I'm sorry to say that this is a typical systemd advocate answer to a much larger problem.
Having to reboot your operating system to enable text based logging for a specific service, or for all services is not reasonable. Hand-editing the boot modules to change the next reboot is _extremely_ dangerous manual editing capable of fracturing your system, and the boot console is not available on many virtualized or remotely managed systems nor without jumping through extraordinary hoops.
Root login access, or sudo access, does not mean a developer or programmer has console access. And even if console access is available, many server class systems take quite long periods to go through hardware scanning and present only a few seconds to manually modify the boot options, and some remote management systems that nominally provide remote console access take so long to restart or reconnect after a reboot that the necessary boot options have passed by before they could be selected.
On top of this, frequently rebooting Linux systems triggers the counters on the partitions that trigger an fsck at boot time after a specific number of reboots. An unscheduled fsck on a larger storage system can make the reboot take _hours_, and can also require console access to accept or reject the fsck options. This can cause a system without console access to fail to reboot, even if the boot loader has been manually loaded, and take the system online until manual keyboad and console access can be scheduled.
> That is a good thing to keep in mind, since nobody is running systemd except by choice.
I'm afraid that's nonsense. systemd has been a penalty in work I do. If I could have more contemporary versions of Linux based daemsns and packages without systemd, such as entire LAMP stacks, I'd accept them in a moment. Debugging failed daemon startups with systemd has repeatedly proven painful, due to the binary log formats and the difficulty of deducing the actual daemon startup commands to run them in a debugger.
> Because the whiners don't have a use-case. systemd is modular, but it tends to come with all the modules packaged together
I'm afraid it's not. The dependencies among components are very strong, and it's quite difficult to segregate out one component for de-activation or non-installation unless you compile with that feature de-activated, in which case you must recompile to re-enable the future. It's very difficult to install only the components you want due to the interdependencies.
A chance to hear formal speech, listen to debates, and pick up accents is an invaluable aid to immigration. Potential immigrants do study the language quite hard when they know it will affect their income and chances of immigration.
The Timothy Zahn books from the Star Wars universe have actually been quite good. They're gone further with the old characters, introduced fascinating new characters, and told much broader stories. I don't believe they're considered "Star Wars canon", but parts of them have certainly shown up in the games. The Thrawn trilogy, itself, included _wonderful_ exploration of other racess' personalities and how individuals are much more difficult to predict than large groups of their native races. It also explored why "clones" might be a limited or difficult resource to handle in a universe with "The Force".
It's very easy after a disaster to claim that an unfunded or ignored project would have prevented the disaster. Since the whistleblowers in the article are talking about the 9/11 terrorist attack, it seems a bit late. to be blowing whistles on it now.
It does seem clear that the NSA suffered, and is suffering, from Jerry Pournell's "Iron Law of Bureaucracy"
>> First, there will be those who are devoted to the goals of the organization.
>> Secondly, there will be those dedicated to the organization itself
The amount of money, time, and manpower burned on oversampling incredible amounts of personal traffic would seem much better focused on parts of the world, and populations, where the monitoring is likely to bear more fruit. But that doesn't expand the NSA itself and its overall capacity.
> The US Constitution is about a set of enumerated powers. I don't see any of the enumerated powers that would allow the US government to intervene in society to "reduce the risk of homicides"
Please re-read the "preamble" of the constitution. Quoted below::
“We the People of the United States, in Order to form a more perfect Union, establish Justice, insure domestic Tranquility, provide for the common defence, promote the general Welfare, and secure the Blessings of Liberty to ourselves and our Posterity, do ordain and establish this Constitution for the United States of America.”
I see at least three phrases that easily include "reducing the risk of homicides", especially the one about "insure domestic Tranquility".
> It is your error to interpret such a phrase as an authorization for the government to take away individual liberties of law abiding citizen in order to reduce perceived risks
It's part of how the laws get created. "Law abiding" takes on fascinating meaning when the laws infringe on those "individual liberties", and the grounds for the law are challenged. Whether you believe that individual liberty is the ultimate goal of government, the legislature, the courts, the government, and most of civilized history does not. It's one of _several_ goals of most governments, and a laudable one, but not an absolute goal.
> First, those two goals cannot be achieved simultaneously: you can't have a society that is governed rationally and that simultaneously is free. Second, even if rational government were desirable, it couldn't be realized in practice; the more rational you attempt to make government, the more corrupt it becomes.
Perfection isn't possible. _Anything_ taken to excess becomes corrupt. It's why governments change and respond to the people who make them up, whether peacefully or on occasion violently. And it's why they attempt to balance the rights of individuals against the safety of others: it's historically one of the clearest reasons for law and governmental intervention. Gun laws, zoning laws for hazardous businesses, fences around railroads, and eminent domain used to confiscate land in flood zones and erect dams are all classic examples of government doing exactly that.
Look again, please. systemd breaks stable network configurations by unnecessarily replacing dhcp, it breaks daemon-startup debugging, it breaks decades of log analysis tools designed to work with text based rather than proprietezed binary logging format, it's repeatedly broken kernel startups, it's broken the stable model of attached storage being mounted under/media, and the attempts to replace all of "/etc" with a "stateless Linux" model is breaking tools that never volunteered to have anything to do with systemd. It's also breaking cross-platform compatibility of daemon initialization configurations.
A "light Linux user" may not see these issues becuase you wouldn't necessarily be debugging failed daemons, writing cross-platform tools, or trying to integrate stable business software with this latest fad for configurations.
This is why you don't trust the mixed "cable modem" devices as anything but a cable modem. Many of them also include firewall, DHCP, and wifi features. Unfortunately, the extra "features" help make them more vulnerable to this kind of remote maintenance access password abuse.
>> At any cost, no. But reducing homicides is _definitely_ one of the roles of government.
> I must have missed that part of the US Constitution.
The constitution is about means, not purposes. "Life, liberty, and the pursuit of happiness" are more clearly defined as goals in the Declaration of Independence, which has fascinating notes about what a government is supposed to do, not merely about how to do it. The famous Second Amendment was created to prevent that goal from preventing the existence of local militias.
> Making policy based on risk estimates (or probabilities in general) is inherently unjust and is not compatible with a free society.
Oh, my. I can't possibly sum up the irrationality of your analyses any better than this.
The CIA got pushed, hard, to support excuses to invade Iraq, and was under tremendous pressure from the White House to support a war. Examine the recently released interviews with former CIA analyst Ben Bonk, at http://www.newsweek.com/2015/0...
Afghanistan is considered part of the Middle East by many, and has strong language and cultural ties with the rest of the Muslim cultures of the Middle East. It's a key land route between the Middle East and Asia, which is one reason it's been so valuable and so often invaded by other nations.
And yes, Russia's invasion was marked by disaster, as was the more recent US invasion, and the previous British invasion.
> Of course, none of this even amounts to a political argument.
Of course it does. Even very strange concepts are the source of political arguments.
> That is, reducing homicide rates at any cost should not be the goal of government.
At any cost, no. But reducing homicides is _definitely_ one of the roles of government. Failure to do so is a reason why governments fall.
> Politically, whether you decide to off your neighbor with your gun has no bearing on my interest or right to own a gun.
And of course it does. The _risks_ of gun ownership are why societies make laws about it. What those laws should be are a matter of considerable political argument.
http://ajph.aphapublications.o..., an American Journal of Public Health study, which finds a very direct correlation, and stacks of similar studies in their references.. The state by state results can be obscured by many factors, such as the age or income of the state populations, so it can be difficult to ensure that homicide rates can be strongly linked to homicide. But they seem to have done their homework. And the homicide rates in Europe and other gun controlled parts of the world are notably lower than the US, with reasonably good correlation with levels of gun control.
Part of the key to understanding the gun control claims and "refutations" is to watch the language very carefully. The claim that "states with the tightest gun control laws have the highest crime rates" tends to reverse cause and effect. The higher crime rates happen from poverty, racism, gangs, and other social problems, and the tighter gun control goes hand in hand with trying to _control_ those. So the statistics can be quite confusing unless one spends time correlating the other factors.
> As to things like the Paris incident, it occurs to me to wonder how easy it would have been to wander through a crowded venue shooting people at random if some of those people had been armed themselves...
The homicide rate with guns in the USA shows the difficulty. The idea that an "armed society is a politie society" was explored by Robert Heinlein in a number of his stories. In real life, the frequency of domestic violence and of violent neighborhood brawls remains so high that the deaths from household violence far outnumber those which might be saved by making personal firearms widely available.
Your other suggestions have similar difficulties. "Raising everyone's standard of living" requires steps, like educating women and reducing the gross disparities between ruling elites and grinding poverty, and birth control to prevent simply creating more starving poor, with profound social and religious consequences that are often resisted by those currently in power.
> Because someone can point a very portable RFID reader at a pile of tagged notes and get all the serial numbers, number of notes of each type and total value of the notes in a few second
Not without raising the price of the bills quite a lot. Reading a few distinct RFID tags in a small box is one task, reading 100 distinct RFID's out of a stacked box of paper is still well beyond the limit of any RFID technology. And they can be magnetically obliterated trivially. (nd get all the serial numbers, number of notes of each type and total value of the notes in a few second)
Putting an RFID reader where a money sorter passes individual bills might be effective in reading the bills casually, but RFID tags are still fragile. They don't take bending or folding well, and a great deal of money gets folded a great deal in normal handling. RFID tags also remain expensive: Adding $0.25 to the cost of making each bill would upset the US or EU mints quite a lot.
> The problem with mod_perl was one of perception:
The problem with the mod_perl update was one of naming and numbering. The version that supported httpd, sometimes called "Apache 2", was _among_ the mod_perl 1.9xxx releases, and it was very difficult to tell which "1.9xxx" release was compatible with what. Older perl modules that were only compatible with the older releases had to all be updated to require a release _less than_ 1.9xxx, or they would break when installed with "cpan install". And if your software required 2 modules, one of them with a recent release compatible only with the old mod_perl, and the other compatible only with the newer mod_perl, you had to trace back each individual module to find a version compatible with the dependencies of the other. This was hellish for web developers, and it _kept happening_ anytime anyone introduced or updated anything that used mod_perl.
The version compatible with Apache 1.3 should have retained the name "mod_perl", and it could have been supported until Debian finally stopped publishing Apache 1.3 packages alongside HTTPD 2.x packages. The new mod_perl could, and should, have been named mod_perl2 to reflect its incompatibility with the old codebase. This would have protected old modules compatible only with the older codebases and preserved forward compatibility. Newer packages could, and would, have used the new codebase by specifically calling it, without having to deduce the specific mod_perl 1.9xxx release where compatibility broke.
You're quite correct, such an internship is abusive. Someone interning at a "branding company" and being abused seems to be exactly the sort of life lesson that would be better learned in an internship than as a college graduate. But the experience for some students of "work free for a semester and get no classes, no income for it, and still pay tuition" sems to be an abuse by both the company providing the internship, and the school.
Slashdot Mirror
I'm sorry to say that this is a typical systemd advocate answer to a much larger problem.
Having to reboot your operating system to enable text based logging for a specific service, or for all services is not reasonable. Hand-editing the boot modules to change the next reboot is _extremely_ dangerous manual editing capable of fracturing your system, and the boot console is not available on many virtualized or remotely managed systems nor without jumping through extraordinary hoops.
Root login access, or sudo access, does not mean a developer or programmer has console access. And even if console access is available, many server class systems take quite long periods to go through hardware scanning and present only a few seconds to manually modify the boot options, and some remote management systems that nominally provide remote console access take so long to restart or reconnect after a reboot that the necessary boot options have passed by before they could be selected.
On top of this, frequently rebooting Linux systems triggers the counters on the partitions that trigger an fsck at boot time after a specific number of reboots. An unscheduled fsck on a larger storage system can make the reboot take _hours_, and can also require console access to accept or reject the fsck options. This can cause a system without console access to fail to reboot, even if the boot loader has been manually loaded, and take the system online until manual keyboad and console access can be scheduled.
How portable is the shim? I'd welcome backports of some contemporary Linux daemons to run on older Linux systems without systemd.
This is confirmed, along with the difficulty of unfurling the systemd init tools to try and start the fractured daemon cleanly in a debugger.
An open source or freeware kernel, with matching open source or freeware drivers, and robust virtualization.
> That is a good thing to keep in mind, since nobody is running systemd except by choice.
I'm afraid that's nonsense. systemd has been a penalty in work I do. If I could have more contemporary versions of Linux based daemsns and packages without systemd, such as entire LAMP stacks, I'd accept them in a moment. Debugging failed daemon startups with systemd has repeatedly proven painful, due to the binary log formats and the difficulty of deducing the actual daemon startup commands to run them in a debugger.
> Because the whiners don't have a use-case. systemd is modular, but it tends to come with all the modules packaged together
I'm afraid it's not. The dependencies among components are very strong, and it's quite difficult to segregate out one component for de-activation or non-installation unless you compile with that feature de-activated, in which case you must recompile to re-enable the future. It's very difficult to install only the components you want due to the interdependencies.
A chance to hear formal speech, listen to debates, and pick up accents is an invaluable aid to immigration. Potential immigrants do study the language quite hard when they know it will affect their income and chances of immigration.
The Timothy Zahn books from the Star Wars universe have actually been quite good. They're gone further with the old characters, introduced fascinating new characters, and told much broader stories. I don't believe they're considered "Star Wars canon", but parts of them have certainly shown up in the games. The Thrawn trilogy, itself, included _wonderful_ exploration of other racess' personalities and how individuals are much more difficult to predict than large groups of their native races. It also explored why "clones" might be a limited or difficult resource to handle in a universe with "The Force".
It's very easy after a disaster to claim that an unfunded or ignored project would have prevented the disaster. Since the whistleblowers in the article are talking about the 9/11 terrorist attack, it seems a bit late. to be blowing whistles on it now.
It does seem clear that the NSA suffered, and is suffering, from Jerry Pournell's "Iron Law of Bureaucracy"
>> First, there will be those who are devoted to the goals of the organization.
>> Secondly, there will be those dedicated to the organization itself
The amount of money, time, and manpower burned on oversampling incredible amounts of personal traffic would seem much better focused on parts of the world, and populations, where the monitoring is likely to bear more fruit. But that doesn't expand the NSA itself and its overall capacity.
> The US Constitution is about a set of enumerated powers. I don't see any of the enumerated powers that would allow the US government to intervene in society to "reduce the risk of homicides"
Please re-read the "preamble" of the constitution. Quoted below::
“We the People of the United States, in Order to form a more perfect Union, establish Justice, insure domestic Tranquility, provide for the common defence, promote the general Welfare, and secure the Blessings of Liberty to ourselves and our Posterity, do ordain and establish this Constitution for the United States of America.”
I see at least three phrases that easily include "reducing the risk of homicides", especially the one about "insure domestic Tranquility".
> It is your error to interpret such a phrase as an authorization for the government to take away individual liberties of law abiding citizen in order to reduce perceived risks
It's part of how the laws get created. "Law abiding" takes on fascinating meaning when the laws infringe on those "individual liberties", and the grounds for the law are challenged. Whether you believe that individual liberty is the ultimate goal of government, the legislature, the courts, the government, and most of civilized history does not. It's one of _several_ goals of most governments, and a laudable one, but not an absolute goal.
> First, those two goals cannot be achieved simultaneously: you can't have a society that is governed rationally and that simultaneously is free. Second, even if rational government were desirable, it couldn't be realized in practice; the more rational you attempt to make government, the more corrupt it becomes.
Perfection isn't possible. _Anything_ taken to excess becomes corrupt. It's why governments change and respond to the people who make them up, whether peacefully or on occasion violently. And it's why they attempt to balance the rights of individuals against the safety of others: it's historically one of the clearest reasons for law and governmental intervention. Gun laws, zoning laws for hazardous businesses, fences around railroads, and eminent domain used to confiscate land in flood zones and erect dams are all classic examples of government doing exactly that.
Look again, please. systemd breaks stable network configurations by unnecessarily replacing dhcp, it breaks daemon-startup debugging, it breaks decades of log analysis tools designed to work with text based rather than proprietezed binary logging format, it's repeatedly broken kernel startups, it's broken the stable model of attached storage being mounted under /media, and the attempts to replace all of "/etc" with a "stateless Linux" model is breaking tools that never volunteered to have anything to do with systemd. It's also breaking cross-platform compatibility of daemon initialization configurations.
A "light Linux user" may not see these issues becuase you wouldn't necessarily be debugging failed daemons, writing cross-platform tools, or trying to integrate stable business software with this latest fad for configurations.
Check the bottom of the man pages. There are usually related commands there.
This is why you don't trust the mixed "cable modem" devices as anything but a cable modem. Many of them also include firewall, DHCP, and wifi features. Unfortunately, the extra "features" help make them more vulnerable to this kind of remote maintenance access password abuse.
>> At any cost, no. But reducing homicides is _definitely_ one of the roles of government.
> I must have missed that part of the US Constitution.
The constitution is about means, not purposes. "Life, liberty, and the pursuit of happiness" are more clearly defined as goals in the Declaration of Independence, which has fascinating notes about what a government is supposed to do, not merely about how to do it. The famous Second Amendment was created to prevent that goal from preventing the existence of local militias.
> Making policy based on risk estimates (or probabilities in general) is inherently unjust and is not compatible with a free society.
Oh, my. I can't possibly sum up the irrationality of your analyses any better than this.
> The CIA had proof, right?
The CIA got pushed, hard, to support excuses to invade Iraq, and was under tremendous pressure from the White House to support a war. Examine the recently released interviews with former CIA analyst Ben Bonk, at http://www.newsweek.com/2015/0...
The judge can "change that" too. Reporting criminal activity by the government is a pretty good defense.
Afghanistan is considered part of the Middle East by many, and has strong language and cultural ties with the rest of the Muslim cultures of the Middle East. It's a key land route between the Middle East and Asia, which is one reason it's been so valuable and so often invaded by other nations.
And yes, Russia's invasion was marked by disaster, as was the more recent US invasion, and the previous British invasion.
> Of course, none of this even amounts to a political argument.
Of course it does. Even very strange concepts are the source of political arguments.
> That is, reducing homicide rates at any cost should not be the goal of government.
At any cost, no. But reducing homicides is _definitely_ one of the roles of government. Failure to do so is a reason why governments fall.
> Politically, whether you decide to off your neighbor with your gun has no bearing on my interest or right to own a gun.
And of course it does. The _risks_ of gun ownership are why societies make laws about it. What those laws should be are a matter of considerable political argument.
Please excuse me, I had a bit of trouble when adding the URL for the article summary. It's at http://ajph.aphapublications.o... .
http://ajph.aphapublications.o..., an American Journal of Public Health study, which finds a very direct correlation, and stacks of similar studies in their references.. The state by state results can be obscured by many factors, such as the age or income of the state populations, so it can be difficult to ensure that homicide rates can be strongly linked to homicide. But they seem to have done their homework. And the homicide rates in Europe and other gun controlled parts of the world are notably lower than the US, with reasonably good correlation with levels of gun control.
Part of the key to understanding the gun control claims and "refutations" is to watch the language very carefully. The claim that "states with the tightest gun control laws have the highest crime rates" tends to reverse cause and effect. The higher crime rates happen from poverty, racism, gangs, and other social problems, and the tighter gun control goes hand in hand with trying to _control_ those. So the statistics can be quite confusing unless one spends time correlating the other factors.
> As to things like the Paris incident, it occurs to me to wonder how easy it would have been to wander through a crowded venue shooting people at random if some of those people had been armed themselves...
The homicide rate with guns in the USA shows the difficulty. The idea that an "armed society is a politie society" was explored by Robert Heinlein in a number of his stories. In real life, the frequency of domestic violence and of violent neighborhood brawls remains so high that the deaths from household violence far outnumber those which might be saved by making personal firearms widely available.
Your other suggestions have similar difficulties. "Raising everyone's standard of living" requires steps, like educating women and reducing the gross disparities between ruling elites and grinding poverty, and birth control to prevent simply creating more starving poor, with profound social and religious consequences that are often resisted by those currently in power.
> Because someone can point a very portable RFID reader at a pile of tagged notes and get all the serial numbers, number of notes of each type and total value of the notes in a few second
Not without raising the price of the bills quite a lot. Reading a few distinct RFID tags in a small box is one task, reading 100 distinct RFID's out of a stacked box of paper is still well beyond the limit of any RFID technology. And they can be magnetically obliterated trivially. (nd get all the serial numbers, number of notes of each type and total value of the notes in a few second)
Putting an RFID reader where a money sorter passes individual bills might be effective in reading the bills casually, but RFID tags are still fragile. They don't take bending or folding well, and a great deal of money gets folded a great deal in normal handling. RFID tags also remain expensive: Adding $0.25 to the cost of making each bill would upset the US or EU mints quite a lot.
> The problem with mod_perl was one of perception:
The problem with the mod_perl update was one of naming and numbering. The version that supported httpd, sometimes called "Apache 2", was _among_ the mod_perl 1.9xxx releases, and it was very difficult to tell which "1.9xxx" release was compatible with what. Older perl modules that were only compatible with the older releases had to all be updated to require a release _less than_ 1.9xxx, or they would break when installed with "cpan install". And if your software required 2 modules, one of them with a recent release compatible only with the old mod_perl, and the other compatible only with the newer mod_perl, you had to trace back each individual module to find a version compatible with the dependencies of the other. This was hellish for web developers, and it _kept happening_ anytime anyone introduced or updated anything that used mod_perl.
The version compatible with Apache 1.3 should have retained the name "mod_perl", and it could have been supported until Debian finally stopped publishing Apache 1.3 packages alongside HTTPD 2.x packages. The new mod_perl could, and should, have been named mod_perl2 to reflect its incompatibility with the old codebase. This would have protected old modules compatible only with the older codebases and preserved forward compatibility. Newer packages could, and would, have used the new codebase by specifically calling it, without having to deduce the specific mod_perl 1.9xxx release where compatibility broke.
You're quite correct, such an internship is abusive. Someone interning at a "branding company" and being abused seems to be exactly the sort of life lesson that would be better learned in an internship than as a college graduate. But the experience for some students of "work free for a semester and get no classes, no income for it, and still pay tuition" sems to be an abuse by both the company providing the internship, and the school.