Quoting the austrian school in serious economic discussions is like quoting creationists or flat-earthers. It's pseudo-science to a degree that real economists are embarrassed by them.
The first part of what the say makes sense, but they use that as a springboard to jump to a lot of apparent nonsense. But then, a lot of mainstream economics is also apparently nonsense. Do you have links that would help show why their nonsense is worse that everyone else's nonsense?
I've yet to find a serious economist or capitalist who believes that no action was a valid alternative--we were a hair away from a complete financial halt in the credit/security market which would have quickly halted our entire economy. But, of course, most people haven't taken basic economic courses (much less advanced ones).
What about the entire Austrian school, which holds that government meddling is what caused the crisis and more meddling can only make things worse?
Borrowing ideas then in vogue among private sector consultants and CEOs to "reengineer" organizations to make them more "customer-driven," Congress instructed the patent office, which had always been funded from government revenues, to now pay its own way through fees charged to applicants, and to make the process of winning a patent easier on them.
Trouble being, the patent applicants aren't actually the customers -- they're the product. The customers would be all the rest of us, who get the benefits of the (supposedly) increased innovation the patent system (supposedly) provides.
where your employees bill time by the tenth of an hour
So, 800 time slots per employee for two weeks.
You have to tabulate all the hours for each employee for the month, and then allocate each hour spent on each day to each client, each client's job, each phase of said job, and each task under that job. Spread that across 1000 active clients with 1-2 jobs each, many with multiple phases, and all with multiple task codes.
A single lookup on <1e6 rows that probably all fit in memory takes 10ms (or 3ms if we assume 24 hours days instead of 8 hour days)? Did they forget an index, or write all the calculations and business logic in shell script or something? Processing extents instead of blocks would also help, you won't usually switch projects every 6 minutes...
If a voter sees their vote registered wrong, it has to be a mistake. Maybe the machine isn't calibrated properly for that voter's viewing angle (my guess), or maybe some nefarious programmer f-ed up and called SwitchVote() in the wrong place, but it's still a mistake. Either way it sounds like they can just touch the screen in a slightly different place to get the result they want, so any possible nefarious intent should be ineffective.
Maybe the machines should do the "touch the center of the X" thing with every single voter to make sure they're properly calibrated for the viewing angle of each voter. Most public computer kiosks I've seen have very thick covers / empty space / whatever between the touch surface and the actual display, being too tall / short could easily result in a half inch or more offset from where you thought you touched.
Should I be against spam for any other reason than I am annoyed by them?
I don't think spam should be any more illegal than billboards, flyers, or direct mailings.
Sure, as long as it has some consistent header field that lets me filter it all out without dropping any legit mail. But somehow, I don't think the people spamming from botnets are going to care about little details like that...
Complete bullshit article, doesn't offer any useful information beyond a completely obvious conclusion -- the more features that are added to a given piece of software, the higher the demands on your PC.
I would think that those increased demands should be mostly in the form of slightly (a few MB) higher memory requirements to store the extra code for those features. Adding new functionality should not impact existing functionality. Haven't you heard of the zero-cost principle (idea from C++ and apparently Perl, "you don't pay for (as in take a performance hit from) what you don't use")?
The only freedom BSD license gives is to allow the distributor able to make the software less beneficial to the customer than GPL'd software.
The various copyleft licenses can't generally be mixed (the most publicized example is probably how GPLv2-only code can't be combined with GPLv3 code), whereas BSD style licensing lets you mix the code with whatever you want.
It's a pity it takes an economic crisis to get companies to look into a better way of doing business.
The problem is that the rest of the time, the government is fixing the interest rate too low and putting a disincentive on any sort of real savings/investment. So people are made to only care about the short term (because they're penalized for trying to plan long-term), and they just can't be bothered to look around at what might be a bit scary but better in the long run.
How hard is it to work out that the software distributed in a manner that it benefits people (customers) will eventually gain dominance over software that is distributed in a manner that restricts customers for the benefit of the distributor?
This is why BSD-style licensing will eventually take over the world, after proprietary licensing dies off and causes the GPL and friends to lose steam.;)
Which is more useful to you as a developer? Code that demands to only be used with other code under the exact same style of mandated freedom, or code that can be used for anything (but you know that if you don't share, your potential customers will ignore you)?
We will not likely be able to affect things like TV and Internet usage times,
How long can a TV run from a car/truck battery? If electricity prices varied by time of day and/or your connection's current power draw, it might actually be cost-effective for people to run some daytime things from batteries that they could recharge overnight.
No, that's an "encoding". The main difference is that encrypted information is (supposed to be) unusable without the key even if you know the encryption algorithm, while an encoding doesn't have a key. For instance I think ROT-13 is really more "encoding" than "encryption", because there's no key. But this is a bit fuzzy, since it's just a common case of ROT-n (with variable n), which does have a (very weak) key.
If you think dealing with encryption won't waste $50/mo of each employees productivity, you're mistaken.
My work laptop has full-disc encryption. The only time I notice is when it asks for a boot password or when I have to change the password every couple months. This is completely negligible compared to, say, the time to boot Windows and open all the horribly bloated (and network-aware, so they also take time to connect to the server) applications I have to use.
What magical encryption do you have that doesn't slow the system at all?
It's not the encryption, it's having a system with a processor made in the last 5 years. Spinning plates of rust are already insanely slow, adding symmetric encryption on top of that won't make a difference.
There are still people running legacy systems that do not support encryption. Nor is it fast, easy, cheap, to get them to do so.
I don't think most legacy systems are portable, so they shouldn't be a problem.
Also I could see huge problems later on when the only IT guy who knows the key is fired, hit by the obligatory train, or quits.
And then you learn why you shouldn't do that. Write it down and put it in a safe, or have at least 3-4 people who know it.
Forcing encryption isn't the answer but penalties and legal repercussions if your data stolen is more appropriate. While it is not the right time to politically say this. It is a case where they don't really need government intervention as most companies will regulate themseles on this front especially if they don't have immunity to legal problems if something goes wrong.
Really? This only works if the company can (and does) truly repair any damage caused by a data leak. If the damage is more than the company is worth, or is something that can't be repaired, or the company can litigate people into submission, then there are "negative externalities" and ordinary market forces don't work so well.
It seems like the Democrats are doing the same thing the republicans did after 9/11. Just as after 9/11 the Republicans pushed Security to an extremist state, Democrats are using the financial crisis to push down all those heave regulations down our mouth. Jast as 7 years ago. They went those Damn Democrats were to soft on security and look what happened, now the democrats are going, Those damn republicans they were so soft on regulating companies and look what happened.
How does the DVD player know which lights to dim, or how far to dim them? There's a lot more than just providing data connections ("mechanism"), there also needs to be configuration data ("this light is in that room") and policy ("when a movie starts in some room, dim the lights in that same room by 90%", "if the fire alarm goes off, turn all lights to full brightness regardless of other rules", etc).
Since you need a central controller to implement the policy anyway, why would you want internet connectivity for every individual device?
I wonder what the opposite strategy would do... have the firewall intercept pings, but instead of just dropping them, pretend to be the target and answer them itself.
As a developer, isn't the point to write better/more robust code??
You forgot "more useful". The more restrictions you impose on your customers, the less useful your code is.
I may be ranting here (apologies in advance), but railing on MS for their past business practices (which I don't condone, BTW) is pointless.
To a degree... but if you're relying on their future behavior (ie, not forcing silverlight to be windows-only), looking at past behavior is a decent idea.
I tend to use the best tools available for the technologies that I code for, and Microsoft has some good ones!
Yes, they do. But if I don't want my products to be tied to Windows, I have to either look elsewhere or trust Microsoft (in spite of all their history) to play nice.
Slashdot Mirror
The first part of what the say makes sense, but they use that as a springboard to jump to a lot of apparent nonsense. But then, a lot of mainstream economics is also apparently nonsense. Do you have links that would help show why their nonsense is worse that everyone else's nonsense?
What about the entire Austrian school, which holds that government meddling is what caused the crisis and more meddling can only make things worse?
Borrowing ideas then in vogue among private sector consultants and CEOs to "reengineer" organizations to make them more "customer-driven," Congress instructed the patent office, which had always been funded from government revenues, to now pay its own way through fees charged to applicants, and to make the process of winning a patent easier on them.
Trouble being, the patent applicants aren't actually the customers -- they're the product. The customers would be all the rest of us, who get the benefits of the (supposedly) increased innovation the patent system (supposedly) provides.
where your employees bill time by the tenth of an hour
So, 800 time slots per employee for two weeks.
You have to tabulate all the hours for each employee for the month, and then allocate each hour spent on each day to each client, each client's job, each phase of said job, and each task under that job. Spread that across 1000 active clients with 1-2 jobs each, many with multiple phases, and all with multiple task codes.
A single lookup on <1e6 rows that probably all fit in memory takes 10ms (or 3ms if we assume 24 hours days instead of 8 hour days)? Did they forget an index, or write all the calculations and business logic in shell script or something? Processing extents instead of blocks would also help, you won't usually switch projects every 6 minutes...
So, really, it isn't a UI design issue; it's a voting machine response time/feedback issue, IMO.
How is response time/feedback not a UI issue?
If a voter sees their vote registered wrong, it has to be a mistake. Maybe the machine isn't calibrated properly for that voter's viewing angle (my guess), or maybe some nefarious programmer f-ed up and called SwitchVote() in the wrong place, but it's still a mistake. Either way it sounds like they can just touch the screen in a slightly different place to get the result they want, so any possible nefarious intent should be ineffective.
Maybe the machines should do the "touch the center of the X" thing with every single voter to make sure they're properly calibrated for the viewing angle of each voter. Most public computer kiosks I've seen have very thick covers / empty space / whatever between the touch surface and the actual display, being too tall / short could easily result in a half inch or more offset from where you thought you touched.
Should I be against spam for any other reason than I am annoyed by them?
I don't think spam should be any more illegal than billboards, flyers, or direct mailings.
Sure, as long as it has some consistent header field that lets me filter it all out without dropping any legit mail. But somehow, I don't think the people spamming from botnets are going to care about little details like that...
Complete bullshit article, doesn't offer any useful information beyond a completely obvious conclusion -- the more features that are added to a given piece of software, the higher the demands on your PC.
I would think that those increased demands should be mostly in the form of slightly (a few MB) higher memory requirements to store the extra code for those features. Adding new functionality should not impact existing functionality. Haven't you heard of the zero-cost principle (idea from C++ and apparently Perl, "you don't pay for (as in take a performance hit from) what you don't use")?
Then why does Greenspan say that he was wrong and it needs more regulation?
Because he's part of the problem?
...at which point there's one less truck driving around, and the environment is just a little bit happier. :-p
The only freedom BSD license gives is to allow the distributor able to make the software less beneficial to the customer than GPL'd software.
The various copyleft licenses can't generally be mixed (the most publicized example is probably how GPLv2-only code can't be combined with GPLv3 code), whereas BSD style licensing lets you mix the code with whatever you want.
It's a pity it takes an economic crisis to get companies to look into a better way of doing business.
The problem is that the rest of the time, the government is fixing the interest rate too low and putting a disincentive on any sort of real savings/investment. So people are made to only care about the short term (because they're penalized for trying to plan long-term), and they just can't be bothered to look around at what might be a bit scary but better in the long run.
How hard is it to work out that the software distributed in a manner that it benefits people (customers) will eventually gain dominance over software that is distributed in a manner that restricts customers for the benefit of the distributor?
This is why BSD-style licensing will eventually take over the world, after proprietary licensing dies off and causes the GPL and friends to lose steam. ;)
Which is more useful to you as a developer? Code that demands to only be used with other code under the exact same style of mandated freedom, or code that can be used for anything (but you know that if you don't share, your potential customers will ignore you)?
We will not likely be able to affect things like TV and Internet usage times,
How long can a TV run from a car/truck battery? If electricity prices varied by time of day and/or your connection's current power draw, it might actually be cost-effective for people to run some daytime things from batteries that they could recharge overnight.
The summary is trying to make this about "un-witnessed searches," but this is about dishonest transport employees.
...who only have the opportunity/incentive to be dishonest because of the "un-witnessed searches", yes?
No, that's an "encoding". The main difference is that encrypted information is (supposed to be) unusable without the key even if you know the encryption algorithm, while an encoding doesn't have a key. For instance I think ROT-13 is really more "encoding" than "encryption", because there's no key. But this is a bit fuzzy, since it's just a common case of ROT-n (with variable n), which does have a (very weak) key.
If you think dealing with encryption won't waste $50/mo of each employees productivity, you're mistaken.
My work laptop has full-disc encryption. The only time I notice is when it asks for a boot password or when I have to change the password every couple months. This is completely negligible compared to, say, the time to boot Windows and open all the horribly bloated (and network-aware, so they also take time to connect to the server) applications I have to use.
What magical encryption do you have that doesn't slow the system at all?
It's not the encryption, it's having a system with a processor made in the last 5 years. Spinning plates of rust are already insanely slow, adding symmetric encryption on top of that won't make a difference.
There are still people running legacy systems that do not support encryption. Nor is it fast, easy, cheap, to get them to do so.
I don't think most legacy systems are portable, so they shouldn't be a problem.
Also I could see huge problems later on when the only IT guy who knows the key is fired, hit by the obligatory train, or quits.
And then you learn why you shouldn't do that. Write it down and put it in a safe, or have at least 3-4 people who know it.
Forcing encryption isn't the answer but penalties and legal repercussions if your data stolen is more appropriate. While it is not the right time to politically say this. It is a case where they don't really need government intervention as most companies will regulate themseles on this front especially if they don't have immunity to legal problems if something goes wrong.
Really? This only works if the company can (and does) truly repair any damage caused by a data leak. If the damage is more than the company is worth, or is something that can't be repaired, or the company can litigate people into submission, then there are "negative externalities" and ordinary market forces don't work so well.
It seems like the Democrats are doing the same thing the republicans did after 9/11. Just as after 9/11 the Republicans pushed Security to an extremist state, Democrats are using the financial crisis to push down all those heave regulations down our mouth. Jast as 7 years ago. They went those Damn Democrats were to soft on security and look what happened, now the democrats are going, Those damn republicans they were so soft on regulating companies and look what happened.
WTF?
So use protection.
Or top truncating titles into something ambiguous, I guess.
Trivial?
How does the DVD player know which lights to dim, or how far to dim them? There's a lot more than just providing data connections ("mechanism"), there also needs to be configuration data ("this light is in that room") and policy ("when a movie starts in some room, dim the lights in that same room by 90%", "if the fire alarm goes off, turn all lights to full brightness regardless of other rules", etc).
Since you need a central controller to implement the policy anyway, why would you want internet connectivity for every individual device?
I wonder what the opposite strategy would do... have the firewall intercept pings, but instead of just dropping them, pretend to be the target and answer them itself.
Would giving them back do anything other than encourage network providers to procrastinate on IPv6 for another couple years?
As a developer, isn't the point to write better/more robust code??
You forgot "more useful". The more restrictions you impose on your customers, the less useful your code is.
I may be ranting here (apologies in advance), but railing on MS for their past business practices (which I don't condone, BTW) is pointless.
To a degree... but if you're relying on their future behavior (ie, not forcing silverlight to be windows-only), looking at past behavior is a decent idea.
I tend to use the best tools available for the technologies that I code for, and Microsoft has some good ones!
Yes, they do. But if I don't want my products to be tied to Windows, I have to either look elsewhere or trust Microsoft (in spite of all their history) to play nice.