizens to do anything about it in court.
There have been effective laws agaiinst junk fax in the US and other nations, laws that have withstood constitutional challenges and really helped curb the problem. But until the advertising lobbyists are forced out of the way, there will be no good federal policies on this. These laws passed constitutional muster because they didn't refer to the content: they referred to the fact that the junk fax was unsolicited. Some US states have successfully passed and used such laws, but the CAN-SPAM act simply overrode them when it was passed.
Oh, the plans to manage content seem pretty obvious: control of formats, control of software, and the ability to block access to the hardware for non-MS-controlled software are all parts of Trusted Computing's capabilities. It's just not hte classic censorship that is easily detected as infringing on civil liberties and would gain quick challenges in court. Microsoft is engaging in a long campaign to get as much control as they can for obvious corporate reasons.