Who received the call? What did the caller say exactly? Did the police assess the situation in the neighborhood (if someone's shooting, you should be able to see that maybe something is going on in that neighborhood). Did they attempt to call the residence to contact the intruder? Did they shout to the residence from a distance to establish a contact with the intruder or the hostages? Did they observe and try to analyze the situation from afar? IN the end, we don't know enough details to be able to tell. But too often, police just assume they have to use brute force and show their muscles without using the brain first.
Let alone the stupidity of the hacker whose actions could have easily got someone killed. But what about the police?! Do they just storm in without having a clue? What if it was the wrong address?
This clearly reflects the anti-competitive, cartel-like, stagnant wireless market in North America.
There is no competition because
a. is really costly
b. FCC does not create an environment favorable to growing the market
c. Politicians protect the market
As a result,
1. there is no separation between carriers and hardwares
2. carriers lock the customers in ridiculous contracts
3. early contract termination are very costly
4. plans are ridiculously complex (have you tried comparing two plans from two different carriers?)
5. plans/contracts are among the most expensive in the developed world
The customer really does not have much choices. If Sprint can terminate a contract early, then the customer should be able to charge an early termination fee to Sprint:-)
As usual, like everything that is MS there's hardly any innovation and originality in this. Except perhaps some new patents and the ability to mass market this product. This technology has been worked on for many many years, most notably by Mitsubishi and a lot of university labs.
Of course it will enable unscrupulous realtors, politicians, decision-makers, and corporations to make unethical use of this tool. At the same time, it will also allow the public to have easy and intuitive access to aspects of everyday life they were conveniently kept hidden from them...
As a statistician working in cancer research I feel we are many years ahead. These days it's hard to publish without a statistician review your paper or having a statistician as an author. I still see a lot of bad science everyday and a lot of it is due to a lack of understanding of basic methodology. In my mind, statistics is what gets you closer to good methodology. Anyone in CS who is considering doing any research, be the most basic of it, needs an understanding of statistics and probability. An example in this field would be the numerous web-statistics many web-hosts offer (webalyzer, awstats etc.) Most of the ones I have seen are either wrong, misleading or irrelevant. It would not take much for a good CS person with a good foundation of statistics to come up with some good methodology to analyze this data in a sensible way. Let alone if you are doing simulation studies, as DoofusOfDeath suggests, or want to do some hard core modelling. When I see white papers suggesting how they found out that most people use MS OS as their web-servers, I just laugh. It only takes a few lines of reading to see what bad science or bad research is all about. UNfortunately, most people showing statistics or talking about statistics know nothing about it.
Unfortunately, the opposite is also true. We statisticians need more training in CS. I had to teach myself several programming languages, database, limitations of computing environments, operating systems, etc.
Excel is decent all-purpose spreadsheet, but it excels in nothing. Until a few years ago, Excel had serious well-proven flaws in certain algorithms contained in the Analysis toolpack as well as they way it approximates number. Let alone the help section, which contained many errors. I was told it hasn't changed much. If you are serious on data analysis, inferential analysis, decision making, scenarios generation etc. you should investigate what other options are available. Most likely you will find a package that specializes in your area. To me, if you say you're truly serious about, say, data analysis and still use Excel, clearly I would be highly skeptical about your results.
The reasons so many people use Excel is because they feel confortable with it and/or it is the only software package they know.
P.S. Someone suggested 2+2*3=12 when done with a calculator and ==8 when done in Excel. This is a fairly basic issue: in a calculator you can only do one operation at a time, so first you type 2+2 then you multiply the result by 3. In Excel (or Matlab or Maple, or SAS or R or SPLUS or Stata or...) you have to use appropriate mathematical notation whereby multiplication and division takes precedent over addition and substraction. So there is a difference between (2+2)*3 and 2+2*3
Slashdot Mirror
Who received the call? What did the caller say exactly? Did the police assess the situation in the neighborhood (if someone's shooting, you should be able to see that maybe something is going on in that neighborhood). Did they attempt to call the residence to contact the intruder? Did they shout to the residence from a distance to establish a contact with the intruder or the hostages? Did they observe and try to analyze the situation from afar? IN the end, we don't know enough details to be able to tell. But too often, police just assume they have to use brute force and show their muscles without using the brain first.
Let alone the stupidity of the hacker whose actions could have easily got someone killed. But what about the police?! Do they just storm in without having a clue? What if it was the wrong address?
This clearly reflects the anti-competitive, cartel-like, stagnant wireless market in North America. There is no competition because a. is really costly b. FCC does not create an environment favorable to growing the market c. Politicians protect the market As a result, 1. there is no separation between carriers and hardwares 2. carriers lock the customers in ridiculous contracts 3. early contract termination are very costly 4. plans are ridiculously complex (have you tried comparing two plans from two different carriers?) 5. plans/contracts are among the most expensive in the developed world The customer really does not have much choices. If Sprint can terminate a contract early, then the customer should be able to charge an early termination fee to Sprint :-)
As usual, like everything that is MS there's hardly any innovation and originality in this. Except perhaps some new patents and the ability to mass market this product. This technology has been worked on for many many years, most notably by Mitsubishi and a lot of university labs.
A simple incidence rate like (# attacks)/(total users) for each IP would have been much more informative.
Of course it will enable unscrupulous realtors, politicians, decision-makers, and corporations to make unethical use of this tool. At the same time, it will also allow the public to have easy and intuitive access to aspects of everyday life they were conveniently kept hidden from them...
As a statistician working in cancer research I feel we are many years ahead. These days it's hard to publish without a statistician review your paper or having a statistician as an author. I still see a lot of bad science everyday and a lot of it is due to a lack of understanding of basic methodology. In my mind, statistics is what gets you closer to good methodology. Anyone in CS who is considering doing any research, be the most basic of it, needs an understanding of statistics and probability. An example in this field would be the numerous web-statistics many web-hosts offer (webalyzer, awstats etc.) Most of the ones I have seen are either wrong, misleading or irrelevant. It would not take much for a good CS person with a good foundation of statistics to come up with some good methodology to analyze this data in a sensible way. Let alone if you are doing simulation studies, as DoofusOfDeath suggests, or want to do some hard core modelling. When I see white papers suggesting how they found out that most people use MS OS as their web-servers, I just laugh. It only takes a few lines of reading to see what bad science or bad research is all about. UNfortunately, most people showing statistics or talking about statistics know nothing about it. Unfortunately, the opposite is also true. We statisticians need more training in CS. I had to teach myself several programming languages, database, limitations of computing environments, operating systems, etc.
Aren't {Truth} and {Opinio/Fiction} mutually exclusive?
Excel is decent all-purpose spreadsheet, but it excels in nothing. Until a few years ago, Excel had serious well-proven flaws in certain algorithms contained in the Analysis toolpack as well as they way it approximates number. Let alone the help section, which contained many errors. I was told it hasn't changed much. If you are serious on data analysis, inferential analysis, decision making, scenarios generation etc. you should investigate what other options are available.
Most likely you will find a package that specializes in your area. To me, if you say you're truly serious about, say, data analysis and still use Excel, clearly I would be highly skeptical about your results.
The reasons so many people use Excel is because they feel confortable with it and/or it is the only software package they know.
P.S. Someone suggested 2+2*3=12 when done with a calculator and ==8 when done in Excel. This is a fairly basic issue: in a calculator you can only do one operation at a time, so first you type 2+2 then you multiply the result by 3. In Excel (or Matlab or Maple, or SAS or R or SPLUS or Stata or...) you have to use appropriate mathematical notation whereby multiplication and division takes precedent over addition and substraction. So there is a difference between (2+2)*3 and 2+2*3