Maybe they'd wait if, I dunno, we didn't advertise the details of suspects publicly on the basis of some random allegation that has yet to be proven in court?
Getting back to this specific case, I'm not sure what's more disturbing:
the fact that the guy lost his reputation and his livelihood on the basis of a tenuous link that wrongly affects thousands of people every year, or
the fact that he could build a solid case to refute the charge using only a fairly simple FOI request and matching up the time and place of the criminal use of the card with records proving he was on the other side of the planet at the time, yet the authorities managed to take his computer equipment and such away for several months and take the system took several years to exhonerate him.
And on an unrelated note, what gives with the weird styling on Slashdot since earlier this evening? Loads of HTML formatting, such as the list above, is completely broken.:-(
Sorry, friend, but I think you've misunderstood what ISO 9000 is all about.
It doesn't require you to have a good process, only that you have some process and actually follow it.
Of course, in my business (software development), the only reason to get formal certification for things like ISO 9000 or TickIT compliance is as a means to get contracts with clients (typically governments... surprise!) who insist on it. If anyone thinks it's actually about improving the productivity of businesses or quality of the products they make, consider that almost every practical improvement in modern management of software development processes has moved away from heavyweight formal processes that do everything step-by-step and toward more lightweight, iterative, interactive approaches in one way or another. You don't necessarily have to buy into the trendy things with words like "extreme" and "agile" associated with them, but seriously, the waterfall model died decades ago in everywhere but ISO land.
We already pay a fair rate for the bandwidth we use.
No, you don't. This is the problem.
It's not your fault. You're just paying what the ISP is currently asking, and expecting what they offer to be provided in return.
But in any flat-rate pricing scheme like this, there are winners and losers. For everyone who's above the middle mark, they're getting a good deal. There's nothing unethical about taking advantage of that when it's offered. But it's not a fair rate, because those who are below the middle mark are subsidising those above it.
Even so, there's one, important thing you're missing that the traditional companies do: they provide all the money needed to assemble, print, distribute and market your book, then give you a percentage of what comes in.
But the actual monetary cost of assembly and printing via print-on-demand is close enough to zero that it makes little difference. The major expense is time, and if you've got the time to write a whole book worth publishing, you're obviously not too worried about that. And as I noted before, the distribution and marketing advantages of using a major publishing house are highly overrated if you're not writing the kind of book that's mass market and going to make it into bricks and mortar bookstores, while the percentage you get back is literally an order of magnitude less than you would via POD.
Going POD or self-publishing means that you have to pony up in advance, with no guarantee that you'll recover your investment, let alone make a profit.
I hate to break this to you, but going via an old school publisher is no guarantee you'll ever cover your advance or make a profit, either. In any case, it will take a lot more book sales before you do...
I studied CS in academia and worked in industry for a while, and I don't know what the "white book" is either. I guess I'd better give up and become a lawyer.:-(
Or, just maybe, CS education is about more than name dropping...
Isn't C still part of a basic university computer science education?
Sadly, no. Even in the better universities, it is often covered only briefly, and it's just expected that any self-respecting CS student will learn it. Of course, while any self-respecting CS student should be gaining exposure to a variety of programming languages and tools, part of the point of the formal academic course is to introduce those students to the ideas they need to be able to do this, which is why I think the courses should include some reasonably deep instruction in example languages of various types before assuming the students will go broaden their horizons independently. If the courses won't do that, what are they actually teaching that someone couldn't just learn on their own?
IMHO, C actually has far more place in such courses than C++, which while a useful practical language isn't really an optimal example of anything for teaching purposes (apart from perhaps how to combine different programming ideas to produce a practical language).
Now, who'll be the first one to spot and explain the problem here, preferrably without actually running it?
Probably your static code analysis tool, which would presumably warn you about the potential division by zero, and for that matter the fact that you're dividing using integer arithmetic in the first place.
But neither intrinsic functions or inline assembler are portable, which is unfortunate if (like many high performance libraries) you need to build on many platforms.
I think the more relevant thing is that if you use a decent compiler, most of them will pick up a simple optimisation like this and generate appropriate code just from the C++ these days. Optimisers certainly have a long way to go, but we're well past the point of getting this sort of thing right routinely.
You seem to have this wonderful illusion that going via a real publisher is some sort of guarantee of decent quality. It certainly isn't in programming books: take a look at some more recent titles by Addison-Wesley, previously the source of the majority of good C++ books, and it's hard not to find typos all over the place, amateurish design and typography, and a general lack of editorial quality. Several other professional publishers were never much better than that. And if you think going for a reputable university press would be better than the big professional outfits, guess again: take a look at the guidelines for authors published by some of them (you can often find these on their web sites) and they tend to insist on petty things like conforming to some arbitrary (or sometimes, IMHO, outright incorrect) house styles that do nothing to improve the quality of a work written by a skilled and knowledgable author, before they'll even grace you with the benefits of their immense knowledge and experience (and yes, that was sarcasm: do you realise who these "professional editors" typically are?).
Not so long ago, I considered writing a book, after receiving some favourable comments on a couple of smaller pieces of work I'd done. I looked into what would be involved in going through a mainstream publisher, and came away asking why anyone with decent writing ability and decent knowledge of the presentation side of things would ever use one. I always knew authors didn't normally receive a high percentage of the cover price of a book, but I was shocked at just how little it really is: well under 10% seems pretty typical, and it varies depending on the market. Anyone with some basic knowledge of subjects like book layout and typography could produce a better design than many of the "professionals" do without even trying, using any good DTP package (or even LaTeX, for technical books). I have some contacts at some university publishers, and some of the comments I hear about their editorial teams are just appalling. And it's not even like being published by a reputable press will get you into bricks and mortar stores any more: if I walk into my local Borders, it's full of "computing for fools" and "learn computing in ten seconds" books, but even relatively mainstream "serious" books are in short supply these days and I usually have to look on-line for them. There are almost no more specialised books of the kind you'd typically find from an academic press.
So, if people are going to have to order in or buy from the big on-line vendors to get any serious technical book, and publishers add precious little value in the editorial and presentation departments if you're likely to be as competent as their staff anyway, what real advantage do they bring? You can jump through a few hoops to get things like an ISBN and use a reputable POD organisation with links to a major distribution channel, and then you can have exactly the book you want available through all the same (realistic) channels anyway, but with at least 10x the profit margin on every book sold and with almost no financial liability if your book doesn't sell. Do professional publishers really produce and market books well enough to get an order of magnitude more sales for specialist titles than the self-publishing/POD route? Not if the data from people like O'Reilly and the anecdotes from published authors on the web are anything to go by, they don't.
And of course, just because you're self-publishing, that doesn't mean your quality has to suck. You might be a decent writer and designer yourself, compared to what you'd get from a typical publisher, and you can still bring in professional help for areas where you need it if you're not an expert. It's just done on your terms, with costs you know, and once you've paid your overheads you don't have to keep paying for them with every new book sold.
So tell me again, why would anyone with decent writing and presentation skills go to a professional publisher today, if they aren't writing the kind of mass-market fiction or technical books for idiots that are likely to sell with a really high volume and make it into real world bookshops?
Look, I'm all for keeping an eye out on Wikileaks. I think it serves a very important purpose in a time when a lot of governments - and their people - feel that the withholding of information is a good idea.
The thing is, sometimes withholding information from the general public is a good idea. Not all of the general public are good guys, and there are legitimate reasons for governments to do some things in secret, for a time. The practical problem is that the general public have no way to ascertain whether any information being withheld is being kept from them for legitimate reasons or less savoury purposes if they can't see it. However, there are pragmatic alternative approaches to dealing with this problem via independent oversight that do not involve full disclosure of everything as implied by Wikileaks.
Unless I missed something, the entire report is basically arguing for such a middle ground. I don't see anywhere it says we should throw children into dangerous situations they can't cope with. Rather, it seems (from a first quick scan at least) to be advocating throwing children into somewhat dangerous situations carefully so they can learn to handle them safely in their own right.
This sounds like the kind of common sense you'd get from someone who actually deals with children professionally and sorts out problems in real life. Oh, wait, she is.:-)
Sadly, I gather she's decided that her television programmes weren't necessarily in the interests of the children participating and discontinued them now. That's a pity; they were very informative and seemed to be done quite responsibly from a naive but interested observer's point of view.
From my experience, all you can eat plans have always been cheaper/more convenient to the customer than tiered rates.
They are also usually loss-leaders. If they carry on for long, there are only two likely outcomes: the company goes bust, or a substantial number of customers wind up getting screwed by paying for more than they need.
This is obviously going to be true of broadband ISPs before very long, as "unmetered" access becomes implausible with the rise of on-line multimedia distribution. The current hardware networks simply aren't up to providing that for everyone, so somebody somewhere is going to pay for the extra capacity, and you can bet it won't be the top-tier ISPs! Personally, since I use only a small fraction of my monthly capacity allowance but the next tier down is inadequate for me and most of the decent ISPs around here have similar models, I would be quite happy to go back to paying by the GB downloaded; that way, I wouldn't be paying extra to support all the heavy downloaders at the other end of the scale.
Don't prohibit DRM. That's a technical measure, and completely within the company's right to try to implement. What needs to happen is that the DMCA needs to get thrown out so that it's no longer illegal to break that DRM. Exercising fair-use rights should not be illegal.
I think I agree with the spirit of what you're saying, but I don't agree with the letter. Merely revoking the necessary parts of the DMCA does not help anyone who can't break the DRM technically, because they still can't make fair use of the material. Moreover, it assumes that someone will always be able to break the DRM and willing to share the technology they used to do so. While this is, AFAIK, true of all major DRM platforms today, it would be foolish to assume that it will always remain so.
I think for any fair DRM-related system to work, the law will need to start distinguishing more carefully between someone who owns a legitimate, permanent copy of a work subject to copyright, and someone who accepts an agreement when the material is supplied that restricts their use of it, for example by imposing a time limit.
In the former case, I believe fair use should be closer to a protected right than its current status of affirmative defence: it should be illegal for anyone to actively inhibit exercising fair use, for example by using DRM technology. Moreover, that particular copy should be treated like any physical property, such that the person who paid for it may lend it out to others or sell it on under the doctrine of first sale, provided they do not retain access to or further copies of it themselves at the same time, and there is no scope for imposing additional restrictions via some sort of licence agreement. In essence, as long as electronic material is not duplicated in a way that would be an infringement of copyright if you did the same thing with a book you bought from a store, it should be given the same status in law.
In the case where restrictions are understood and accepted before receiving a copy, however, I am coming round to the view that using DRM to (attempt to) enforce those restrictions is fair game. I suspect the law would have to, in effect, allow someone to voluntarily give up the normal fair use defence in connection with a restricted copy of something as part of the contract under which the material is supplied. This may already be possible in some jurisdictions, but probably falls into the same sort of legal mess as EULAs.
What I'm worried about is a situation where supplying restricted versions of a product without making a full version available becomes the norm, which effectively overcomes the whole idea of mandating a DRM-free version for those who want to exercise their full fair use should-be-rights and leaves us back where we started. However, I suspect this danger could be overcome by a combination of mandatory licensing (if a work subject to copyright is available to a certain market with additional restrictions for $d, it must also be available to the same market without those restrictions and in an unrestricted format for no more than $d*N where N is probably something like 5 or 10) and regulation on labelling (so people actually know when they're paying for restricted material, and understand what the realistic implications are).
The one compelling argument I have seen for not outright prohibiting DRM-style tactics in law is that such tactics allow businesses to experiment with new business models such as on-line rentals, where consumers can have a one-off viewing of something for a much lower price than buying a permanent copy. There is a danger that such models will become the default, removing the option for buying a permanent copy altogether, which is clearly not in consumer interests. However, if this threat can be reliably neutralised (and there's no reason it can't from a legal perspective) then it's likely that introducing the rental model as an additional choice will work in consumers' favour, and market forces will sort out the relative pricing.
Similarly, if Microsoft wanted to sell little more than a kernel as "Windows 7" but then provide additional modules at extra cost for parts of the market that would want them, this again may work out in consumers' favour. They would have to sell a useful core very cheaply, of course, but given the rampant piracy that goes on despite all the product activation mess, that might well be a more economically viable long-term business model for them anyway. There's a whole new upselling market to be had where PC vendors are fighting over the best combinations of hardware and software options, with Microsoft sitting in the background and raking in a little extra from every extra upgrade.
I don't think the kinds of limitation you're describing (actively and arbitrarily disabling parts of the basic system hardware) would ever fly, in the same way that it didn't take long for chip manufacturers to give up on trying to prevent overclocking. But providing basic Windows and then offering things like Media Player or IE as genuine, paid-for add-ons presumably heavily marketed on convenience grounds might go somewhere. Actually, those are probably the two things they wouldn't want to do that with, because of the competition for media formats and web space and because of the high quality, freely available competition, but you get the idea.
Bottom line: modularity would be bad if the baseline was expensive as what we have today and then you had to pay more on top to get the equivalent functionality, but that's not likely to fly in the market. Modularity where you pay little for the baseline but then a few relevant top-up costs for what you actually need might be in both Microsoft's and consumers' interests in the long run.
You mean, it would have made you feel good about your preconceived notions if they had been ratified by this event. Except, oh darn, they were not. So, just like everyone else, you reassure us that the sky might fall next time, and urge us to take your concerns seriously in the absence of evidence that we should.
You write with great confidence, considering that you have absolutely no idea who you're talking to or what you're talking about.
I can't speak for anyone else, but I have personally had my life screwed up for several months because of simple mistakes in government handling of personal information. I nearly lost my home because of a simple human error by some low grade government worker with direct access to personal information about me and inadequate supervsion, working in a system with insufficient safeguards. My notions aren't preconceived, they are born of direct personal experience. My concerns about skies falling (your choice of words, not mine) are born of direct personal experience. You talk about absence of evidence, yet from where I sit, you appear to be the one disconnected from reality: you assume that because something hasn't happened to you, there's no problem for anyone else either.
I believe you're missing my point. Monitoring is good, but prevention is much better.
Governments (or any other organisation for that matter) should only be allowed to hold the very minimum personal information required to do their jobs. At present, they typically collect vastly more information about both their own population and those visiting their countries than they have any need or right to have. Personal information can't be abused if no-one has collected it and given it to a potential abuser.
Access to any personal information that does need to be held should be subject to strict controls. In particular, it should not be possible for a single worker to arbitrarily access anything sensitive, and where access is legitimately required as part of a job, not only should that access be recorded but the people given those jobs should be carefully checked first. At present, far too much access is available to individuals working in certain government departments, financial institutions and the like, and it is often abused. This suggests that insufficient safeguards are in place on either the people being hired or the access to the information. Sure, monitoring and deterrence matter (and I happen to think that deliberate invasion of privacy or abuse of access to personal information should be regarded as very serious crimes and punished accordingly) but these measures are only worth anything after the fact. There is a whole load of work you can do to make it less likely that abuses will happen in the first place.
And no, I don't think certain governments do need anything like as much information about people entering their countries as they currently collect. In fact, I know a significant number of people who now refuse to travel to places like the US for precisely that reason.
You're right that it's unlikely the data in this case could be used effectively to harm high-profile people like presidential candidates, whose international movements are often the subject of public record anyway. Of course, that doesn't mean it couldn't be immensely damaging to Joe Public who doesn't enjoy the same practical protections. Does anyone feel like trying to steal Barack Obama's identity? I doubt that would work out real well for them. But stealing some random member of the public's, and then cleaning out their bank account or committing a crime so it looks like they did it? That's all too real a prospect, because it happens to thousands of people every year, and that figure is growing very fast. Personally, I wouldn't call this "inconsequential".
It would have been nice if the candidates had been personally damaged by whatever data was revealed, obviously not enough that it would compromise their personal safety or anything like that, but enough that they would not want it to happen again, perhaps something that caused them some political embarrassment. Short of the US having something like the UK's child benefit data loss, where almost half the population was affected and the media made merry with the story for weeks, it's hard to see how anything other than directly affecting the people at the top of government will make the point that privacy and security do matter, and the law isn't keeping up with the technology.
Full transparency is a more effective solution than full opacity because it's both easier to achieve, and eliminates abuses by making them uses.
In general, full transparency is not a solution to privacy problems, though, because not everyone has equal power given the information. If a public official knows my name and address, he can look me up on all kinds of databases and, more to the point, make entries on all kinds of databases that may ultimately cause harm to me. If I know his name and address, what am I going to do, go stand outside his house with a sign saying "Abuser of power!"?
I read a much better articulated version of this argument a little while ago, possibly written by someone like Ross Anderson, but I'm afraid I can't find it now. If anyone has it, please do post the link.
What you're overlooking is that not everyone turned up by the system will be a criminal, but even being a suspect in a crime can result in terribly damaging treatment in today's knee-jerk justice system. There have been several recorded cases of someone being arrested erroneously, finding themselves on the DNA database as a result, and later finding themselves harrasssed again in connection with a second crime they did not commit. The danger of this sort of thing is exactly why we have the presumption of innocence, and why treating all citizens as suspects must not be tolerated.
If you think I'm exaggerating and you really have nothing to fear, go ahead and look up some statistics. Here's a few to start you off:
What proportion of people arrested and detained are subsequently convicted of a serious criminal offence?
What proportion of people arrested under anti-terrorism legislation are subsequently convicted of a terrorism-related offence?
What proportion of people stopped in the street and searched under anti-terrorism legislation are subsequently arrested? And then convicted of a serious criminal offence?
What proportion of people on the DNA database are black? What is the proportion of UK citizens as a whole? What proportion of serious criminal offences are committed by black people?
The same questions as above, but about men.
What proportion of suspects identified via the DNA database are subsequently charged and convicted of a serious criminal offence? Looked at from the other side, what proportion of people go through the resulting hassle even though they are innocent?
On average, after a case of complete identity theft, how long does it take someone to get their life back to normal?
What proportion of major government IT projects have been completed on time? On budget? With all the features working? With data held securely?
What proportion of the UK population have had personal information lost by the government? What is the value on the black market of that information, per person?
What are the fastest growing crimes in the UK today?
If you can find reliably sourced answers to those and still honestly believe that we have nothing to fear from the DNA database and its like, then fair enough, you're as entitled to your opinion as anyone else. But I can't help thinking that you've probably never been the wrong side of the kind of government screw up that wastes months of your life to put right and/or completely destroys your career/personal life/professional reputation, and then found that the government is basically immune from any kind of serious repercussions no matter how badly they may wreck your life and how much effort it takes you to put it right, if you ever can.
The list above is an off-the-top-of-my-head list of Open Source projects that I use and rely upon on a regular basis. It has grown significantly over the years, going from a relatively small list of key programs to permeating nearly every aspect of my day-to-day life and work. If you did a similar inventory of the OSS products you use, I wouldn't be surprised if you came up with a similarly growing list.
Sure, I use plenty of OSS, and I'm grateful to those who have contributed to it and then given it away for the benefit of others.
However, the list certainly isn't growing exponentially. There are a few major applications I use frequently — Firefox, Thunderbird, OpenOffice, a few programming-related tools — but I've been using these for a while. I won't need two web browsers next year, four the year after, and eight by 2011.
As I've often observed in these Slashdot discussions, OSS has a few well-known success stories, but when you look at where the priority is, it's always focussed on the same relatively small number of projects. This is, perhaps, almost inevitable given the nature of the beast: volunteers will tend to concentrate their efforts on things that matter to them, which means mass-(geek-)market commodity things like programming tools and Linux and multimedia utilities attract a lot of help and attention, but most other projects don't.
People have listed several dozen of the better known OSS titles in this thread, perhaps trying to convince me that I'm wrong about the lack of evidence for growth in OSS provided by TFA or something. The thing is, I've never disputed that there is growth in the major OSS projects. I'm simply arguing that the article doesn't show that the world of OSS as a whole is growing (by any useful metric) in the same way.
Also, for what it's worth, I think there's sometimes a lack of perspective among Slashdot posters about just how big and diverse the software world is. I don't know whether the parent poster is one of them, but it's a feeling I get from this discussion generally. To put things in perspective, I'd say this thread mentions most of the best-known OSS today. And yet, there are commercial software products that have probably shipped more units alone than all the OSS titles mentioned in this entire thread put together. Pick any major field such as business management, or CAD/CAM/CAE, or games, and there are orders of magnitude more commercial, closed source titles available right now than all the OSS mentioned in this thread put together, and that's just "off-the-shelf software" to make it a fair comparison given the context of this discussion. I'm not bashing the OSS world here, just asking that we keep a sense of perspective in these discussions and try to avoid too much hyperbole.
Ah, OK. If you're talking about anybody providing a service via the web, then sure, it's a significant part of the industry. I assumed you meant Web Services(TM)(R) (patent pending).:-)
What's particularly interesting to note is that web services are the latest craze in software development.
Sorry, I must have missed that memo. How many major name web services actually make money today?
I would wager that the overwhelming majority of software development is still nothing to do with web services, and moreover that those web services that do have real value to someone are mostly (like a lot of software) written for in-house use and not to make money through the software-as-a-service model. I would also wager that of those businesses set up to operate on a software-as-a-service model, very few actually have healthy growth and a sustainable business plan. Indeed, as with OSS and the "free product, paid support" idea, I expect a few major areas will surely have critical mass, but a whole bunch behind them won't.
“Measuring programming progress by lines of code is like measuring aircraft building progress by weight.” — Bill Gates
The rest of us got over this particular naive metric years ago. The fact that lines of OSS code produced are growing exponentially doesn't tell us anything useful about how much useful stuff can now be done with OSS.
Moreover, the rate of growth now is not the interesting thing. The total volume of serious OSS is still relatively small, and so is its growth in absolute terms. The future potential is far more interesting to explore.
For example, if (as TFA tells us) packaged OSS generated revenues of $1.8B in 2006 and this was around 0.7% of total revenue generated from all packaged software sales, then I disagree with the article's claim that the OSS revenue was not trivial compared to the market as a whole. In business terms, 0.7% market share is nothing. On the other hand, if you also say that the OSS revenue is doubling every year while the total remains roughly constant, and you have evidence that this will continue giving exponential growth, then your data suggests that in a few years the OSS revenue very much will be significant.
However, I'm struggling to find data to support those claims on a first quick look at TFA. The pretty pictures just show that the volume of code is going up, which doesn't tell us anything about the value (economic or practical) of what's being written, nor what the future trends for that value are likely to be.
Slashdot Mirror
Maybe they'd wait if, I dunno, we didn't advertise the details of suspects publicly on the basis of some random allegation that has yet to be proven in court?
Getting back to this specific case, I'm not sure what's more disturbing:
And on an unrelated note, what gives with the weird styling on Slashdot since earlier this evening? Loads of HTML formatting, such as the list above, is completely broken. :-(
Sorry, friend, but I think you've misunderstood what ISO 9000 is all about.
It doesn't require you to have a good process, only that you have some process and actually follow it.
Of course, in my business (software development), the only reason to get formal certification for things like ISO 9000 or TickIT compliance is as a means to get contracts with clients (typically governments... surprise!) who insist on it. If anyone thinks it's actually about improving the productivity of businesses or quality of the products they make, consider that almost every practical improvement in modern management of software development processes has moved away from heavyweight formal processes that do everything step-by-step and toward more lightweight, iterative, interactive approaches in one way or another. You don't necessarily have to buy into the trendy things with words like "extreme" and "agile" associated with them, but seriously, the waterfall model died decades ago in everywhere but ISO land.
We already pay a fair rate for the bandwidth we use.
No, you don't. This is the problem.
It's not your fault. You're just paying what the ISP is currently asking, and expecting what they offer to be provided in return.
But in any flat-rate pricing scheme like this, there are winners and losers. For everyone who's above the middle mark, they're getting a good deal. There's nothing unethical about taking advantage of that when it's offered. But it's not a fair rate, because those who are below the middle mark are subsidising those above it.
Even so, there's one, important thing you're missing that the traditional companies do: they provide all the money needed to assemble, print, distribute and market your book, then give you a percentage of what comes in.
But the actual monetary cost of assembly and printing via print-on-demand is close enough to zero that it makes little difference. The major expense is time, and if you've got the time to write a whole book worth publishing, you're obviously not too worried about that. And as I noted before, the distribution and marketing advantages of using a major publishing house are highly overrated if you're not writing the kind of book that's mass market and going to make it into bricks and mortar bookstores, while the percentage you get back is literally an order of magnitude less than you would via POD.
Going POD or self-publishing means that you have to pony up in advance, with no guarantee that you'll recover your investment, let alone make a profit.
I hate to break this to you, but going via an old school publisher is no guarantee you'll ever cover your advance or make a profit, either. In any case, it will take a lot more book sales before you do...
It'd probably have to bark at the STL libraries first, since that's what they usually do in sort() and many other algorithms.
Sorry, I don't follow. What does integer division have to do with sorting?
Integer division is a perfectly well defined and frequently used operation in C/C++.
Sure it is, but it's still inappropriate to ever use it unless you really do want a particular kind of truncation on the answer.
I studied CS in academia and worked in industry for a while, and I don't know what the "white book" is either. I guess I'd better give up and become a lawyer. :-(
Or, just maybe, CS education is about more than name dropping...
Isn't C still part of a basic university computer science education?
Sadly, no. Even in the better universities, it is often covered only briefly, and it's just expected that any self-respecting CS student will learn it. Of course, while any self-respecting CS student should be gaining exposure to a variety of programming languages and tools, part of the point of the formal academic course is to introduce those students to the ideas they need to be able to do this, which is why I think the courses should include some reasonably deep instruction in example languages of various types before assuming the students will go broaden their horizons independently. If the courses won't do that, what are they actually teaching that someone couldn't just learn on their own?
IMHO, C actually has far more place in such courses than C++, which while a useful practical language isn't really an optimal example of anything for teaching purposes (apart from perhaps how to combine different programming ideas to produce a practical language).
Now, who'll be the first one to spot and explain the problem here, preferrably without actually running it?
Probably your static code analysis tool, which would presumably warn you about the potential division by zero, and for that matter the fact that you're dividing using integer arithmetic in the first place.
But neither intrinsic functions or inline assembler are portable, which is unfortunate if (like many high performance libraries) you need to build on many platforms.
I think the more relevant thing is that if you use a decent compiler, most of them will pick up a simple optimisation like this and generate appropriate code just from the C++ these days. Optimisers certainly have a long way to go, but we're well past the point of getting this sort of thing right routinely.
You seem to have this wonderful illusion that going via a real publisher is some sort of guarantee of decent quality. It certainly isn't in programming books: take a look at some more recent titles by Addison-Wesley, previously the source of the majority of good C++ books, and it's hard not to find typos all over the place, amateurish design and typography, and a general lack of editorial quality. Several other professional publishers were never much better than that. And if you think going for a reputable university press would be better than the big professional outfits, guess again: take a look at the guidelines for authors published by some of them (you can often find these on their web sites) and they tend to insist on petty things like conforming to some arbitrary (or sometimes, IMHO, outright incorrect) house styles that do nothing to improve the quality of a work written by a skilled and knowledgable author, before they'll even grace you with the benefits of their immense knowledge and experience (and yes, that was sarcasm: do you realise who these "professional editors" typically are?).
Not so long ago, I considered writing a book, after receiving some favourable comments on a couple of smaller pieces of work I'd done. I looked into what would be involved in going through a mainstream publisher, and came away asking why anyone with decent writing ability and decent knowledge of the presentation side of things would ever use one. I always knew authors didn't normally receive a high percentage of the cover price of a book, but I was shocked at just how little it really is: well under 10% seems pretty typical, and it varies depending on the market. Anyone with some basic knowledge of subjects like book layout and typography could produce a better design than many of the "professionals" do without even trying, using any good DTP package (or even LaTeX, for technical books). I have some contacts at some university publishers, and some of the comments I hear about their editorial teams are just appalling. And it's not even like being published by a reputable press will get you into bricks and mortar stores any more: if I walk into my local Borders, it's full of "computing for fools" and "learn computing in ten seconds" books, but even relatively mainstream "serious" books are in short supply these days and I usually have to look on-line for them. There are almost no more specialised books of the kind you'd typically find from an academic press.
So, if people are going to have to order in or buy from the big on-line vendors to get any serious technical book, and publishers add precious little value in the editorial and presentation departments if you're likely to be as competent as their staff anyway, what real advantage do they bring? You can jump through a few hoops to get things like an ISBN and use a reputable POD organisation with links to a major distribution channel, and then you can have exactly the book you want available through all the same (realistic) channels anyway, but with at least 10x the profit margin on every book sold and with almost no financial liability if your book doesn't sell. Do professional publishers really produce and market books well enough to get an order of magnitude more sales for specialist titles than the self-publishing/POD route? Not if the data from people like O'Reilly and the anecdotes from published authors on the web are anything to go by, they don't.
And of course, just because you're self-publishing, that doesn't mean your quality has to suck. You might be a decent writer and designer yourself, compared to what you'd get from a typical publisher, and you can still bring in professional help for areas where you need it if you're not an expert. It's just done on your terms, with costs you know, and once you've paid your overheads you don't have to keep paying for them with every new book sold.
So tell me again, why would anyone with decent writing and presentation skills go to a professional publisher today, if they aren't writing the kind of mass-market fiction or technical books for idiots that are likely to sell with a really high volume and make it into real world bookshops?
Look, I'm all for keeping an eye out on Wikileaks. I think it serves a very important purpose in a time when a lot of governments - and their people - feel that the withholding of information is a good idea.
The thing is, sometimes withholding information from the general public is a good idea. Not all of the general public are good guys, and there are legitimate reasons for governments to do some things in secret, for a time. The practical problem is that the general public have no way to ascertain whether any information being withheld is being kept from them for legitimate reasons or less savoury purposes if they can't see it. However, there are pragmatic alternative approaches to dealing with this problem via independent oversight that do not involve full disclosure of everything as implied by Wikileaks.
Unless I missed something, the entire report is basically arguing for such a middle ground. I don't see anywhere it says we should throw children into dangerous situations they can't cope with. Rather, it seems (from a first quick scan at least) to be advocating throwing children into somewhat dangerous situations carefully so they can learn to handle them safely in their own right.
This sounds like the kind of common sense you'd get from someone who actually deals with children professionally and sorts out problems in real life. Oh, wait, she is. :-)
Sadly, I gather she's decided that her television programmes weren't necessarily in the interests of the children participating and discontinued them now. That's a pity; they were very informative and seemed to be done quite responsibly from a naive but interested observer's point of view.
From my experience, all you can eat plans have always been cheaper/more convenient to the customer than tiered rates.
They are also usually loss-leaders. If they carry on for long, there are only two likely outcomes: the company goes bust, or a substantial number of customers wind up getting screwed by paying for more than they need.
This is obviously going to be true of broadband ISPs before very long, as "unmetered" access becomes implausible with the rise of on-line multimedia distribution. The current hardware networks simply aren't up to providing that for everyone, so somebody somewhere is going to pay for the extra capacity, and you can bet it won't be the top-tier ISPs! Personally, since I use only a small fraction of my monthly capacity allowance but the next tier down is inadequate for me and most of the decent ISPs around here have similar models, I would be quite happy to go back to paying by the GB downloaded; that way, I wouldn't be paying extra to support all the heavy downloaders at the other end of the scale.
Don't prohibit DRM. That's a technical measure, and completely within the company's right to try to implement. What needs to happen is that the DMCA needs to get thrown out so that it's no longer illegal to break that DRM. Exercising fair-use rights should not be illegal.
I think I agree with the spirit of what you're saying, but I don't agree with the letter. Merely revoking the necessary parts of the DMCA does not help anyone who can't break the DRM technically, because they still can't make fair use of the material. Moreover, it assumes that someone will always be able to break the DRM and willing to share the technology they used to do so. While this is, AFAIK, true of all major DRM platforms today, it would be foolish to assume that it will always remain so.
I think for any fair DRM-related system to work, the law will need to start distinguishing more carefully between someone who owns a legitimate, permanent copy of a work subject to copyright, and someone who accepts an agreement when the material is supplied that restricts their use of it, for example by imposing a time limit.
In the former case, I believe fair use should be closer to a protected right than its current status of affirmative defence: it should be illegal for anyone to actively inhibit exercising fair use, for example by using DRM technology. Moreover, that particular copy should be treated like any physical property, such that the person who paid for it may lend it out to others or sell it on under the doctrine of first sale, provided they do not retain access to or further copies of it themselves at the same time, and there is no scope for imposing additional restrictions via some sort of licence agreement. In essence, as long as electronic material is not duplicated in a way that would be an infringement of copyright if you did the same thing with a book you bought from a store, it should be given the same status in law.
In the case where restrictions are understood and accepted before receiving a copy, however, I am coming round to the view that using DRM to (attempt to) enforce those restrictions is fair game. I suspect the law would have to, in effect, allow someone to voluntarily give up the normal fair use defence in connection with a restricted copy of something as part of the contract under which the material is supplied. This may already be possible in some jurisdictions, but probably falls into the same sort of legal mess as EULAs.
What I'm worried about is a situation where supplying restricted versions of a product without making a full version available becomes the norm, which effectively overcomes the whole idea of mandating a DRM-free version for those who want to exercise their full fair use should-be-rights and leaves us back where we started. However, I suspect this danger could be overcome by a combination of mandatory licensing (if a work subject to copyright is available to a certain market with additional restrictions for $d, it must also be available to the same market without those restrictions and in an unrestricted format for no more than $d*N where N is probably something like 5 or 10) and regulation on labelling (so people actually know when they're paying for restricted material, and understand what the realistic implications are).
Not a pretty picture.
I'm not so sure; the context matters.
The one compelling argument I have seen for not outright prohibiting DRM-style tactics in law is that such tactics allow businesses to experiment with new business models such as on-line rentals, where consumers can have a one-off viewing of something for a much lower price than buying a permanent copy. There is a danger that such models will become the default, removing the option for buying a permanent copy altogether, which is clearly not in consumer interests. However, if this threat can be reliably neutralised (and there's no reason it can't from a legal perspective) then it's likely that introducing the rental model as an additional choice will work in consumers' favour, and market forces will sort out the relative pricing.
Similarly, if Microsoft wanted to sell little more than a kernel as "Windows 7" but then provide additional modules at extra cost for parts of the market that would want them, this again may work out in consumers' favour. They would have to sell a useful core very cheaply, of course, but given the rampant piracy that goes on despite all the product activation mess, that might well be a more economically viable long-term business model for them anyway. There's a whole new upselling market to be had where PC vendors are fighting over the best combinations of hardware and software options, with Microsoft sitting in the background and raking in a little extra from every extra upgrade.
I don't think the kinds of limitation you're describing (actively and arbitrarily disabling parts of the basic system hardware) would ever fly, in the same way that it didn't take long for chip manufacturers to give up on trying to prevent overclocking. But providing basic Windows and then offering things like Media Player or IE as genuine, paid-for add-ons presumably heavily marketed on convenience grounds might go somewhere. Actually, those are probably the two things they wouldn't want to do that with, because of the competition for media formats and web space and because of the high quality, freely available competition, but you get the idea.
Bottom line: modularity would be bad if the baseline was expensive as what we have today and then you had to pay more on top to get the equivalent functionality, but that's not likely to fly in the market. Modularity where you pay little for the baseline but then a few relevant top-up costs for what you actually need might be in both Microsoft's and consumers' interests in the long run.
You mean, it would have made you feel good about your preconceived notions if they had been ratified by this event. Except, oh darn, they were not. So, just like everyone else, you reassure us that the sky might fall next time, and urge us to take your concerns seriously in the absence of evidence that we should.
You write with great confidence, considering that you have absolutely no idea who you're talking to or what you're talking about.
I can't speak for anyone else, but I have personally had my life screwed up for several months because of simple mistakes in government handling of personal information. I nearly lost my home because of a simple human error by some low grade government worker with direct access to personal information about me and inadequate supervsion, working in a system with insufficient safeguards. My notions aren't preconceived, they are born of direct personal experience. My concerns about skies falling (your choice of words, not mine) are born of direct personal experience. You talk about absence of evidence, yet from where I sit, you appear to be the one disconnected from reality: you assume that because something hasn't happened to you, there's no problem for anyone else either.
I believe you're missing my point. Monitoring is good, but prevention is much better.
Governments (or any other organisation for that matter) should only be allowed to hold the very minimum personal information required to do their jobs. At present, they typically collect vastly more information about both their own population and those visiting their countries than they have any need or right to have. Personal information can't be abused if no-one has collected it and given it to a potential abuser.
Access to any personal information that does need to be held should be subject to strict controls. In particular, it should not be possible for a single worker to arbitrarily access anything sensitive, and where access is legitimately required as part of a job, not only should that access be recorded but the people given those jobs should be carefully checked first. At present, far too much access is available to individuals working in certain government departments, financial institutions and the like, and it is often abused. This suggests that insufficient safeguards are in place on either the people being hired or the access to the information. Sure, monitoring and deterrence matter (and I happen to think that deliberate invasion of privacy or abuse of access to personal information should be regarded as very serious crimes and punished accordingly) but these measures are only worth anything after the fact. There is a whole load of work you can do to make it less likely that abuses will happen in the first place.
And no, I don't think certain governments do need anything like as much information about people entering their countries as they currently collect. In fact, I know a significant number of people who now refuse to travel to places like the US for precisely that reason.
You see, The protections worked.
You seem to have an interesting idea of what it means for the protections to "work". It's a bit like saying, "Hi, I've just shot you dead."
I mean outside discarding the information and not keeping a record of who enters and leave the country, the protections worked quite well.
Gosh, there's a thought. Maybe there's a moral here about allowing governments to conduct systematic mass surveillance?
You're right that it's unlikely the data in this case could be used effectively to harm high-profile people like presidential candidates, whose international movements are often the subject of public record anyway. Of course, that doesn't mean it couldn't be immensely damaging to Joe Public who doesn't enjoy the same practical protections. Does anyone feel like trying to steal Barack Obama's identity? I doubt that would work out real well for them. But stealing some random member of the public's, and then cleaning out their bank account or committing a crime so it looks like they did it? That's all too real a prospect, because it happens to thousands of people every year, and that figure is growing very fast. Personally, I wouldn't call this "inconsequential".
It would have been nice if the candidates had been personally damaged by whatever data was revealed, obviously not enough that it would compromise their personal safety or anything like that, but enough that they would not want it to happen again, perhaps something that caused them some political embarrassment. Short of the US having something like the UK's child benefit data loss, where almost half the population was affected and the media made merry with the story for weeks, it's hard to see how anything other than directly affecting the people at the top of government will make the point that privacy and security do matter, and the law isn't keeping up with the technology.
Full transparency is a more effective solution than full opacity because it's both easier to achieve, and eliminates abuses by making them uses.
In general, full transparency is not a solution to privacy problems, though, because not everyone has equal power given the information. If a public official knows my name and address, he can look me up on all kinds of databases and, more to the point, make entries on all kinds of databases that may ultimately cause harm to me. If I know his name and address, what am I going to do, go stand outside his house with a sign saying "Abuser of power!"?
I read a much better articulated version of this argument a little while ago, possibly written by someone like Ross Anderson, but I'm afraid I can't find it now. If anyone has it, please do post the link.
What you're overlooking is that not everyone turned up by the system will be a criminal, but even being a suspect in a crime can result in terribly damaging treatment in today's knee-jerk justice system. There have been several recorded cases of someone being arrested erroneously, finding themselves on the DNA database as a result, and later finding themselves harrasssed again in connection with a second crime they did not commit. The danger of this sort of thing is exactly why we have the presumption of innocence, and why treating all citizens as suspects must not be tolerated.
If you think I'm exaggerating and you really have nothing to fear, go ahead and look up some statistics. Here's a few to start you off:
If you can find reliably sourced answers to those and still honestly believe that we have nothing to fear from the DNA database and its like, then fair enough, you're as entitled to your opinion as anyone else. But I can't help thinking that you've probably never been the wrong side of the kind of government screw up that wastes months of your life to put right and/or completely destroys your career/personal life/professional reputation, and then found that the government is basically immune from any kind of serious repercussions no matter how badly they may wreck your life and how much effort it takes you to put it right, if you ever can.
The list above is an off-the-top-of-my-head list of Open Source projects that I use and rely upon on a regular basis. It has grown significantly over the years, going from a relatively small list of key programs to permeating nearly every aspect of my day-to-day life and work. If you did a similar inventory of the OSS products you use, I wouldn't be surprised if you came up with a similarly growing list.
Sure, I use plenty of OSS, and I'm grateful to those who have contributed to it and then given it away for the benefit of others.
However, the list certainly isn't growing exponentially. There are a few major applications I use frequently — Firefox, Thunderbird, OpenOffice, a few programming-related tools — but I've been using these for a while. I won't need two web browsers next year, four the year after, and eight by 2011.
As I've often observed in these Slashdot discussions, OSS has a few well-known success stories, but when you look at where the priority is, it's always focussed on the same relatively small number of projects. This is, perhaps, almost inevitable given the nature of the beast: volunteers will tend to concentrate their efforts on things that matter to them, which means mass-(geek-)market commodity things like programming tools and Linux and multimedia utilities attract a lot of help and attention, but most other projects don't.
People have listed several dozen of the better known OSS titles in this thread, perhaps trying to convince me that I'm wrong about the lack of evidence for growth in OSS provided by TFA or something. The thing is, I've never disputed that there is growth in the major OSS projects. I'm simply arguing that the article doesn't show that the world of OSS as a whole is growing (by any useful metric) in the same way.
Also, for what it's worth, I think there's sometimes a lack of perspective among Slashdot posters about just how big and diverse the software world is. I don't know whether the parent poster is one of them, but it's a feeling I get from this discussion generally. To put things in perspective, I'd say this thread mentions most of the best-known OSS today. And yet, there are commercial software products that have probably shipped more units alone than all the OSS titles mentioned in this entire thread put together. Pick any major field such as business management, or CAD/CAM/CAE, or games, and there are orders of magnitude more commercial, closed source titles available right now than all the OSS mentioned in this thread put together, and that's just "off-the-shelf software" to make it a fair comparison given the context of this discussion. I'm not bashing the OSS world here, just asking that we keep a sense of perspective in these discussions and try to avoid too much hyperbole.
Ah, OK. If you're talking about anybody providing a service via the web, then sure, it's a significant part of the industry. I assumed you meant Web Services(TM)(R) (patent pending). :-)
What's particularly interesting to note is that web services are the latest craze in software development.
Sorry, I must have missed that memo. How many major name web services actually make money today?
I would wager that the overwhelming majority of software development is still nothing to do with web services, and moreover that those web services that do have real value to someone are mostly (like a lot of software) written for in-house use and not to make money through the software-as-a-service model. I would also wager that of those businesses set up to operate on a software-as-a-service model, very few actually have healthy growth and a sustainable business plan. Indeed, as with OSS and the "free product, paid support" idea, I expect a few major areas will surely have critical mass, but a whole bunch behind them won't.
“Measuring programming progress by lines of code is like measuring aircraft building progress by weight.” — Bill Gates
The rest of us got over this particular naive metric years ago. The fact that lines of OSS code produced are growing exponentially doesn't tell us anything useful about how much useful stuff can now be done with OSS.
Moreover, the rate of growth now is not the interesting thing. The total volume of serious OSS is still relatively small, and so is its growth in absolute terms. The future potential is far more interesting to explore.
For example, if (as TFA tells us) packaged OSS generated revenues of $1.8B in 2006 and this was around 0.7% of total revenue generated from all packaged software sales, then I disagree with the article's claim that the OSS revenue was not trivial compared to the market as a whole. In business terms, 0.7% market share is nothing. On the other hand, if you also say that the OSS revenue is doubling every year while the total remains roughly constant, and you have evidence that this will continue giving exponential growth, then your data suggests that in a few years the OSS revenue very much will be significant.
However, I'm struggling to find data to support those claims on a first quick look at TFA. The pretty pictures just show that the volume of code is going up, which doesn't tell us anything about the value (economic or practical) of what's being written, nor what the future trends for that value are likely to be.