Bruce made a throwaway comment that NSA should be spying on foreign governments and leaders. Obviously, more effort should be put on enemies and major rivals than others. What is not wanted is the NSA spying on Americans in America.
I am not be surprised at the amount of useful intelligence that could be gleaned from social media sites. To expect your average ISIS dim bulb not to post actionable intelligence on a social media site is ludicrous. If they will post very personal details why would anyone expect them not post something juicier.
Whether the story referenced is true, plausible but not proven would be my verdict.
Agreed with the strong possibility of reasonable doubt in many cases.
A comment on the Simpson case; when I heard how the LAPD handle evidence - no proper chain of custody - I could not see how any competent jury would convict him. In fact, if Ito had a pair he would have slammed the DA's and the LAPD.
I suspect with the newer languages being developed (Go, Julia, and Rust for example) Java in particular and to a lesser extent C/C++ may see declining use anyway. When Java's APIs are being aggressively litigated many smaller developer houses and solo developers may not pay up and look for some other language.
The book "Statistics Done Wrong" by Alex Reinhart (No Starch Press) covers this (without any math). All too often studies are designed without considering the sample size required to reduce false positive and negative rates to a small number and the fact that most papers do not publish the full statistical treatment.
The study mimics, deliberately, the fact that too many medical studies are done using excessively small sample sizes. This means there is a tendency for false positive and negatives. Also, they are chastising the numerical illiteracy of the press who do not even try to understand statistics.
There is a little appreciated point about this debate: many chemicals in food have what chemists call stereoisomers. They have the same chemical components but in 3-d the structures are either left or right handed. This handedness may or may not be important; Thalidomide is a drug where one stereoisomer causes birth defects and the other is a useful drug. However this does not necessarily have anything to do with artificial or natural.
Sony's problems reflect badly on the movie industry to a degree but the movie industry has an image problem from before. Hollywood accounting is well known. Disney is in the movie/entertainment industry and I doubt they are clean.
The problem is with this is the disable almost always get the shaft. Part of the problem is the disable are disparate group with all sorts of medical issues. So there is often a fix for one group that could cause problems for others.
People can be divided into 4 groups on computer security: 1. White hats who try find flaws and alert the appropriate people about them so the flaws can be fixed (Roberts). This is relatively small group because of the skills needed to find the flaws. 2. The security aware who try keep up with the field but lack the deep technical knowledge to routinely find flaws. The is not a particularly large group but an important because they often provide a link to educate others about best practices. Readers of Techdirt and similar blogs are in this group. 3. The average user who does not keep up with most security issues and may not understand their implications. They are heavily reliant on their technically aware friends and family members for advice, training, and support. This by far the largest group of users. Often they confuse the white hats with the black hats, especially by the technical illiterates in the criminal injustice system. 4. The black hats who use flaws to harm others usually financially but occasionally in other ways. Many black hats are script kiddies who do not have the technical skill to find the flaws. They are reliant on more skill black hats to find them and determine how to exploit. This is a very small but dangerous group.
Fiber evidence and similar evidence is used to exclude not to prove. Say you find there are 30 sort of viable suspects for a crime. Fiber evidence proves that 25 could not have done the crime, you are now left with 5 good suspects to investigate.
Now multiple different types of fibers (hair, carpet, clothing) combined could narrow the list done to 1 or 2 people. But you should have other evidence to tie the case together.
The real scandal is not the dodgy fiber evidence but how much of the other evidence is equally dodgy.
Juries can only decide based on the evidence and testimony they have. If these are flawed the jurors are liable to convict an innocent person because, essentially, they were lied to and had no point of reference to prove the lie. If OJ Simpson did not have money he would have convicted because "minor" issues like improper sample chain of custody would never have come out. In science, the analysis is only as good as the sample it is provenance.