Unix, when it was new, was radical in that everything was in ordinary ascii text files. Everyone else "knew" that you had to work in binary, have binary config databases, binary file systems with dozens of record type and so on. With each binary format you had to provide a binary editor and/or debugger. If something broke, you needed a high priest of that particular religion just to debug it, much less fix it.
Note how many Unixes you see for each machine running GCOS or PRIMOS. Of all the machines of the day that still exist, note that most z/OS files are simple EBCDIC. Over time, that square wheel quietly went away.
When the PC came along, application designers once again started doing everything in binary, plus the occasional DOS text file. If something broke, you needed to go back to the vendor, because programs didn't come with binary editors. Or you could get a high priest of a particular order to take it apart in a debugger.
And, just to add injury to insult, a 64-bit binary floating point zero is four times the length of an ascii zero followed by a space or newline. Spreadsheet files in binary were ~ 4 times larger that the ones in DOS text my company used (;-)) Turns out spreadsheet files are mostly empty, or mostly contain zeros.
Over time, lots of config files and data files became ascii or UTF-8, and a huge number fo data files became html or xml text files. And that square wheel went away.
Let a hypertext file be a sequence of bytes, separated by newline characters. Let the text be a sequence of bytes, optionally using multiple bytes per character, as in UTF-8.
Verily I say unto you, let it be simple, lest it be stupid. Round wheels are better than square, or ones that are just flat on one side
The U.S. military addressed all the problems except covert channels (now called DLP) in the Orange Book, back in 1985, the days of the dinosaurs and mainframes.
Alas, it was relatively hard to admin, requiring two people to do almost anything, and proving the completeness and sufficiency of the policies was exceedingly hard using the techniques of the day. The good thing was it was easy to use such a system. I used Multics, which was running at B2 and didn't even know security was tight. I later took the week course on how to admin Trusted Solaris and admined a couple of B1 machines. My brain tended to bleed out my ears, I kept running out of audit disk until I turned audit down to a week and I badly broke the two-person rule.
I suspect the difficulty and cost of running secure systems, and the cost of having two-person signoffs in computing as well as accounting killed the governments' desire to be reasonably secure against insiders.
The mechanisms to implement MAC and much of the rest still exist in the NSA security-enhanced Linux, but the work of creating categories and levels to keep users out of each other's pockets, and managing them and the sysadmins so they can't conspire to sneak data out is too expensive for any organization to shoulder as a cost, even the NSA.
Your cheque is an order to the bank to pay $X to Y. If your government outlaws Y, the bank cannot honour your order.
At the moment, banks have a smallish list of countries and companies that have been outlawed, and so the bank cannot pay tme anything. These are organizations/countries claimed to be in of support of terrorism. If the government in question can argue VPNs enable terrorism, they can add VPN companies to the list.
... you are, when, who your friends are, where they are, who their friends are and so on. In fact, it's everything about you except what you said to Aunt Martha in the letter.
To be fair, it is wonderful for tracking spies. If you start with one known spy, it helps find others. See Using Metadata to Find Paul Revere", written from the point of view of the British.
It's less useful for identifying spies (or bombers) from a cloud of data, because if you start with Aaron Swartz, you do get a list of supposedly suspicious people to investigate. Unfortunately, for the spy-trackers, they're actually innocent bystanders.
--dave
[You can always charge them and see if they suicide. Just like the old trick of throwing a witch in a pond and seeing if they float. If they do, they're evil and you kill them. If they don't, they drown. Either way, you get rid of them.]
Just like a swimming pool, keeping records that someone else might want is an attractive nuisance: people you don't want will go snooping around in them. And just like a swimming pool, it you that's liable when someone uses them without your permission.
At the moment, it's ISPs that find themselves having to cough up DHCP records to courts: give the criminals a week or two and they'll be writing exploits to get at Facebook, Google+ and your local video store, just like they've been doing for people who have lists of credit-card numbers.
If the citizen was making false complaints to Homeland Security, that would be actionable, but this is TDEC complaining to HS that the citizen is a terrorist. IMHO, this would render the TDEC official the person wasting Staatssicherheit's time with a false complaint. If that's terrorism, then Mr. Smith is encouraging and/or procuring terrorism by having his staff make the reports.
It also suffers from the birthday paradox: if you're looking for a lot of people (ie, N terrorists) in a large database (ie, M drivers licences), the probability of a false positive is multiplied by a factor of roughly N factorial, where N is the number of terrorists you're looking for. This caused the German Federal Security folks and Siemens to cancel an effort to use facial recognition in airports.
The law is predicated on the results being desirable: monopolies are illegal unless they fit the terms of the (U.S.) copyright act, authorized by the constitution. If the result is a catastrophy, the law is unconstitutional.
When you get past modus ponens into universal and existential quantifiers.
Or, as the medievals would say, when "one crosses the bridge of fools" from the trivia to the quadrivia (:-))
You end up needing it later, anyway... I struggled with logic initially, and wondered if I'd ever need it. I ended doing a ton of it in ADL just to test some libraries! And as a performance engineer, I rapidly discovered that DBAs don't know deMorgan's law, hoping that the query optimized does (;-))
Definitely start small, but make sure you measure the response time and load from the beginning, as close to the user as possible.
The load will tell you how many users you're gaining, albeit in computer terms, and the response time will tell you if the system is starting to annoy people by slowing down. If you plot RT against load, you'll get the curve you need for capacity planning when and if the program becomes popular.
report
The U.S. requests under our Mutual Law Assistance Treaties for private information re Megaupload parallels the CISPA proposals, and both strike me as wrong-headed (;-)) It is arguably valid for such a process to be followed in cases of copyright infringement, and can be critiqued on the basis of whether it is necessary and sufficient.
However, it suggest that at least the U.S. government is trying to deal with a minor crime, copyright infringement, because they don't know how to deal with major ongoing ones, commercial espionage.
Real "computer crime"is centred around breaking in to people's machines to steal data or crash them to deny the data to its owners. This is done via viruses, root-kits and the like, communicating across the internet to "bot-nets", collections of machines used as accomplices and cut-outs. These in turn are run by "bot master" machines in the hands of the criminals.
To investigate a key-logger (snooping) virus running on the machine of your chief counsel, you need to trace the connections across the internet from the infected machine to the "bot" and thence to the master. This requires cooperation of the police in the jurisdictions where the machines are and the ISPs they are connected to, to trace the connections between machines. To the best of my knowledge, that is barely in discussion at ICANN, and is nowhere part of the law or practice.
Only once that is done does one need to identify persons, and only one person, the criminal operating the master, and seize the machine for evidence, possibly in a foreign country.
All the other human beings in the story are victims, whom we do not need to identify, but merely transmit a warning to via their ISP. Once we have seized the master machine, we know the IP addresses (and ISPs) of the people who are being attacked, and the IP addresses of the people whose machines have been taken over by viruses to become the bot-net. Without breaching confidentiality, an ISP can forward a message that they are infected by a criminal's virus, and in extreme cases require the machine to be cleaned of infectious before being allowed to connect to the ISPs other customers.
I'm just a bit horrified at our American cousins: right now, people are stealing corporate information, collecting credit-card numbers and sabotaging centrifuges using techniques that neither the police, legislators nor courts are paying any attention to. Instead they are prosecuting a drop-box operator for a misdemeanor.
They remind me of the story of the drunk looking for his car-keys under the street-light, instead of in the dark garage where he dropped them.
Logically, a police force anywhere should need a wiretap warrant to read your (electronic) mail, and you have a duty to password-protect your email (:-)) At the moment this hasn't been tested in court, even in Canada.
--dave
The world of end-user programming is larger that one would think on first glance, although in the case of spreadsheets it looks like functional-languages-with-globals (;-))
I've seen occasional graphic languages (POLs) that could be used in more general ways than spreadsheets: one needs to find one that solves an interesting problem everyone faces.
--dave
In addition, their actions have been found to be a fraud upon the courts in some cases.
Applying to the courts for an order to identify people on the grounds you will sue them, and then extorting payments instead, makes the initial application fraudulent.
In the U.S. and the U.K., this has led to legal or law-society actions against the fraudulent plaintiffs. In Canada, as we just passed a law to limit such suits, it may lead to stronger measures.
A legal system based on private actions (suits) always has that problem: one based on crimes being investigated by police doesn't (but has a quality control problem with the police: quid custoet custodes?).
Wordpress replied that "the blog owner has already informed us that the subpoena will be challenged. Per our policies, we will not turn over any information (including on commentors) until that challenge has been decided by the courts."
In the era in which the Constitution was written, this wasn't a common practice. A good federal SLAPP law might classify then actions as an abuse of power, or a fraud upon the court, depending on whether one was a lawyer or a plaintiff, in order to make the punishment fit the crime.
Anyone want to start a White House petition for criminal sanctions against anyone who defrauds a court?
--dave (I can't, I'd have to start a 24 Sussex Drive petition) c-b
This is Mr. Harper, whose Reform/Alliance/Conservative party supposedly stands for individual liberty over the oppression of the corporations, fairness to Canadians and loyalty to the principles of peace, order and good government.
Their actions are a betrayal of their own members, the legalization of the oppression of citizens by corporations and the subversion of Canadian judicial independence from the U.S.
I'm sure his antediluvian core supporters will think this is a good idea, but I suspect anyone under fifty will find it at least a little suspicious!
This is really rather important: More emphasis! More emphasis!
To people in Sweden, this seemed mind-bogglingly odd: in the small Scandinavian country, private entrepreneurs had been fibering apartment blocks wholesale for years. I had fiber in my own apartment in 1999, and keep enjoying a 100 megabit-connection with several static, public IPs – from where you’re reading this article, as I run my server from home.
The take-home from this is that telcos have a conflict of interest, while hydro companies have underused poles in your neighbourhood.
Slashdot Mirror
Unix, when it was new, was radical in that everything was in ordinary ascii text files. Everyone else "knew" that you had to work in binary, have binary config databases, binary file systems with dozens of record type and so on. With each binary format you had to provide a binary editor and/or debugger. If something broke, you needed a high priest of that particular religion just to debug it, much less fix it.
Note how many Unixes you see for each machine running GCOS or PRIMOS. Of all the machines of the day that still exist, note that most z/OS files are simple EBCDIC. Over time, that square wheel quietly went away.
When the PC came along, application designers once again started doing everything in binary, plus the occasional DOS text file. If something broke, you needed to go back to the vendor, because programs didn't come with binary editors. Or you could get a high priest of a particular order to take it apart in a debugger.
And, just to add injury to insult, a 64-bit binary floating point zero is four times the length of an ascii zero followed by a space or newline. Spreadsheet files in binary were ~ 4 times larger that the ones in DOS text my company used (;-)) Turns out spreadsheet files are mostly empty, or mostly contain zeros.
Over time, lots of config files and data files became ascii or UTF-8, and a huge number fo data files became html or xml text files. And that square wheel went away.
Let a hypertext file be a sequence of bytes, separated by newline characters. Let the text be a sequence of bytes, optionally using multiple bytes per character, as in UTF-8.
Verily I say unto you, let it be simple, lest it be stupid. Round wheels are better than square, or ones that are just flat on one side
--dave
The U.S. military addressed all the problems except covert channels (now called DLP) in the Orange Book, back in 1985, the days of the dinosaurs and mainframes.
Alas, it was relatively hard to admin, requiring two people to do almost anything, and proving the completeness and sufficiency of the policies was exceedingly hard using the techniques of the day. The good thing was it was easy to use such a system. I used Multics, which was running at B2 and didn't even know security was tight. I later took the week course on how to admin Trusted Solaris and admined a couple of B1 machines. My brain tended to bleed out my ears, I kept running out of audit disk until I turned audit down to a week and I badly broke the two-person rule.
I suspect the difficulty and cost of running secure systems, and the cost of having two-person signoffs in computing as well as accounting killed the governments' desire to be reasonably secure against insiders.
The mechanisms to implement MAC and much of the rest still exist in the NSA security-enhanced Linux, but the work of creating categories and levels to keep users out of each other's pockets, and managing them and the sysadmins so they can't conspire to sneak data out is too expensive for any organization to shoulder as a cost, even the NSA.
--dave
Your cheque is an order to the bank to pay $X to Y. If your government outlaws Y, the bank cannot honour your order.
At the moment, banks have a smallish list of countries and companies that have been outlawed, and so the bank cannot pay tme anything. These are organizations/countries claimed to be in of support of terrorism. If the government in question can argue VPNs enable terrorism, they can add VPN companies to the list.
--dave
To be fair, it is wonderful for tracking spies. If you start with one known spy, it helps find others. See Using Metadata to Find Paul Revere", written from the point of view of the British.
It's less useful for identifying spies (or bombers) from a cloud of data, because if you start with Aaron Swartz, you do get a list of supposedly suspicious people to investigate. Unfortunately, for the spy-trackers, they're actually innocent bystanders.
--dave
[You can always charge them and see if they suicide. Just like the old trick of throwing a witch in a pond and seeing if they float. If they do, they're evil and you kill them. If they don't, they drown. Either way, you get rid of them.]
Just like a swimming pool, keeping records that someone else might want is an attractive nuisance: people you don't want will go snooping around in them. And just like a swimming pool, it you that's liable when someone uses them without your permission.
At the moment, it's ISPs that find themselves having to cough up DHCP records to courts: give the criminals a week or two and they'll be writing exploits to get at Facebook, Google+ and your local video store, just like they've been doing for people who have lists of credit-card numbers.
--dave
If the citizen was making false complaints to Homeland Security, that would be actionable, but this is TDEC complaining to HS that the citizen is a terrorist. IMHO, this would render the TDEC official the person wasting Staatssicherheit's time with a false complaint. If that's terrorism, then Mr. Smith is encouraging and/or procuring terrorism by having his staff make the reports.
It also suffers from the birthday paradox: if you're looking for a lot of people (ie, N terrorists) in a large database (ie, M drivers licences), the probability of a false positive is multiplied by a factor of roughly N factorial, where N is the number of terrorists you're looking for. This caused the German Federal Security folks and Siemens to cancel an effort to use facial recognition in airports.
The law is predicated on the results being desirable: monopolies are illegal unless they fit the terms of the (U.S.) copyright act, authorized by the constitution. If the result is a catastrophy, the law is unconstitutional.
That is what is important to the courts.
--dave
When you get past modus ponens into universal and existential quantifiers. Or, as the medievals would say, when "one crosses the bridge of fools" from the trivia to the quadrivia (:-))
--dave
You end up needing it later, anyway... I struggled with logic initially, and wondered if I'd ever need it. I ended doing a ton of it in ADL just to test some libraries! And as a performance engineer, I rapidly discovered that DBAs don't know deMorgan's law, hoping that the query optimized does (;-))
--dave
For some reason it got longer than what I had tried to post... --dave
Definitely start small, but make sure you measure the response time and load from the beginning, as close to the user as possible.
The load will tell you how many users you're gaining, albeit in computer terms, and the response time will tell you if the system is starting to annoy people by slowing down. If you plot RT against load, you'll get the curve you need for capacity planning when and if the program becomes popular. report
--dave
[In part from a reply to http://www.slaw.ca/2013/04/04/access-to-server-data-for-foreign-criminal-investigative-purposes/ at Slaw]
The U.S. requests under our Mutual Law Assistance Treaties for private information re Megaupload parallels the CISPA proposals, and both strike me as wrong-headed (;-)) It is arguably valid for such a process to be followed in cases of copyright infringement, and can be critiqued on the basis of whether it is necessary and sufficient.
However, it suggest that at least the U.S. government is trying to deal with a minor crime, copyright infringement, because they don't know how to deal with major ongoing ones, commercial espionage.
Real "computer crime"is centred around breaking in to people's machines to steal data or crash them to deny the data to its owners. This is done via viruses, root-kits and the like, communicating across the internet to "bot-nets", collections of machines used as accomplices and cut-outs. These in turn are run by "bot master" machines in the hands of the criminals.
To investigate a key-logger (snooping) virus running on the machine of your chief counsel, you need to trace the connections across the internet from the infected machine to the "bot" and thence to the master. This requires cooperation of the police in the jurisdictions where the machines are and the ISPs they are connected to, to trace the connections between machines. To the best of my knowledge, that is barely in discussion at ICANN, and is nowhere part of the law or practice.
Only once that is done does one need to identify persons, and only one person, the criminal operating the master, and seize the machine for evidence, possibly in a foreign country.
All the other human beings in the story are victims, whom we do not need to identify, but merely transmit a warning to via their ISP. Once we have seized the master machine, we know the IP addresses (and ISPs) of the people who are being attacked, and the IP addresses of the people whose machines have been taken over by viruses to become the bot-net. Without breaching confidentiality, an ISP can forward a message that they are infected by a criminal's virus, and in extreme cases require the machine to be cleaned of infectious before being allowed to connect to the ISPs other customers.
I'm just a bit horrified at our American cousins: right now, people are stealing corporate information, collecting credit-card numbers and sabotaging centrifuges using techniques that neither the police, legislators nor courts are paying any attention to. Instead they are prosecuting a drop-box operator for a misdemeanor.
They remind me of the story of the drunk looking for his car-keys under the street-light, instead of in the dark garage where he dropped them.
–dave
The proponents want you to think that: in fact, the non-"anal probe"* companies will object to this variant. Time for another "paint it black" day!
--dave
* Thanks to wierd_w for the term!
Canada now requires a wiretap warrant to ready stored texts on pohone-company servers, which is harder to get than a regular one. See http://www.cbc.ca/news/technology/story/2013/03/27/technology-telus-text-messages-scc-decision.html [www.cbc.ca]
The Ontario appeal court separately ruled that one needs to put a password/passcode on your phone to demonstrate that you have and expect privacy in the data it contains. See http://www.cbc.ca/news/canada/ottawa/story/2013/02/21/ottawa-cell-phone-users-beware.html [www.cbc.ca]
Logically, a police force anywhere should need a wiretap warrant to read your (electronic) mail, and you have a duty to password-protect your email (:-)) At the moment this hasn't been tested in court, even in Canada. --dave
Canada now requires a wiretap warrant, which is harder to get than a regular one. See http://www.cbc.ca/news/technology/story/2013/03/27/technology-telus-text-messages-scc-decision.html
In a separate decision, the Ontario appeal court ruled that one needs to put a password/passcode on your phone to demonstrate that you have and expect privacy in the data it contains. See http://www.cbc.ca/news/canada/ottawa/story/2013/02/21/ottawa-cell-phone-users-beware.html
--dave
The world of end-user programming is larger that one would think on first glance, although in the case of spreadsheets it looks like functional-languages-with-globals (;-))
I've seen occasional graphic languages (POLs) that could be used in more general ways than spreadsheets: one needs to find one that solves an interesting problem everyone faces.
--dave
In addition, their actions have been found to be a fraud upon the courts in some cases.
Applying to the courts for an order to identify people on the grounds you will sue them, and then extorting payments instead, makes the initial application fraudulent.
In the U.S. and the U.K., this has led to legal or law-society actions against the fraudulent plaintiffs. In Canada, as we just passed a law to limit such suits, it may lead to stronger measures.
--dave
--dave
Wordpress replied that "the blog owner has already informed us that the subpoena will be challenged. Per our policies, we will not turn over any information (including on commentors) until that challenge has been decided by the courts."
--dave c-b
--dave
In the era in which the Constitution was written, this wasn't a common practice. A good federal SLAPP law might classify then actions as an abuse of power, or a fraud upon the court, depending on whether one was a lawyer or a plaintiff, in order to make the punishment fit the crime.
Anyone want to start a White House petition for criminal sanctions against anyone who defrauds a court?
--dave (I can't, I'd have to start a 24 Sussex Drive petition) c-b
This is Mr. Harper, whose Reform/Alliance/Conservative party supposedly stands for individual liberty over the oppression of the corporations, fairness to Canadians and loyalty to the principles of peace, order and good government.
Their actions are a betrayal of their own members, the legalization of the oppression of citizens by corporations and the subversion of Canadian judicial independence from the U.S.
I'm sure his antediluvian core supporters will think this is a good idea, but I suspect anyone under fifty will find it at least a little suspicious!
This is really rather important: More emphasis! More emphasis!
--dave
In the article Trusting Telcos With Internet Is Like Trusting Fox With Henhouse, Rick writes
The take-home from this is that telcos have a conflict of interest, while hydro companies have underused poles in your neighbourhood.