As I acknowledged, taxation may be a necessary evil, at least for now. But there is no logical basis for assuming it must always be so.
Consider that whether you call it patronage, charity, philanthropy, or something else, many people from the minimum wage hired help to billionaires already give a substantial chunk of their personal income to causes they consider worthwhile. I personally believe that most people are decent, and understand that some public/common services are in everyone's interests, and would be willing to contribute to society without being forced given a reasonable system for doing so.
Moreover, many people (sometimes whole countries) pay privately for services that other people (sometimes whole countries) consider essential public services. So demonstrably a lot of things that "big government" does don't necessarily have to be funded from the public purse.
And crucially, the modern world has technology and communication tools that could allow people to collaborate and direct their money to what they considered worthwhile causes on a scale never before seen in human history, without imposing a heavy burden on anyone who was willing to contribute but didn't have time to learn about the different possibilities in detail. We have the means to build a system where people could be involved in directing what their money supported to some extent if they wanted to, while still allowing others to choose a sensible "default" distribution for any donations as recommended by the government of the day.
I don't see us getting away from taxation any time in the near future, as it would require a huge cultural shift to make the kind of system I described above a success. However, I see no reason in principle that we couldn't have an alternative system where everyone understood that key services needed to be supported and gave what they could/what the services needed to do that. This would also have the useful side effect of making it very obvious when an important service was under-funded so it could attract more support, which is something that doesn't always happen today because under-funded public services appear to be "someone else's problem" if the expectation is that they are supported via tax revenues.
Speaking as someone in the UK who does run his own business and therefore can do those things you described, I would like to point out that the situation is not nearly as one-sided as you are suggesting.
For a start, you can only do those things if you are not considered an employee. You can't just decide to call yourself a contractor and magically opt out of the normal tax/NI system when you're still doing a regular job for an employer. People used to try and do this, but with the introduction of the IR35 rules a decade or so ago, HMRC (the tax authority in the UK) can go after people who are merely "disguised employees" to get the missing tax income back.
Moreover, for those of us who really are running our own businesses rather than being someone else's employee, anything we save through not making the same payments as a typical full-time employee is usually far outweighed by what we lose because no-one pays us for time off, we have to provide our own equipment and training, we're responsible for paying all the basic bills like heating/lighting/phone/Internet, and so on.
There are certainly some great perks to running your own small business against being an employee. In a very real sense, you get to control your own destiny. You can be much more flexible with when and how you work, and fit paying the rent around your lifestyle rather than the other way around. I think our education and employment systems don't do enough to promote alternative careers to those who might enjoy and benefit from them.
But mysteriously saving a fortune in tax isn't a good reason to go down this path, and if that's the only reason someone is doing it then they're going to be sadly disappointed. Contractors often do make more money than their employed counterparts, but contrary to the impression given by certain mainstream media outlets, the real financial benefits don't come from dodging a bit of NI, they come from having much more flexibility to negotiate what you're paid based on the value you actually generate for customers/clients rather than being stuck with a market value salary where your employer keeps most of the profits in return for the guaranteed wage. Of course, you also bear all the risk if your client cancels the contract at 24 hours' notice and you still have rent to pay, etc, so it's not for everyone. And at the risk of stating the obvious, when you make more money, you're also paying more tax on the income one way or another, so having highly skilled contractors commanding relatively high rates of pay a big win for government tax revenues overall.
As a final data point, just to put the "contractors outside PAYE are just dodging tax" meme to bed, a real tax accountant was interviewed by a major news outlet recently and asked to calculate how much tax an individual earning a fairly high income of (IIRC) £200,000 per year would pay as an employee vs. as a freelancer using typical alternative arrangements. It turned out that this person would actually pay slightly more in tax than their employed counterpart, and the whole thing is only a net loss to the taxman because the employer isn't paying NI either. So, when the BBC or government departments or whoever got caught paying highly compensated people via these arrangements, it wasn't actually the (non-)employees who were benefitting.
I don't accept your premise that businesses don't tolerate ethics that aren't mandated by law. There is nothing in the law in any major Western jurisdiction I'm familiar with that compels the executive management of a business to maximise profit above all else, and numerous big businesses do support things they don't strictly have to with real money. Probably they do it out of "enlightened self-interest", but that might just be making life better for their employees, who in turn are likely to be more productive, or developing good relations with their neighbours, because often what goes around comes around.
As for the ethics of making such a huge voluntary contribution to government coffers, I suppose that depends on whether you think the government will spend that money wisely. Given that the governments of the West haven't exactly got a stellar record when it comes to spending money as their respective populations would like, any argument that a company should pay more taxes merely because it's somehow "the right thing to do" is fairly weak, IMHO.
With the kind of money we're talking about here, you might get more benefit for both the company and the population as a whole if you funded bursaries for bright kids from poor backgrounds to go to university and study STEM subjects, or if you funded new leisure facilities/transport infrastructure/whatever in the city where your employees live, or if you paid a bigger dividend to shareholders, many of which might be pension funds, which in turn are going to be worth money to those retiring so the older generations can keep their heating on this winter. Amazingly, it turns out that all those big public company profits don't just disappear into some black hole or their directors' pockets, they actually go to shareholders, which includes people like you if you have a pension plan or any other investment based on stocks and shares.
What they are highlighting is not the fact that is illegal (it's not), but rather that it's unfair, which it is.
All taxation is unfair. Taxation is, essentially, legalised theft. It is, arguably, a necessary evil for the effective functioning of society at least until we develop better ways to collaborate in our common interest, but at the end of the day you're still taking away someone's money whether they like it or not.
And so, when people talk about a tax system being "unfair", what they usually mean is that they are forced to pay more than they want to and they think someone else should be forced to pay more so they can pay less. And when politicians talk about a tax system being "fair", what they usually mean is that the wealth is redistributed in favour of those who might vote for them and away from those who never would.
Actually, from the expression on his face, I'm pretty sure he knew exactly what I was talking about. But now I'm wondering whether he also heard about the issue we're discussing here a few hours before I did...
As you say, it's trivially easy, and has a huge upside. It also has showed me that the big companies really aren't a problem
Tesco are the biggest spammers of any business I know in the UK. I fail to see how their actions are legal to start with, and they ignore any attempt to unsubscribe.
Ironically, I never actually bought anything from their download site, due to what seemed to be a bug where you could sign up with N characters of e-mail address but only provide N-1 to log in. My e-mail address was N characters long.
One of these days, I am actually going to go after them on principle. Companies like that should know better, and should be made into examples when they don't.
TicketMaster (or possibly venues they deal with to whom they have given my TM-specific e-mail address, I'm not sure) are another repeat offender who are big enough that they ought to know their legal obligations when it comes to use/abuse of e-mail.
Not only is your point entirely fair, but the GP post was optimistic anyway. If this device is network enabled for legitimate reasons (streaming catch-up TV services, say) but also phones home for firmware updates and/or permits installable apps that actually change the software running internally, that's going to be non-trivial to firewall against abuse even if you have some idea what you're doing.
Put me in the camp that wants a TV to be a TV, without including an ad hoc, informally-specified, bug-ridden, slow implementation of the half-breed bastard child of a PVR and its iPad mistress.
I don't remember seeing your earlier post the first time around, but coincidentally I was in the electronics store just yesterday, and I saw one of these Samsung TVs with the marketing junk covered with stuff about the integrated camera/mic. I actually joked with the guy from the store that Samsung had imported someone from north of the border who still thought 1984 was a reference manual. And then today I log onto Slashdot and find this...
Even if there was blanket CCTV coverage available live to the police, they would only be using it to find and convict criminals.
Ah, yes, if we have nothing to hide then we have nothing to fear. Except incompetence or malice, that is.
I admire your optimism, but having personally been on the wrong side of a government screw-up involving mistaken identity (in my case, tax-related rather than criminal), I can assure you that they do make mistakes. Moreover, I can also testify that even if your life is being turned upside down as a result, and even if the situation described by their collective databases is clearly absurd and the records are obviously contradictory, there might still be no-one on the other side who gets informed automatically or is even in a position to help you when contacted directly.
I can also tell you from personal experience that although CCTV might theoretically be monitoring an area, whether anyone in authority can be bothered to look it up even if you're reporting a significant theft is far from certain. And of course if there's any hint that the authorities themselves might have done anything wrong, the cameras were magically all down for maintenance. So CCTV is of limited value to individuals who are the victims of crime or abuse. Moreover, evidence about CCTV reducing crime levels often says on page 2 that the crime levels in neighbouring areas without CCTV increased by a similar amount, so the benefits as a deterrent are questionable as well.
If you're not guilty of anything, then society has a far worse problem than the physical means that are used to fit you up.
Yes it does. Our justice system is deeply flawed in many ways.
But that's life, to some extent. There are always elements of "necessary evil" in government and there are always balances to be struck between the need to penalise the guilty effectively and the need to protect the innocent from becoming collateral damage. The number of people required to get everything in government right and the amount of money required to fund it is prohibitive, assuming you could ever find enough willing and able people to achieve it.
So, while in principle I would be happy to see a constitution-level law that says any government department or agent with any kind of statutory powers must also provide a fast and effective means of correcting mistakes or forfeit the right to use those powers at all (up to and including things like revoking HMRC's right to collect taxes or revoking the police's right to arrest people), I recognise that such a black/white system is extremely unlikely to be practical. The negative consequences of losing the good work would probably outweigh the benefits of preventing bad work. Instead, I prefer to limit the powers the government has to those it absolutely requires and the specific parts of government that absolutely require them, and in particular to restrict or abolish any powers that would be easy to abuse whether deliberately or inadvertently unless there is a clearly demonstrable benefit to having them anyway. I have yet to be convinced that such a clearly demonstrable benefit exists for almost anything involving words like security and surveillance.
No, it tells you that the voters in the referendum thought they'd prefer the current kind of Parliament more than the alternative they were being offered. You can't logically read any more into the result than that.
In fact, a significant part of the "no" campaign was arguing that if the electorate positively endorsed AV, that would be even worse than sticking with the previous arrangement, because then anyone who wanted any form of actual PR later would have to overcome the claims of "but the people voted for the system we've got".
I would agree that the advocates for change, particularly the Lib Dems, were totally politically outplayed by those who stood to gain from leaving things as they were. The end result might still be an equally solid "no longer on the table" response to any attempt at reform in the near future.
Just for the record, CCTV isn't nearly as simple an ethical issue as you're implying there. While the cameras installed a few years ago generate low quality imagery, modern ones can film you in glorious HD and full colour from a considerable distance. Moreover, facial recognition technology exists that could match you up against those handy computer-friendly photos you have to provide for passports and driving licences these days with a useful level of accuracy, meaning the authorities could literally establish a system to record your every move within areas covered by CCTV and stick it in one big database. (Don't think they'd do it? The national vehicle number plate scanning systems were installed by law enforcement quietly without ever really raising the issue in Parliament.)
Combine that with the really sinister stuff like snooping into conversations from a considerable distance or analysing people's movements (literally, the way they walk) to guess whether they have dubious intentions, and you've got all the key elements for a national thoughtcrime screening process. CCTV and related technologies are not your friend if you believe in civil liberties and the freedoms of individuals over the power of the state.
Obviously there are genuine concerns about security and crime prevention/detection that have to balance all of this. The difficult thing is figuring out what a reasonable balance is, given that technology is far ahead of the related ethics debates and any attempt to codify those ethics in actual laws.
For those not familiar with UK politics, I'll just point out here that some of the claims in the parent AC post are objectively wrong. For example, not all Lib Dem MPs reneged on the tuition fees commitment (the pledge mentioned by the parent poster).
Can I assume from this that you personally are appalled and angry when you buy a car but it doesn't come with detailed technical schematics, you buy a TV but they don't include the maintenance manual for engineers, or you buy a book but the original DTP files aren't thrown in? Would you rather have no car at all, no TV at all, no books at all?
There are things businesses do that go too far and we should be wary of supporting or encouraging them inadvertently. For me personally, the lines are often around things like invading privacy or installing malware-like DRM/product activation tools, which can (and sometimes have) actually cause bad consequences for me.
But if someone sells me software that does something useful to me, does it well, and doesn't do anything shady on the side, I frankly don't care whether I have the source code. I was never going to spend a whole load of my spare time learning my way around a complex system just to make some little change anyway, so I have lost nothing. And most of the general population wouldn't have the skills and knowledge to do that even if they wanted to, nor know how to reliably hire someone who did, so it's worth even less to them.
Last did anything extraordinary a few thousand years ago, but his followers still keep talking about it as if it was the pinnacle of human development? Erm... check.
I think this is the thing that bugs me more than anything else about RMS and the people who talk about how far Free Software has got us and how dangerous proprietary software is. Those of us who develop proprietary software are bringing many benefits to many people, despite all the potential pitfalls, and amazingly many of our users actually like us for it and are happy to enjoy some benefit they wouldn't otherwise have. Meanwhile, RMS is allegedly still reading the web via transcription onto stone carvings or something. If he symbolises anything, it's the total failure of the culture he wants to create to do better than the culture the rest of us live in.
The GPL is one of the most used software licences in the world because of its viral nature, getting there first, and being trendy for a while with young geeks. None of the above means those young geeks wouldn't have gone out and enthused and shared and built cool stuff if the GPL wasn't around.
That's what we did in my day. We called it freeware, and if someone asked if they could see how you did something in your code then you probably took it as a compliment and sent them the file, and no-one gave a damn about copyright on their toy code or plastering boilerplate legalese all over source files.
Plenty of people still do it. The Internet is full of software in online repos/code sharing sites that was presumably put there for anyone to use but doesn't even have a licence attached, just as people share their graphics demos or upload videos of them playing music to YouTube or post stuff on Facebook/Twitter/their blogs.
I don't really believe that RMS's philosophy or the ethical goals of the FSF are shared by many people in the grand scheme of things. A lot of people use OSS because it's free as in beer and legal, and a lot of people contribute to it because they're happy to give their code away and someone gave them some code with that LICENCE file thing once so they do the same.
FWIW, it sounds like we agree on what should happen. Local councillors and council workers having powers normally reserved to organisations like the police, security services and HMRC makes me nervous for a lot of reasons... starting with the fact that I've actually met some of those council people, and of course the numerous well-documented cases where councils have abused privileged access to sensitive data or otherwise blatantly overstepped their authority.
Although in practicality, it is probably comparable to the C-style error handling overhead that Darinbob mentioned. People forget that error handling in C is not free.
Exactly. The point isn't whether using an exception magically requires no code, it's whether whatever the exception requires is different to what you'd have to do anyway if you manually implemented error codes and clean-up logic to get the same behaviour.
For what little it's worth, I do sympathise with your argument. Local authorities in the UK tend to have far more powers than I would like and/or to have far more flexibility in how to implement things than seems sensible when they're really only representing a tiny fraction of the population as a whole yet could affect many other people as well. Thus we get everything from factually unsupported 20mph limits outside all the posh houses (because the local councillor, who doesn't drive anyway, is in reality elected by a handful of rich people and pushes it through, even though almost everyone in the city and beyond uses that road) to postcode lotteries for just about anything under local authority control: medical treatments, schooling, you name it.
I also sympathise with disproportionate punishments for tiny things that are genuine mistakes anyone could have made. I don't think this sort of behaviour makes the world a better place for anyone, and I very much doubt it even makes a significant amount of money for the people doing it.
But I'm not sure either of these points really contradicts the original discussion, where I was noting that, as a matter of fact and for better or worse, it is not the police who do the kind of enforcement we were talking about in much of the UK today.
If your argument is that exceptions are relatively slow if you're using a platform where the implementation of exceptions is relatively slow, then I can't really argue with that!:-p
But the original comment I replied to was making a blanket statement about how to use exceptions, and was talking specifically about C++: "Exceptions are usually quite expensive, especially in C++ compared to just returning an error code." That simply isn't true, for the reasons I explained before.
Sorry, I'm not quite sure what point you're trying to make here. If you're arguing that sometimes an exception-like mechanism can be useful for returning control in expected cases as well as unexpected ones, then I agree with you, as long as the mechanism in the particular language/environment you're using is suitable of course.
Exception handling in quite a lot of implementations involves overhead in every function entry and exit.
Really? If that's true, then either those implementations are decades behind the state of the art or they are doing something more than merely propagating a different kind of result in the way we've been talking about for most of this Slashdot discussion.
Any given routine can only ever succeed or fail. When it returns control, either it hands back the result and/or causes the side effects you asked for, or it doesn't.
I'm not sure what realistic scenarios you had in mind when describing those other cases, but they sound like simply applying the same principles at a lower level to me: something causes a low level function to complete, and a higher level function detects that and then carries on with whatever behaviour is appropriate for the overall design/algorithm being implemented. If the higher level code can recover gracefully from any problems lower down, the overall algorithm might still succeed. If the higher level code can't continue its work once a lower level component has failed, at some point the higher level function is going to indicate failure as well.
As far as I know, C++ compilers have gone in the direction of making exceptions free at the expense of high cost if they are thrown.
Not necessarily. Several major C++ compilers implement exceptions using some variation of jump tables, where the exceptional paths are effectively precomputed. Unless you start doing tricky things like calling functions indirectly via pointers, where obviously you can't precompute everything as effectively, there is negligible (time) cost even if the exception is thrown. Effectively, you just call any relevant destructors that would have been encountered as you unwound the stack (each of which you would necessarily also have run even if you had returned from each function one by one) and then jump directly to the nearest matching handler. The main trade-off is that a lot of space may be used to hold those jump tables.
(This isn't necessarily true of implementations of similar exception features in other languages/environments, or even of all C++ compilers on all platforms, but it does demonstrate what is possible.)
Those are problems (if they're problems at all) only because Microsoft's particular managed environment imposes all kinds of overheads and mixes up hardware/OS/managed runtime functionality with a language feature. Not one of those examples is a fundamental or essential problem with the concept of exceptions.
Grab a stack trace by interpreting metadata emitted by the compiler to guide our stack unwind.
That can be done with simple table look-ups, a negligible overhead for something we're only running once where we're probably already talking about unwinding several stack frames and running tidy up code along the way. As I mentioned before, I have spent significant time looking at and profiling generated code in this area, and I'm telling you that jump tables as used in many even slightly modern C++ compilers do this faster than a return code approach can unwind a stack one frame at a time using return codes.
Run through a chain of handlers up the stack, calling each handler twice.
That sounds like an architectural compromise made by that specific managed environment. Presumably there is some reason in.Net that calling each handler twice is necessary, but in general it isn't. And since the only thing handlers generally need to do until you get to the final one is tidy up side effects by freeing resources and the like, which you also need to do if you're unwinding manually, there's no additional overhead there at all.
Compensate for mismatches between SEH, C++ and managed exceptions.
Again, that's an overhead if you're running in Microsoft's managed environment, not a problem with exceptions as such. You're talking about a non-standard behaviour projecting OS-level functionality (SEH) onto a language level feature (C++ exceptions), which is apples to oranges.
Allocate a managed Exception instance and run its constructor. Most likely, this involves looking up resources for the various error messages.
As opposed to either not returning the useful information at all (if you just dump a trivial return code on the caller) or constructing an object with more information to pass back by some other means (which will incur essentially the same overheads to build it, set up error messages, etc.). Again, there's no fundamental overhead implied by exceptions here. It's the same work you'd do to return the same information using return codes anyway.
Probably take a trip through the OS kernel. Often take a hardware exception.
Again, this is conflating hardware/OS level functions with exceptions as a tool in a programming language. Just because Microsoft project one onto the other, that doesn't mean everyone else does or wants to.
Notify any attached debuggers, profilers, vectored exception handlers and other interested parties.
So now I'm not allowed to use the best tool for the job if it's inconvenient for people writing developer tools? I put it to you that in this case, the development tools have the problem. And in any case, presumably this overhead only applies if any such development tools are actually in use, so in production code it's a no-op.
This is light years away from returning a -1 from your function call.
It's also light years away from what most people mean when they talk about exceptions as a language feature.
Exceptions are inherently non-local, and if there's an obvious and enduring trend for today's architectures, it's that you must remain local for good performance.
Programming any non-trivial system is inherently non-local, particularly if you use things like modular designs and functional decomposition. You can use exceptions, which dovetail neatly with that approach, or you can throw the logic into your big ball of mud architecture, and not fit in the cache anyway.
Slashdot Mirror
As I acknowledged, taxation may be a necessary evil, at least for now. But there is no logical basis for assuming it must always be so.
Consider that whether you call it patronage, charity, philanthropy, or something else, many people from the minimum wage hired help to billionaires already give a substantial chunk of their personal income to causes they consider worthwhile. I personally believe that most people are decent, and understand that some public/common services are in everyone's interests, and would be willing to contribute to society without being forced given a reasonable system for doing so.
Moreover, many people (sometimes whole countries) pay privately for services that other people (sometimes whole countries) consider essential public services. So demonstrably a lot of things that "big government" does don't necessarily have to be funded from the public purse.
And crucially, the modern world has technology and communication tools that could allow people to collaborate and direct their money to what they considered worthwhile causes on a scale never before seen in human history, without imposing a heavy burden on anyone who was willing to contribute but didn't have time to learn about the different possibilities in detail. We have the means to build a system where people could be involved in directing what their money supported to some extent if they wanted to, while still allowing others to choose a sensible "default" distribution for any donations as recommended by the government of the day.
I don't see us getting away from taxation any time in the near future, as it would require a huge cultural shift to make the kind of system I described above a success. However, I see no reason in principle that we couldn't have an alternative system where everyone understood that key services needed to be supported and gave what they could/what the services needed to do that. This would also have the useful side effect of making it very obvious when an important service was under-funded so it could attract more support, which is something that doesn't always happen today because under-funded public services appear to be "someone else's problem" if the expectation is that they are supported via tax revenues.
Speaking as someone in the UK who does run his own business and therefore can do those things you described, I would like to point out that the situation is not nearly as one-sided as you are suggesting.
For a start, you can only do those things if you are not considered an employee. You can't just decide to call yourself a contractor and magically opt out of the normal tax/NI system when you're still doing a regular job for an employer. People used to try and do this, but with the introduction of the IR35 rules a decade or so ago, HMRC (the tax authority in the UK) can go after people who are merely "disguised employees" to get the missing tax income back.
Moreover, for those of us who really are running our own businesses rather than being someone else's employee, anything we save through not making the same payments as a typical full-time employee is usually far outweighed by what we lose because no-one pays us for time off, we have to provide our own equipment and training, we're responsible for paying all the basic bills like heating/lighting/phone/Internet, and so on.
There are certainly some great perks to running your own small business against being an employee. In a very real sense, you get to control your own destiny. You can be much more flexible with when and how you work, and fit paying the rent around your lifestyle rather than the other way around. I think our education and employment systems don't do enough to promote alternative careers to those who might enjoy and benefit from them.
But mysteriously saving a fortune in tax isn't a good reason to go down this path, and if that's the only reason someone is doing it then they're going to be sadly disappointed. Contractors often do make more money than their employed counterparts, but contrary to the impression given by certain mainstream media outlets, the real financial benefits don't come from dodging a bit of NI, they come from having much more flexibility to negotiate what you're paid based on the value you actually generate for customers/clients rather than being stuck with a market value salary where your employer keeps most of the profits in return for the guaranteed wage. Of course, you also bear all the risk if your client cancels the contract at 24 hours' notice and you still have rent to pay, etc, so it's not for everyone. And at the risk of stating the obvious, when you make more money, you're also paying more tax on the income one way or another, so having highly skilled contractors commanding relatively high rates of pay a big win for government tax revenues overall.
As a final data point, just to put the "contractors outside PAYE are just dodging tax" meme to bed, a real tax accountant was interviewed by a major news outlet recently and asked to calculate how much tax an individual earning a fairly high income of (IIRC) £200,000 per year would pay as an employee vs. as a freelancer using typical alternative arrangements. It turned out that this person would actually pay slightly more in tax than their employed counterpart, and the whole thing is only a net loss to the taxman because the employer isn't paying NI either. So, when the BBC or government departments or whoever got caught paying highly compensated people via these arrangements, it wasn't actually the (non-)employees who were benefitting.
I don't accept your premise that businesses don't tolerate ethics that aren't mandated by law. There is nothing in the law in any major Western jurisdiction I'm familiar with that compels the executive management of a business to maximise profit above all else, and numerous big businesses do support things they don't strictly have to with real money. Probably they do it out of "enlightened self-interest", but that might just be making life better for their employees, who in turn are likely to be more productive, or developing good relations with their neighbours, because often what goes around comes around.
As for the ethics of making such a huge voluntary contribution to government coffers, I suppose that depends on whether you think the government will spend that money wisely. Given that the governments of the West haven't exactly got a stellar record when it comes to spending money as their respective populations would like, any argument that a company should pay more taxes merely because it's somehow "the right thing to do" is fairly weak, IMHO.
With the kind of money we're talking about here, you might get more benefit for both the company and the population as a whole if you funded bursaries for bright kids from poor backgrounds to go to university and study STEM subjects, or if you funded new leisure facilities/transport infrastructure/whatever in the city where your employees live, or if you paid a bigger dividend to shareholders, many of which might be pension funds, which in turn are going to be worth money to those retiring so the older generations can keep their heating on this winter. Amazingly, it turns out that all those big public company profits don't just disappear into some black hole or their directors' pockets, they actually go to shareholders, which includes people like you if you have a pension plan or any other investment based on stocks and shares.
What they are highlighting is not the fact that is illegal (it's not), but rather that it's unfair, which it is.
All taxation is unfair. Taxation is, essentially, legalised theft. It is, arguably, a necessary evil for the effective functioning of society at least until we develop better ways to collaborate in our common interest, but at the end of the day you're still taking away someone's money whether they like it or not.
And so, when people talk about a tax system being "unfair", what they usually mean is that they are forced to pay more than they want to and they think someone else should be forced to pay more so they can pay less. And when politicians talk about a tax system being "fair", what they usually mean is that the wealth is redistributed in favour of those who might vote for them and away from those who never would.
Actually, from the expression on his face, I'm pretty sure he knew exactly what I was talking about. But now I'm wondering whether he also heard about the issue we're discussing here a few hours before I did...
As you say, it's trivially easy, and has a huge upside. It also has showed me that the big companies really aren't a problem
Tesco are the biggest spammers of any business I know in the UK. I fail to see how their actions are legal to start with, and they ignore any attempt to unsubscribe.
Ironically, I never actually bought anything from their download site, due to what seemed to be a bug where you could sign up with N characters of e-mail address but only provide N-1 to log in. My e-mail address was N characters long.
One of these days, I am actually going to go after them on principle. Companies like that should know better, and should be made into examples when they don't.
TicketMaster (or possibly venues they deal with to whom they have given my TM-specific e-mail address, I'm not sure) are another repeat offender who are big enough that they ought to know their legal obligations when it comes to use/abuse of e-mail.
Not only is your point entirely fair, but the GP post was optimistic anyway. If this device is network enabled for legitimate reasons (streaming catch-up TV services, say) but also phones home for firmware updates and/or permits installable apps that actually change the software running internally, that's going to be non-trivial to firewall against abuse even if you have some idea what you're doing.
Put me in the camp that wants a TV to be a TV, without including an ad hoc, informally-specified, bug-ridden, slow implementation of the half-breed bastard child of a PVR and its iPad mistress.
I don't remember seeing your earlier post the first time around, but coincidentally I was in the electronics store just yesterday, and I saw one of these Samsung TVs with the marketing junk covered with stuff about the integrated camera/mic. I actually joked with the guy from the store that Samsung had imported someone from north of the border who still thought 1984 was a reference manual. And then today I log onto Slashdot and find this...
Even if there was blanket CCTV coverage available live to the police, they would only be using it to find and convict criminals.
Ah, yes, if we have nothing to hide then we have nothing to fear. Except incompetence or malice, that is.
I admire your optimism, but having personally been on the wrong side of a government screw-up involving mistaken identity (in my case, tax-related rather than criminal), I can assure you that they do make mistakes. Moreover, I can also testify that even if your life is being turned upside down as a result, and even if the situation described by their collective databases is clearly absurd and the records are obviously contradictory, there might still be no-one on the other side who gets informed automatically or is even in a position to help you when contacted directly.
I can also tell you from personal experience that although CCTV might theoretically be monitoring an area, whether anyone in authority can be bothered to look it up even if you're reporting a significant theft is far from certain. And of course if there's any hint that the authorities themselves might have done anything wrong, the cameras were magically all down for maintenance. So CCTV is of limited value to individuals who are the victims of crime or abuse. Moreover, evidence about CCTV reducing crime levels often says on page 2 that the crime levels in neighbouring areas without CCTV increased by a similar amount, so the benefits as a deterrent are questionable as well.
If you're not guilty of anything, then society has a far worse problem than the physical means that are used to fit you up.
Yes it does. Our justice system is deeply flawed in many ways.
But that's life, to some extent. There are always elements of "necessary evil" in government and there are always balances to be struck between the need to penalise the guilty effectively and the need to protect the innocent from becoming collateral damage. The number of people required to get everything in government right and the amount of money required to fund it is prohibitive, assuming you could ever find enough willing and able people to achieve it.
So, while in principle I would be happy to see a constitution-level law that says any government department or agent with any kind of statutory powers must also provide a fast and effective means of correcting mistakes or forfeit the right to use those powers at all (up to and including things like revoking HMRC's right to collect taxes or revoking the police's right to arrest people), I recognise that such a black/white system is extremely unlikely to be practical. The negative consequences of losing the good work would probably outweigh the benefits of preventing bad work. Instead, I prefer to limit the powers the government has to those it absolutely requires and the specific parts of government that absolutely require them, and in particular to restrict or abolish any powers that would be easy to abuse whether deliberately or inadvertently unless there is a clearly demonstrable benefit to having them anyway. I have yet to be convinced that such a clearly demonstrable benefit exists for almost anything involving words like security and surveillance.
Which doesn't matter if there is still at least one watching you that isn't.
No, it tells you that the voters in the referendum thought they'd prefer the current kind of Parliament more than the alternative they were being offered. You can't logically read any more into the result than that.
In fact, a significant part of the "no" campaign was arguing that if the electorate positively endorsed AV, that would be even worse than sticking with the previous arrangement, because then anyone who wanted any form of actual PR later would have to overcome the claims of "but the people voted for the system we've got".
I would agree that the advocates for change, particularly the Lib Dems, were totally politically outplayed by those who stood to gain from leaving things as they were. The end result might still be an equally solid "no longer on the table" response to any attempt at reform in the near future.
Just for the record, CCTV isn't nearly as simple an ethical issue as you're implying there. While the cameras installed a few years ago generate low quality imagery, modern ones can film you in glorious HD and full colour from a considerable distance. Moreover, facial recognition technology exists that could match you up against those handy computer-friendly photos you have to provide for passports and driving licences these days with a useful level of accuracy, meaning the authorities could literally establish a system to record your every move within areas covered by CCTV and stick it in one big database. (Don't think they'd do it? The national vehicle number plate scanning systems were installed by law enforcement quietly without ever really raising the issue in Parliament.)
Combine that with the really sinister stuff like snooping into conversations from a considerable distance or analysing people's movements (literally, the way they walk) to guess whether they have dubious intentions, and you've got all the key elements for a national thoughtcrime screening process. CCTV and related technologies are not your friend if you believe in civil liberties and the freedoms of individuals over the power of the state.
Obviously there are genuine concerns about security and crime prevention/detection that have to balance all of this. The difficult thing is figuring out what a reasonable balance is, given that technology is far ahead of the related ethics debates and any attempt to codify those ethics in actual laws.
For those not familiar with UK politics, I'll just point out here that some of the claims in the parent AC post are objectively wrong. For example, not all Lib Dem MPs reneged on the tuition fees commitment (the pledge mentioned by the parent poster).
And the apology was funny...
Can I assume from this that you personally are appalled and angry when you buy a car but it doesn't come with detailed technical schematics, you buy a TV but they don't include the maintenance manual for engineers, or you buy a book but the original DTP files aren't thrown in? Would you rather have no car at all, no TV at all, no books at all?
There are things businesses do that go too far and we should be wary of supporting or encouraging them inadvertently. For me personally, the lines are often around things like invading privacy or installing malware-like DRM/product activation tools, which can (and sometimes have) actually cause bad consequences for me.
But if someone sells me software that does something useful to me, does it well, and doesn't do anything shady on the side, I frankly don't care whether I have the source code. I was never going to spend a whole load of my spare time learning my way around a complex system just to make some little change anyway, so I have lost nothing. And most of the general population wouldn't have the skills and knowledge to do that even if they wanted to, nor know how to reliably hire someone who did, so it's worth even less to them.
Last did anything extraordinary a few thousand years ago, but his followers still keep talking about it as if it was the pinnacle of human development? Erm... check.
I think this is the thing that bugs me more than anything else about RMS and the people who talk about how far Free Software has got us and how dangerous proprietary software is. Those of us who develop proprietary software are bringing many benefits to many people, despite all the potential pitfalls, and amazingly many of our users actually like us for it and are happy to enjoy some benefit they wouldn't otherwise have. Meanwhile, RMS is allegedly still reading the web via transcription onto stone carvings or something. If he symbolises anything, it's the total failure of the culture he wants to create to do better than the culture the rest of us live in.
The GPL is one of the most used software licences in the world because of its viral nature, getting there first, and being trendy for a while with young geeks. None of the above means those young geeks wouldn't have gone out and enthused and shared and built cool stuff if the GPL wasn't around.
That's what we did in my day. We called it freeware, and if someone asked if they could see how you did something in your code then you probably took it as a compliment and sent them the file, and no-one gave a damn about copyright on their toy code or plastering boilerplate legalese all over source files.
Plenty of people still do it. The Internet is full of software in online repos/code sharing sites that was presumably put there for anyone to use but doesn't even have a licence attached, just as people share their graphics demos or upload videos of them playing music to YouTube or post stuff on Facebook/Twitter/their blogs.
I don't really believe that RMS's philosophy or the ethical goals of the FSF are shared by many people in the grand scheme of things. A lot of people use OSS because it's free as in beer and legal, and a lot of people contribute to it because they're happy to give their code away and someone gave them some code with that LICENCE file thing once so they do the same.
FWIW, it sounds like we agree on what should happen. Local councillors and council workers having powers normally reserved to organisations like the police, security services and HMRC makes me nervous for a lot of reasons... starting with the fact that I've actually met some of those council people, and of course the numerous well-documented cases where councils have abused privileged access to sensitive data or otherwise blatantly overstepped their authority.
Although in practicality, it is probably comparable to the C-style error handling overhead that Darinbob mentioned. People forget that error handling in C is not free.
Exactly. The point isn't whether using an exception magically requires no code, it's whether whatever the exception requires is different to what you'd have to do anyway if you manually implemented error codes and clean-up logic to get the same behaviour.
For what little it's worth, I do sympathise with your argument. Local authorities in the UK tend to have far more powers than I would like and/or to have far more flexibility in how to implement things than seems sensible when they're really only representing a tiny fraction of the population as a whole yet could affect many other people as well. Thus we get everything from factually unsupported 20mph limits outside all the posh houses (because the local councillor, who doesn't drive anyway, is in reality elected by a handful of rich people and pushes it through, even though almost everyone in the city and beyond uses that road) to postcode lotteries for just about anything under local authority control: medical treatments, schooling, you name it.
I also sympathise with disproportionate punishments for tiny things that are genuine mistakes anyone could have made. I don't think this sort of behaviour makes the world a better place for anyone, and I very much doubt it even makes a significant amount of money for the people doing it.
But I'm not sure either of these points really contradicts the original discussion, where I was noting that, as a matter of fact and for better or worse, it is not the police who do the kind of enforcement we were talking about in much of the UK today.
If your argument is that exceptions are relatively slow if you're using a platform where the implementation of exceptions is relatively slow, then I can't really argue with that! :-p
But the original comment I replied to was making a blanket statement about how to use exceptions, and was talking specifically about C++: "Exceptions are usually quite expensive, especially in C++ compared to just returning an error code." That simply isn't true, for the reasons I explained before.
Sorry, I'm not quite sure what point you're trying to make here. If you're arguing that sometimes an exception-like mechanism can be useful for returning control in expected cases as well as unexpected ones, then I agree with you, as long as the mechanism in the particular language/environment you're using is suitable of course.
Exception handling in quite a lot of implementations involves overhead in every function entry and exit.
Really? If that's true, then either those implementations are decades behind the state of the art or they are doing something more than merely propagating a different kind of result in the way we've been talking about for most of this Slashdot discussion.
Any given routine can only ever succeed or fail. When it returns control, either it hands back the result and/or causes the side effects you asked for, or it doesn't.
I'm not sure what realistic scenarios you had in mind when describing those other cases, but they sound like simply applying the same principles at a lower level to me: something causes a low level function to complete, and a higher level function detects that and then carries on with whatever behaviour is appropriate for the overall design/algorithm being implemented. If the higher level code can recover gracefully from any problems lower down, the overall algorithm might still succeed. If the higher level code can't continue its work once a lower level component has failed, at some point the higher level function is going to indicate failure as well.
As far as I know, C++ compilers have gone in the direction of making exceptions free at the expense of high cost if they are thrown.
Not necessarily. Several major C++ compilers implement exceptions using some variation of jump tables, where the exceptional paths are effectively precomputed. Unless you start doing tricky things like calling functions indirectly via pointers, where obviously you can't precompute everything as effectively, there is negligible (time) cost even if the exception is thrown. Effectively, you just call any relevant destructors that would have been encountered as you unwound the stack (each of which you would necessarily also have run even if you had returned from each function one by one) and then jump directly to the nearest matching handler. The main trade-off is that a lot of space may be used to hold those jump tables.
(This isn't necessarily true of implementations of similar exception features in other languages/environments, or even of all C++ compilers on all platforms, but it does demonstrate what is possible.)
Those are problems (if they're problems at all) only because Microsoft's particular managed environment imposes all kinds of overheads and mixes up hardware/OS/managed runtime functionality with a language feature. Not one of those examples is a fundamental or essential problem with the concept of exceptions.
Grab a stack trace by interpreting metadata emitted by the compiler to guide our stack unwind.
That can be done with simple table look-ups, a negligible overhead for something we're only running once where we're probably already talking about unwinding several stack frames and running tidy up code along the way. As I mentioned before, I have spent significant time looking at and profiling generated code in this area, and I'm telling you that jump tables as used in many even slightly modern C++ compilers do this faster than a return code approach can unwind a stack one frame at a time using return codes.
Run through a chain of handlers up the stack, calling each handler twice.
That sounds like an architectural compromise made by that specific managed environment. Presumably there is some reason in .Net that calling each handler twice is necessary, but in general it isn't. And since the only thing handlers generally need to do until you get to the final one is tidy up side effects by freeing resources and the like, which you also need to do if you're unwinding manually, there's no additional overhead there at all.
Compensate for mismatches between SEH, C++ and managed exceptions.
Again, that's an overhead if you're running in Microsoft's managed environment, not a problem with exceptions as such. You're talking about a non-standard behaviour projecting OS-level functionality (SEH) onto a language level feature (C++ exceptions), which is apples to oranges.
Allocate a managed Exception instance and run its constructor. Most likely, this involves looking up resources for the various error messages.
As opposed to either not returning the useful information at all (if you just dump a trivial return code on the caller) or constructing an object with more information to pass back by some other means (which will incur essentially the same overheads to build it, set up error messages, etc.). Again, there's no fundamental overhead implied by exceptions here. It's the same work you'd do to return the same information using return codes anyway.
Probably take a trip through the OS kernel. Often take a hardware exception.
Again, this is conflating hardware/OS level functions with exceptions as a tool in a programming language. Just because Microsoft project one onto the other, that doesn't mean everyone else does or wants to.
Notify any attached debuggers, profilers, vectored exception handlers and other interested parties.
So now I'm not allowed to use the best tool for the job if it's inconvenient for people writing developer tools? I put it to you that in this case, the development tools have the problem. And in any case, presumably this overhead only applies if any such development tools are actually in use, so in production code it's a no-op.
This is light years away from returning a -1 from your function call.
It's also light years away from what most people mean when they talk about exceptions as a language feature.
Exceptions are inherently non-local, and if there's an obvious and enduring trend for today's architectures, it's that you must remain local for good performance.
Programming any non-trivial system is inherently non-local, particularly if you use things like modular designs and functional decomposition. You can use exceptions, which dovetail neatly with that approach, or you can throw the logic into your big ball of mud architecture, and not fit in the cache anyway.