One of the big reasons for this is that outside the USA, generally people do not pay to receive calls on mobile phones; the caller pays a higher cost to call a mobile number than a landline instead (at least in theory, although inclusive minutes deals make this increasingly not the case for either the USA or rest of world). One of the principles that seems to be broadly applied in the numbering systems used in most countries is that you should be able to tell whether a number is an "expensive" one or not by looking at the prefix. Allowing higher cost for calls to mobiles would break this principle (it also makes sense logically, since mobiles are non-geographical so giving them a geographical prefix is a bit weird).

And individuals get plea bargains. Same difference really, just the available penalties for corporate infringement are more limited.

One other fact which appears to have been massively under-reported is that, from what I understand, their definition of "metadata" includes location data for cellphones (ie at least which tower you were connected to, and potentially a tower signal-strength triangulated position). Simply knowing where you made your calls from (and where the recipient was) can allow someone to infer an awful lot about what might have been said on those calls. Especially if they can then cross-reference that with e.g. credit card records etc.

gscan2pdf is OK, but if you want to do this seriously then you're probably going to want a reasonably fast sheet-fed scanner (I got a Fujitsu ScanSnap S1500, which is supported by SANE and can scan at 18-20 pages/36-40 sides per minute) with a button so that you can go through a whole stack of paper quickly with minimal keyboard/mouse interaction to slow you down. This led me to setting up scanbuttond (which just gained official support for the ScanSnap but there was a patch floating around somewhere for a while before that) with a custom script.

Make sure you OCR your documents to make them searchable then run an indexer (I like recoll but KDE and GNOME both have their own desktop search solutions as well). I've found the best OCR engine on Linux seems to be tesseract, but there are a couple of others you can try. The process took me a while to get right and is a bit painful - the script which scanbuttond runs runs scanadf to scan to a string of image files per side and puts them in a processing directory. I then have another batch-processing script I run once I'm done with a pile of papers while I go and get a cup of tea which runs unpaper then tesseract on them, then hocr2pdf to convert each page individually into a searchable PDF file then finally pdftk to concatenate all the pages together into a scanned document. I split the two parts of the process out because the OCR bit can take some time and this way I can get maximum throughput on the scanner itself without needing to wait for the rest to catch up. If I could be bothered then I could make the scanning script run my de-batching script once only and have it pick up new files as they are dropped in the directory but it's not that much of an effort really.

I then sort my PDFs into a hierarchical directory structure once they've been OCRd (and at this point they get indexed as well for searching).

If you're on Windows/Mac then the software that comes with the ScanSnap will pretty much do all this for you; although it's better to scan with OCR disabled then use Acrobat to batch-OCR the PDFs later for the same reason. Add a decent desktop search solution like an old version of Copernic (or possible Windows Search) and all is good.

But if the "cur" directory in the Maildir is corrupted then you're back to the same problem; in fact potentially worse depending on how resistant to corruption your filesystem is (will that trash the whole directory or just a part of it? If a single file has corruption in the middle of it can you still read before and after the corruption?). The correct solution to corruption concerns like that is to make good backups. Maildir has advantages over mbox in terms of the consequences of a crash/segfault/whatever while your mail client is writing the mailbox causing an issue but for an archive you won't ever write to it so this shouldn't be an issue.

Completely agree here. The Canon 10-22 seems to have a better reputation than the rather-maligned 17-40 (although I personally like the Tokina 11-16/2.8). And the Canon 17-55/2.8 has a very good reputation in terms of "normal" zoom lenses - probably equal to the 24-70/2.8 which is the FF "equivalent" only it's lighter and has IS (albeit that f/2.8 on FF is arguably really like f/2.0 on APS-C as FF has both shallower DoF and better low-light capability). Plus you can generally use all the FF lenses but just take the crop factor into account to buy slightly different lenses - e.g. you would buy the Canon 85/1.2 (or the Sigma 85/1.4) and use it on APS-C where you would be using the 135/2.0 on FF and buy something like a 50/1.2 or 50/1.4 where you would be using one of the 85s on FF.

APS-C (and APS-H for that matter although Canon have discontinued that line of sensors now) also allows you to "get away with" more on FF-designed lenses since you lose the "worst" parts of the frame which are typically the corners. So it's not all bad. Plus the extra pixel density in reach-limited situations means more reach (since there's no 47MP FF camera yet the 7D still has more reach than any other camera right now).

APS-C loses out to FF mainly in terms of low-light photography as it's at least a stop "worse" due to physical limitations - and this is why e.g. the 5D range is so popular with wedding photographers who are constantly available light challenged. Many people also talk in terms of the "quality" of each pixel being worse - i.e. at the same resolution, crop is noisier than FF since the pixels are physically smaller - although a lot of that difference can be made up for in post-processing; so another reason the FF cameras are more popular among pros is that it reduces workload in post-processing to get decent results.

That's 20 years from the filing date though, not 20 years from the date of invention/date of first spend on R&D. There are often several years of research which happen before the filing date.

Actually, this is exactly what used to happen before (roughly, and depending on where you live) the early to mid 19th century. The earliest firefighters in modern times were either volunteers or employed on a private contractual basis (ie they would literally turn up at the scene of a fire and try to negotiate payment before putting it out). As insurance developed in the 17th century, naturally insurers started to provide their own firefighters to reduce the losses sustained to fire. The insurers in London, for example, set up a system after the Great Fire of 1666 whereby each had their own group of firemen and they placed "fire insurance marks" on each house so that they could identify whether their unit was supposed to fight a particular fire or not. Eventually the usual pressures of commerce meant that these units usually merged into a single unit covering the whole of London across multiple insurers in the early to mid 19th century, however, still under a model of privately funded provision.

What happened next could be viewed as an example of "corporate welfare"... the insurers lost large sums in a few particularly bad fires and they decided as a result of this that they would lobby the government to provide a beefed-up firefighting service at taxpayers' expense. At the same time, there was a growing movement to "profesionalise" the remaining voluntary provision in other parts of the world which led to them becoming paid rather than voluntary. Following the model set in the insurer-led markets, these areas paid their firefighters out of the public purse.

I would suggest that it seems the right thing to do to fund fire defence by extracting the costs directly from insurers on an incident basis rather than simply relying on general taxation - i.e. if my house catches fire, my insurer would then have to pay the government back the cost involved in calling the fire brigade out (you can argue about the corner case of how to deal with people who are uninsured and whether you fund their costs from general taxation, a levy on those who are insured, or by trying to pursue them individually). One benefit is that the insurers then have even more incentive (beyond just the threat of loss) to ensure that fire prevention measures are adequate. You could also compare this to the idea that the court system should be self-funding through filing fees etc. Just because it's a legitimate use of a government monopoly, doesn't mean it has to be funded through general taxation.

Completely agree with this. The multimonitor support on Radeon is much, much better than nVidia and that's why I moved over as well. I wouldn't say I've had any big "issues" but ATI's driver support (at least on Linux, using the Catalyst drivers) has been a little bit disappointing - I had to stay on an old version of X.org for a while because of the amd64 Xv issue forcing me to use an older driver for example.

No, "free schools" are a special type of state school and "free" means that they are free from a number of the diktats usually imposed upon the rest of our state-funded schools, including the requirement to adhere to the national (government-mandated) curriculum. They are a new thing in the past year or two. The idea was to get rid of some of the bureaucracy involved in founding a school so that groups of parents and other people could more easily open their own new schools to create more competition in the state-funded sector which in turn would drive up standards across the board.

That's not completely true. The UK does have a legal recognition of the right to free speech, both uncodified (it is accepted as part of the common law) and as codified through the Human Rights Act (amongst other statutes). It's just that the exceptions to this right are rather broader here than in many other countries (notably the US).

I disagree that this is going to make politicians less sensitive to requests to narrow those exceptions. It's all about how it's framed. When everyone is being asked to tighten their belts either through higher taxes or lower service provision from the public sector (or both), it seems pretty difficult to justify the police (and, if it goes that far, the crown prosecutors and the courts) spending their time on things as trivial as this. It really doesn't help their case when they are complaining about cuts to their resources.

I don't think that means that this law will be removed from the statute books (unfortunately). But I wouldn't be surprised if police forces and individual officers are told (perhaps quietly, perhaps not) to use a bit more common sense about these types of cases when exercising their discretion.

I'm not sure which numbers you were looking at, but I think they are rather a lot higher than that, even if not officially acknowledged as such. Possibly you have confused "being shot by the police" with "being killed by the police" (although even then the number is far, far higher than that).

Between 2000 and 2011 there were nearly 6,000 deaths in police custody in the UK. Now, some (perhaps even many or most) of those will be unavoidable - perhaps people who would have died anyway even if not in police custody. Then, some are down to negligence (although I'd argue that in many case that is just as bad as malfeasance - if I was at home and vulnerable to some medical condition e.g. diabetes then it's much more likely someone would be around who would watch and look after me properly). But I find it very difficult to believe that given such a large number of cases there is no significant element of either bad intent or intentinoal recklessness, because it really is a shockingly high number - for context, it is not terribly far off the total number of murders recorded in the UK in the same period.

Looking just at shootings - there seem to be on average about 6 or 7 a year in recent years - e.g. here is a list of some of them. There are in fact multiple recent cases where the police have literally shot naked and unarmed people (and faced only relatively minor consequences as a result) and several more where they have shot unarmed people. Even in this case, which would appear to be about as clear-cut a case as they come, the officers were acquitted and retained their jobs in the police, albeit not on firearm duties.

Finally, I'd like to say that the fact that police can apparently get away with murder should worry you, for two reasons. Firstly - not because you might be murdered by the police yourself (that is still very unlikely), but because it means they might be likely to get away with far lesser crimes (like assaulting you, planting drugs on you, or making up a traffic offence because they decide they don't like the look of you) much more easily. Secondly - because it is indicative of a force who don't see their primary loyalty as being to the victims of crime (and to thus solving crime) but rather to looking after their own. If you were a victim of crime, would you want a force where the officers thought people who didn't pull their weight to solve it effectively should be protected from public scrutiny?

If anything, we should be holding police officers, especially firearms officers, to a higher standard than we do the general public because we grant them additional powers and privileges and entrust them to use those responsibly while paying them out of the public purse.