The Graphene one. The others are just the usual BS from people clueless about how computers work and what they can and cannot do.
Spoken like a true programmer or sysadmin with no knowledge of statistics, modelling, machine learning or data analysis. I know, because I was one (and I still write code and maintain servers). But I've also moved into the above fields, and it's a completely different world. The discrete math and logic you use in programming are completely useless here, and the things you can do and the hurdles you come across are very different from the ones you see in programming. Of course, you still have to implement your models and analyses, and you get all the usual issues there (plus things like numerical instability), but even if the software is running fine you'll have things like parameter identifiability, difficulties in comparing models, lack of data in the places where you need it, conceptual problems with the models that can only be solved by making them more complex, which leads to lack of data problems and the need for massive amounts of compute power, and so on. These are the things they will be trying to tackle, and they have nothing to do with the limitations of Turing-style computers.
I do remain sceptical about having a chat with a Turing-level AI any time soon, but data analysis, modelling and inference methods are getting better and better (see Google Search, Watson) and I don't think that continued research into these things is a waste of money. Neither do Google, Facebook, Microsoft, the US government, and the EU apparently.
Finally, here's another EU project in this direction that is both scary and interesting.
In Kim Stanley Robinson's Mars trilogy, the northern ocean is filled with fresh water from the molten polar ice cap, while the rivers take up salt from the rocks they flow over, so there are salty rivers flowing into a fresh water ocean. I'm not sure how realistic that is, but it doesn't seem completely illogical.
As artist impressions go, I prefer this one, by Daein Ballard over the one in the article.
Maybe the Linux Foundation (or someone else, they're the first that come to mind) could do a similar thing to raise money for improving the Linux graphics and wireless stacks? How much improvement could we get for a million USD? Or perhaps there are individual developers out there who would do what Poul-Henning Kamp did? I'd be happy to contribute to such an initiative. Kickstart it?
I'm typing this on a Dell Latitude E6410, which is on that list (albeit with nVidia graphics I think, but Intel support is better, right? That's why I ordered it, anyway). When I first got this machine, it was also on the list, but Ubuntu 10.04 LTS (the most recent LTS) wouldn't even boot on it, just gave a black screen. Apparently there were multiple issues with the Intel graphics drivers, with both the E6410 and the E6510. Now it did seem that Canonical was giving those bugs some attention, but it still took many months for them to be fixed for most users. Then there was one last patch and it started working for me as well...until the next (ordinary, stable) update which broke it again. I ended up running 11.04 with a backported kernel that I didn't dare upgrade.
Then the touchpad (ALPS, not Synaptic) wasn't recognised as a touchpad, which they "solved" with a patch that sent a magic command combo to the device to switch it into imPS/2 mode. Result: scrolling worked, but it's still not recognised as a touchpad, and you still cannot configure it as such.
So yes, Canonical has a list, but I'd interpret that as "we'll try a bit harder to fix these machines, but they're otherwise just as broken as everything else", not as "tested and working".
Generally speaking, I'm seeing a lot of comments saying that there are no issues with server hardware. The OP didn't mention it, but it seems he/she was asking for desktop hardware, whose most critical functionality is (non-trivial) graphics hardware and, if it's a laptop, wifi. In my experience those things remain difficult for modern Linux kernels, especially on new hardware (i.e. anything you bought in the last year, maybe two), and there doesn't seem to be much progress either. As others have said though, Intel hardware seems to be your best bet.
How many false positives do you get though? In a classifier, having a high true positive rate is good, but only if it comes with a low false positive rate. It seems that in this case, perhaps there are a few too many false positives.
I've never played or even seen Minecraft, but I've heard of it and I could envision downloading this and having a look to see if it's interesting enough. Unfortunately, the local branch of the RIAA has sued my ISP into blocking the Pirate Bay, so I can't. Honestly, I was unlikely to buy this in the first place, but now I'm certainly not going to. Which I suppose serves 2 Player Productions right for daring to try to compete, even if it's in a small niche market, with the big media conglomerates. Still a shame though.
Whatever the motivation, I think this is one of the strengths of the scientific method and thus, one of the reasons for its success: we aren't quick to publishing until it is just right, and therefore, perhaps our best approximation of the "truth" we can muster.
On the other hand: What if you did publish your work early and often, not as concerned with slowly and deliberately ensuring everything is just right before spreading the information -- Not keeping quiet just so that you can be the one with the badge of "1st"? Why, then worldwide cooperation could kick in. Perhaps other interested parties would help you prove or disprove the results much more quickly. Thus, accelerating the speed of scientific progress.
You are assuming that scientists only ever communicate by publishing papers in journals. That's incorrect, there is a lot (and I mean a lot) of informal communication and collaboration, by email and phone and through presentations and posters at conferences. Our knowledge is now so vast and much research being done so multidisciplinary that it's nearly impossible for any single person to know enough to really cover all the aspects of a particular investigation. In my field, you'd need to be a good programmer, an expert statistician, an experienced and knowledgeable (field- and theoretical) biologist, and a good systems analyst/modeller. Such people don't exist, so work is done in teams with each member contributing their specific expertise. When you get a weird result, you go and talk to your colleague about it, and try to figure it out together, and you keep going together until you feel that you really understand what's going on. And then you write the paper, it gets published, and then hopefully it won't turn out to have been a fluke or a mistake or not representative of the wider area of research. If the whole team can't figure it out, you might publish a "Hey, that's weird?" paper, as was done here.
If people published everything they did immediately, we'd get so many publications that it would be impossible to keep up with all of it. The whole situation would be similar to the "Linus doesn't scale" problem in Linux kernel development a few years ago, where Linus Torvalds was inundated with patches and couldn't keep up. They solved that by appointing lieutenants, who filter and aggregate contributions. Publishing papers works the same way, you solve the smaller problems locally, and publish bigger and better-vetted results, so that everyone else doesn't waste their precious time on solving other people's small problems and consequently invalid results. Also, people wouldn't waste their precious time on writing up all those small problems, and peer-reviewing them, and so on. Writing a paper is not like writing a post on the Internet (something that some climate change deniers often conveniently forget), it takes serious time and effort by a group of people to make sure that the results are really of good enough quality. You don't want to waste that effort on trivial things.
I've got an MSc in CompSci, and I'm now doing a PhD in a biology department. I teach a programming course using Matlab, and I've recently started using R to do my own analysis stuff, mainly because it's popular and I'd like to stay compatible with the rest of the field, as well as use some specialised software that works with it. I have to say, being used to real programming languages (such as C++, I'm not counting Matlab here, although see below) I'm quite frustrated with R. Function names are generally different from other languages and to me at least unintuitive, and the documentation is too often extremely vague and difficult to search.
For example, the function match() returns the offsets of entries in a vector that match a given object. But what exactly constitutes a match, well, according to help(match) that is "to some extent a matter of definition". It goes on to give an example or two, but that definition remains elusive. Or look at this gem from help(as.vector):
Value: For ‘as.vector’, a vector (atomic or of type list). All attributes are removed from the result if it is of an atomic mode, but not in general for a list result. The default method handles 24 input types and 12 values of ‘type’: the details of most coercions are undocumented and subject to change.
First, a "vector of type list" is actually just a list. In R, a vector is an ordered collection of elements all of the same type, while a list is an ordered collection of elements of (possibly) different types. So, by the normal Liskov rules, one could say that a vector is a kind of list in which the types of the elements are all identical. According to the R language definition however, a list is a kind of vector. In practice, lists and vectors are used in rather different ways so their exact relation is not so relevant, and it doesn't make much sense for the help page to throw them together like this. Second, apparently attributes are removed for atomic vectors, but not "in general" for lists. This is a somewhat arbitrary inconsistency, and it leaves the reader to wonder if there are specific attributes or lists for which this doesn't count. But the kicker is in the last sentence: not only are the exact workings of this function explicitly undocumented, they are also subject to change without notice! Note that these are not functions from some obscure package that I pulled off of somebodies blog. They are core language functions, and unfortunately these examples are not exceptional. A colleague of mine recently had his whole analysis suddenly return weird results after a routine update of an add-on package, because someone decided to swap the order of the longitude and latitude arguments to a function for no particular reason.
That's not to say that R is not usable, but in my opinion is is unsuitable for any kind of programming, and perhaps unsuitable for programmers. R is a powerful, extensible system for statistical analysis, with a command line interface. If you consider your text files with R code as reference notes rather than as source files, and if you use R interactively, copy-pasting lines from your notes and checking after every couple of operations that it's actually doing what you think it is, then you can do useful things with R. Looking around me, that is in fact how most people use it, and what I've taken to doing as well, although I can't resist attempting to automate things here and there.
Comparing R to Matlab, in my eyes there's no contest in terms of ease of use. The Matlab help files are professionally written and tell you what you need to know in enough detail to be useful, and that difference alone makes it a lot better. The language itself is also a bit more sensible, at least to me, being designed as an easier-to-use alternative to FORTRAN, where R is based on LISP. I was originally considering moving my course to R from Matlab, since I don't like to teach proprietary software, but now that I have some experience with R I'm pretty sure
Well, it doesn't have to be technically the exact same set of bits, or even an exactly identical interface. What he seems to be getting at is that he wants to shift from a model of personal computing where you own a personal computer that you use in the privacy of your home and have full control of, to a model where you consume a set of interlinked services that are provided partially by devices you lease (e.g. a smart phone), partially by devices you own (tablet), and partially by servers run by third parties, with ultimate control of your actions and your data mostly in the hands of the service providers. This, of course, is the antithesis of the idea of Free Software.
Further down, he states "If your way of seeing the world IS genuinely more productive, effective, efficient, insightful and usable, then you should be confident that you will win in the long term" which is either naive or disingenuous, and incidentally a nice example of why RMS dislikes the term Open Source so much. Looking at the current state of GNU/Linux on the desktop and comparing that to Windows XP and Windows 7, I'm not so sure that GNU/Linux is any better (disclaimer, I have been running GNU/Linux exclusively since 1999, so I'm comparing between my machines and my coworkers'). But that is not the point, and that is not what the question was about. The question was about freedom, about controlling the computation and communication that is done on your behalf by your equipment. The fact that he sidesteps this question in much the same way that Jono Bacon sidestepped RMS's criticism of the new integrated Amazon search engine says a lot.
So here's one more attempt to get through to the fine folks at Canonical: the question is not about whether I want to be able to buy things from Amazon easily. The question is where else in the OS you are sending things that I type into my personal computer to some server on the Internet without me knowing about it, and how I can trust you to not do that without my knowledge or permission in the future? That is what Free Software is about. Unfortunately, Canonical seems to have a very different agenda.
I'd say this is like a cat leaping into the air to catch a bird, or an Osprey diving down into the water to catch a fish. Seems like many species happily leave their domain temporarily if there's food to be had. Still, interesting that these fish have picked up the idea (maybe it's the "cat" in their name?), and anything that gets rid of pigeons is a good thing:-).
There's another issue with the UN, which is that the International Telecommunications Union, the UN organisation that would take over control of the Internet, is run by the old national telephone companies. They're extremely conservative and would be very happy to grab control and kill all innovation, or at least slow it down a lot. See Mother Earth, Mother Board for how control of the undersea cable system that ties together most of the 'net was wrested from them (and because it's an excellent read). Imagine being only allowed to connect "approved equipment" to the Internet, the way it used to be with phone networks until not that long ago. It's bureaucracy that we should be afraid of, maybe more so than censorship (after all, the US has been seizing domain names as well).
Also, don't forget to go by Egan's homepage to play some quantum soccer. It has a collection of Java applets demonstrating various things from his books, and some free stories too. Seriously geeky!
And KDE has dedicated releases to deceased contributors as well. Why not? A small note in the release notes and perhaps in the About box would be the most tasteful option in my opinion, but it's your grandmother, so you decide.
Personally I am shocked that Microsoft is taking this action. It's the number-one most consumer-friendly thing I've ever heard of them doing -- literally.
It's pretty obvious to me: Google and Facebook, two very big Microsoft competitors, get most of their revenues from online advertising. Microsoft is doing this to damage them, not to promote their user's privacy (remember Windows Media Player phoning home every file played?). The only goal here is to make the services offered by Google and Facebook look more expensive by emphasising that while they seem free, you're paying with your privacy. Remember the whole Linux Total Cost of Ownership FUD campaign? Same old, except that people don't own licences to Google or Facebook software, so there's no ownership and they've had to approach it a bit differently. Meanwhile of course, everyone still pays Microsoft tax on every new computer they buy whether they want Windows or not, but let's not mention that...
I suspect that for most people, the reason to use a mind map rather than UML would not be UML's complexity. UML scales pretty well: a bunch of boxes with names in them and lines between them to signal what is related to what is a valid UML class diagram already. Instead, I think the problem with UML is that it forces you to think very carefully about what exactly it is that you're going to create. Back in the 1960's and 1970's, people like Donald Knuth and Edsger Dijkstra advocated careful thinking about software, rigid specification, and proving correct any important algorithms. They saw software as a mathematical construct, and the exercise of building software as akin to proving theorems.
Fast forward to the Internet age. Software is everywhere, and rare is the project where the customer can tell you clearly what they want. The small cadre of people who are capable of the precise and abstract thought required to do programming the mathematical way is not by far big enough to write all the software that the world needs, so even if customers could make rigid specifications, most programmers would find them written in an alien language. So we have adopted a biological rather than mathematical approach: specifications are never exact, software is always broken, but it's okay because the software has an immune system (we call it vendor support), which fixes up errors continuously. In such a world, maybe a mind map is as formal a description as you need.
Personally, I used UML to describe a logical data model in my last big project. I was the only one with a formal CS background in the project, but everyone understood the diagrams just fine. I had to explain a few more advanced things to some people, but it was no problem. And we did think everything through very carefully, and so far the whole thing is holding up very well because of that. In my opinion, even in the age of agile development and web technologies, careful thought is still invaluable in software development, and a diagram language that lets you specify a bit more detail when you need it is a very useful tool. I'll stick with UML.
That's not quite right. If the cyclist is 14 or younger, the car driver('s compulsory insurance) is 100% liable for all damages due to the accident. If the cyclist is over 14, they're presumed to have some responsibility for avoiding an accident as well, and the car driver's liability is at least 50% and at most 100%, depending on the circumstances. Note that that is only about the (civil) damages, any criminal negligence or reckless driving charge or some such is dealt with separately.
My girlfriend was hit by a taxi while riding her bike in Amsterdam a couple of years ago. The driver was distracted by the passenger, and was (legally, but it was literally his first time doing so) driving in the tram lane. Road side obstacles and the curvature of the road (it was on a relatively small but high bridge) kept them from seeing each other until it was too late. The prosecutor's office decided not to prosecute the criminal case as it was mostly just the dangerous crossing that was at fault, and the taxi was not speeding. After a couple of letters from my girlfriend's lawyer the car driver's insurance did assume 100% liability, as they were the more guilty party.
In the end, what this law mostly means is that damages from car/bike or car/pedestrian accidents are paid for collectively by car drivers, through their insurance premiums. And it results in cyclists always getting the right of way, generally ignoring traffic lights, and sometimes riding quite recklessly, especially here in Amsterdam. You get used to it...
(My girlfriend recovered well and still doesn't wear a helmet. We never did find out if the windscreen survived.)
The difference is that when Bain did it, it was because opening jobs overseas was the only way to save a company. When Obama did it, it was for.... Actually, I have no idea why Obama would invest in Brazilian oil and not Gulf of Mexico oil. Of course, like I said, Obama shipped jobs overseas with YOUR money against your will. Romney did not.
The Netherlands has quite a substantial amount of natural gas in fields in the north of the country. We've been exploiting those fields, but not as much as we could, and we're buying gas from abroad to fulfil the rest of our need. Essentially, we're treating it partially as a strategic reserve; when fossil fuels run out, we'll have a buffer supply to get us through a transition to other types of energy. Gas production is done by a commercial company, but they pay the government for the privilege, and that money ends up in a special fund for infrastructure investment, i.e. new motorways, train lines, and even some for improvements in education as well IIRC.
Anyway, it may not be a bad idea at all for the US to save some of its domestic supply for later, and get oil from elsewhere while those other countries are still willing to sell. Obama strikes me as smart enough to think ahead strategically. Or else the oil is Brazil is pumped up by an American company and they lobbied for some more subsidies...
If we all had to show our faces, I'm sure we'd be a little more civil toward one another. Personally, I don't think I run a very high risk of ending up in the situation that this guy was in, since I value my online anonymity too much.
That's an interesting case of Prisoner's Dilemma there. If everyone is anonymous, nobody can stalk anybody, but there will be relatively many rude people. If only some people are anonymous (and potentially rude) and others aren't (and are civil), then the civil folks run the risk of being stalked by anonymous rude people. If nobody is anonymous, nobody can stalk anybody, at least not while escaping punishment, and we'll have fewer rude people.
In the Prisoner's Dilemma, there are several solutions to this. One is having an external authority that keeps both prisoner's from talking. Another strategy that works well (in the sense of fostering cooperation) is tit-for-tat with random altruism, where in each round of Prisoner's Dilemma you cooperate if the other person cooperated with you in the previous one, and betray if they betrayed you, except that you occasionally give them a break and cooperate even if they betrayed you. This latter feature keeps the system from getting stuck in endless loops of betrayal. The analogues in the case of Internet anonymity would be a real-name policy, and letting users selectively share their personal information with others. Note that for the user to follow the tit-for-tat strategy, users have to know that it's the same person who betrayed them last time, so you can't have multiple accounts per person (another reason for a real-name policy). Of course, there are also good valid uses of anonymity and anonymous speech, and there should be a place for them as well.
I realize that for many, the temptation to spread their personal misery is just too great, and so they troll, which is really just a cry for attention -- something they probably didn't get enough of growing up.
Exactly, and that is why the solution that Traynor chose is the right solution, and it's quite admirable that he did so. But even if all he would consider would be self-interest, this would have been the correct solution. Prisons and other punishment don't solve the problem of crime in a society. They just repress the symptoms, and they deflect and channel feelings of revenge into something more civilised than a lynching. If you actually want to solve the crime problem, you have to prevent people from becoming criminals, because once they are there is very little you can do to turn them into good citizens again. That means keeping a very close eye on 12-18 year olds who are running with the wrong crowd, who live in unstable families, and so on, and intervening where necessary. Keep them on the straight during puberty and adolescence, make sure they finish school and find a job and start a family, that's basically all you have to do (see also the recent find-hackers-girlfriends item). And for the ones for which it's too late, and the ones with mental disorders, well, so far prisons and closed institutions is the best we can do it seems. But if Traynor had had this kid sent to prison, the kid would have been out in a couple years, and would probably have got right back into stalking people or worse. The path he chose to take gives Traynor a much better chance of never having to deal with this again.
1) "Wherever Possible" means that they know FOSS doesn't have a solution to every little problem. However, if it comes down to word processors, databases, web browsers, etc then there are numerous FOSS alternatives.
The Netherlands' government has been operating on a "comply or explain" principle for years. All government agencies are required to use open source software and free standards, or else explain why they don't. All the government agencies I've seen in the past couple of years (municipalities, provinces, and a couple other government and semi-government organisations) use Microsoft software everywhere, with the exception of the databases, for which they use Oracle. Spatial planning is done with the proprietary ESRI stack. The only open source is usually the CMS they use for the web site, either an existing one or a home grown one which they open sourced themselves, and they have an ODF plugin installed in Word so that they can fulfil their legal requirement to be able to communicate using the ODF standard. Of course, everyone uses doc and docx.
I think the main reason is that they simply don't employ any real IT staff, just a few technicians who know how to swap out a machine and which phone number to call the supplier on when something breaks. It's difficult to find people who, given a bunch of open source software, can actually fix things themselves, and those people are expensive. Getting external support for FOSS is also not easy unless it's for something extremely mainstream. The FOSS GIS stack is getting quite capable for example, but I think there are only a handful of companies world wide who offer support for the thing, and they're all pretty small and on the other side of the world from here. So ESRI and Oracle Spatial it is.
So, which the initiative is great, and all sovereign governments should be using Free software on general principle, I'm afraid that this is not going to change much in France.
And perhaps more importantly, that increased CO2 increases temperatures, which induces coral bleaching, and it increases the oceans' acidity, which compounds the problem. At least 20% of the world's coral reefs have already been destroyed by climate change and other human activity. Since we're not doing anything to mitigate climate change, Google Maps may well end up the only place where our grandchildren will be able to see coral reefs. So kudos to Google and CSS for at least saving the view.
Actually, after the acquisition Sun Microsystems, Inc. and Oracle USA, Inc. were merged to form Oracle America, Inc. So strictly speaking, Oracle is Sun. I wholly agree though that we need to know for how long they knew about this before passing judgement.
There are some interesting parallels here. Google is starting to look more and more like an operating system, with the menu bar at the top and the integration of a lot of their services into a desktop-like interface. And in a way, the "start menu" for this operating system is Google Search (it is after all the one at google.com). So the question then is, are they allowed to bundle other applications with this operating system, or should they allow others to compete with their own applications? In that sense it's similar to the whole Windows/IE bundling case. And in fact, Google could argue just like Microsoft did (although MS made some ridiculous claims about it being technically impossible to remove IE) that the embedded Maps is not a separate service at all, but that Search simply has an embedded viewer for search results that are geographical locations, which happens to be powered by the same technology as Maps.
Of course, what matters legally is the effect the thing has on the markets, not any kind of technical consideration. In that case, Google Search is a near-monopoly in the search market, and it's conceivable that its embedding of Google Maps to display results advantages Google Maps over other mapping services. I'm not sure how you would prove that (and have no idea what the standard of proof would be here), but if it turns out to be the case, then Google could remedy it by offering any other mapping services an open API that they can use to register their mapping service with Google, with Google then giving the user the option to choose a mapping service for showing embedded search results. That would be similar to the IE solution.
As for Google being evil, right now the EU is investigating if there is a crime at all. Antitrust law is a murky thing; there is no exact borderline where a market leader becomes a monopolist and where integrating services or products becomes too big a distortion of the market. So let's wait for the EU opinion first. Then, let's see how Google handles it. Will they work with the regulators to find an acceptable solution and implement it quickly, or will they try to lie, sue and lobby their way out of it like Microsoft did? I'd say that their reaction of a potential complaint constitutes a much better test of their character than just the fact that the EU has decided to investigate something.
As current species move toward the poles, more heat tolerant species are generated at the equator. And life backfills the change with more life. Such is at it has always been. Our dynamic world has never been stable, and should never be.
Heat tolerant species are generated? By whom? And more importantly, how quickly? Plants generally live quite long (annuals excepted obviously) and take quite a while to reproduce, especially in extreme circumstances. That means that evolution goes slowly. Climate change is currently happening extremely quickly. Hand waving doesn't make that problem go away.
The supposition of AGC has ever been that a static climate is a desirable thing, despite the fact that such a thing has never existed in the history of the Earth, and never will.
Humans haven't existed for most of the history of the Earth. Like all species we have a climatic niche that we fit into (hint: deserts, rain forests and polar regions are very sparsely populated). Now we're changing the planet to reduce the amount of space that falls within that climatic niche. Has the planet seen similar circumstances before? Sure, and it was fine. But what's important to us is that our species survives and thrives. Arguing that the planet will be fine is like saying that it doesn't matter when your only car breaks down halfway during your trip because the road will still be there. That's as correct as it is useless.
I can't tell where the factor of 2.6 comes from, it's not in the abstract and the article doesn't explain. I can't access the full article, but the abstract at least does not say how big the effect of BPA is relative to the factors that they controlled for. It may well be that the effect of say caloric intake is much larger still than that of BPA. And as the abstract states at the end, it may be that the obese children have a different mix of foods, something that wasn't controlled for. Perhaps the obese children are obese because their parents are more lazy and set them a bad example for getting some exercise, and the lazy parents also serve them more canned and packaged foods because they're too lazy to cook, so they have a higher BPA level.
It's also interesting to note that there is no significant difference between the second, third and fourth quartiles of urinary BPA, i.e. what they found was that (a bunch of things like ethnicity, sex, income, caloric intake, etc. being equal), if you're in the 25% of people with the lowest urinary BPA, you have about a 10% chance of being obese, while if you're in the 75% of people with highest urinary BPA then you have about a 20% chance of being obese. So perhaps a better way of summarising that result is to say that people with a low level of urinary BPA have, most other things being equal, only half the chance of being obese compared to people with an average or high level of urinary BPA. Perhaps the BPA acts as a kind of enabler, i.e. getting a certain minimum amount of it will change something in your biochemistry, but once you cross that threshold getting more of it doesn't make any more difference? Or perhaps it points to some statistical fluke?
And that the fashion photographer wasn't actually a fashion photographer but a professional skydiver? Say it ain't so!
If the phone can actually do image stabilisation and it's not much worse than what's shown in the ad (regardless of how the ad was produced) then I don't see how this is misleading beyond the fact that it's an advertisement, and thus by definition intentionally misleading. Lies, damned lies, statistics, and advertising...
Slashdot Mirror
The Graphene one. The others are just the usual BS from people clueless about how computers work and what they can and cannot do.
Spoken like a true programmer or sysadmin with no knowledge of statistics, modelling, machine learning or data analysis. I know, because I was one (and I still write code and maintain servers). But I've also moved into the above fields, and it's a completely different world. The discrete math and logic you use in programming are completely useless here, and the things you can do and the hurdles you come across are very different from the ones you see in programming. Of course, you still have to implement your models and analyses, and you get all the usual issues there (plus things like numerical instability), but even if the software is running fine you'll have things like parameter identifiability, difficulties in comparing models, lack of data in the places where you need it, conceptual problems with the models that can only be solved by making them more complex, which leads to lack of data problems and the need for massive amounts of compute power, and so on. These are the things they will be trying to tackle, and they have nothing to do with the limitations of Turing-style computers.
I do remain sceptical about having a chat with a Turing-level AI any time soon, but data analysis, modelling and inference methods are getting better and better (see Google Search, Watson) and I don't think that continued research into these things is a waste of money. Neither do Google, Facebook, Microsoft, the US government, and the EU apparently.
Finally, here's another EU project in this direction that is both scary and interesting.
In Kim Stanley Robinson's Mars trilogy, the northern ocean is filled with fresh water from the molten polar ice cap, while the rivers take up salt from the rocks they flow over, so there are salty rivers flowing into a fresh water ocean. I'm not sure how realistic that is, but it doesn't seem completely illogical.
As artist impressions go, I prefer this one, by Daein Ballard over the one in the article.
Maybe the Linux Foundation (or someone else, they're the first that come to mind) could do a similar thing to raise money for improving the Linux graphics and wireless stacks? How much improvement could we get for a million USD? Or perhaps there are individual developers out there who would do what Poul-Henning Kamp did? I'd be happy to contribute to such an initiative. Kickstart it?
I'm typing this on a Dell Latitude E6410, which is on that list (albeit with nVidia graphics I think, but Intel support is better, right? That's why I ordered it, anyway). When I first got this machine, it was also on the list, but Ubuntu 10.04 LTS (the most recent LTS) wouldn't even boot on it, just gave a black screen. Apparently there were multiple issues with the Intel graphics drivers, with both the E6410 and the E6510. Now it did seem that Canonical was giving those bugs some attention, but it still took many months for them to be fixed for most users. Then there was one last patch and it started working for me as well...until the next (ordinary, stable) update which broke it again. I ended up running 11.04 with a backported kernel that I didn't dare upgrade.
Then the touchpad (ALPS, not Synaptic) wasn't recognised as a touchpad, which they "solved" with a patch that sent a magic command combo to the device to switch it into imPS/2 mode. Result: scrolling worked, but it's still not recognised as a touchpad, and you still cannot configure it as such.
So yes, Canonical has a list, but I'd interpret that as "we'll try a bit harder to fix these machines, but they're otherwise just as broken as everything else", not as "tested and working".
Generally speaking, I'm seeing a lot of comments saying that there are no issues with server hardware. The OP didn't mention it, but it seems he/she was asking for desktop hardware, whose most critical functionality is (non-trivial) graphics hardware and, if it's a laptop, wifi. In my experience those things remain difficult for modern Linux kernels, especially on new hardware (i.e. anything you bought in the last year, maybe two), and there doesn't seem to be much progress either. As others have said though, Intel hardware seems to be your best bet.
How many false positives do you get though? In a classifier, having a high true positive rate is good, but only if it comes with a low false positive rate. It seems that in this case, perhaps there are a few too many false positives.
I've never played or even seen Minecraft, but I've heard of it and I could envision downloading this and having a look to see if it's interesting enough. Unfortunately, the local branch of the RIAA has sued my ISP into blocking the Pirate Bay, so I can't. Honestly, I was unlikely to buy this in the first place, but now I'm certainly not going to. Which I suppose serves 2 Player Productions right for daring to try to compete, even if it's in a small niche market, with the big media conglomerates. Still a shame though.
Whatever the motivation, I think this is one of the strengths of the scientific method and thus, one of the reasons for its success: we aren't quick to publishing until it is just right, and therefore, perhaps our best approximation of the "truth" we can muster.
On the other hand: What if you did publish your work early and often, not as concerned with slowly and deliberately ensuring everything is just right before spreading the information -- Not keeping quiet just so that you can be the one with the badge of "1st"? Why, then worldwide cooperation could kick in. Perhaps other interested parties would help you prove or disprove the results much more quickly. Thus, accelerating the speed of scientific progress.
You are assuming that scientists only ever communicate by publishing papers in journals. That's incorrect, there is a lot (and I mean a lot) of informal communication and collaboration, by email and phone and through presentations and posters at conferences. Our knowledge is now so vast and much research being done so multidisciplinary that it's nearly impossible for any single person to know enough to really cover all the aspects of a particular investigation. In my field, you'd need to be a good programmer, an expert statistician, an experienced and knowledgeable (field- and theoretical) biologist, and a good systems analyst/modeller. Such people don't exist, so work is done in teams with each member contributing their specific expertise. When you get a weird result, you go and talk to your colleague about it, and try to figure it out together, and you keep going together until you feel that you really understand what's going on. And then you write the paper, it gets published, and then hopefully it won't turn out to have been a fluke or a mistake or not representative of the wider area of research. If the whole team can't figure it out, you might publish a "Hey, that's weird?" paper, as was done here.
If people published everything they did immediately, we'd get so many publications that it would be impossible to keep up with all of it. The whole situation would be similar to the "Linus doesn't scale" problem in Linux kernel development a few years ago, where Linus Torvalds was inundated with patches and couldn't keep up. They solved that by appointing lieutenants, who filter and aggregate contributions. Publishing papers works the same way, you solve the smaller problems locally, and publish bigger and better-vetted results, so that everyone else doesn't waste their precious time on solving other people's small problems and consequently invalid results. Also, people wouldn't waste their precious time on writing up all those small problems, and peer-reviewing them, and so on. Writing a paper is not like writing a post on the Internet (something that some climate change deniers often conveniently forget), it takes serious time and effort by a group of people to make sure that the results are really of good enough quality. You don't want to waste that effort on trivial things.
I've got an MSc in CompSci, and I'm now doing a PhD in a biology department. I teach a programming course using Matlab, and I've recently started using R to do my own analysis stuff, mainly because it's popular and I'd like to stay compatible with the rest of the field, as well as use some specialised software that works with it. I have to say, being used to real programming languages (such as C++, I'm not counting Matlab here, although see below) I'm quite frustrated with R. Function names are generally different from other languages and to me at least unintuitive, and the documentation is too often extremely vague and difficult to search.
For example, the function match() returns the offsets of entries in a vector that match a given object. But what exactly constitutes a match, well, according to help(match) that is "to some extent a matter of definition". It goes on to give an example or two, but that definition remains elusive. Or look at this gem from help(as.vector):
First, a "vector of type list" is actually just a list. In R, a vector is an ordered collection of elements all of the same type, while a list is an ordered collection of elements of (possibly) different types. So, by the normal Liskov rules, one could say that a vector is a kind of list in which the types of the elements are all identical. According to the R language definition however, a list is a kind of vector. In practice, lists and vectors are used in rather different ways so their exact relation is not so relevant, and it doesn't make much sense for the help page to throw them together like this. Second, apparently attributes are removed for atomic vectors, but not "in general" for lists. This is a somewhat arbitrary inconsistency, and it leaves the reader to wonder if there are specific attributes or lists for which this doesn't count. But the kicker is in the last sentence: not only are the exact workings of this function explicitly undocumented, they are also subject to change without notice! Note that these are not functions from some obscure package that I pulled off of somebodies blog. They are core language functions, and unfortunately these examples are not exceptional. A colleague of mine recently had his whole analysis suddenly return weird results after a routine update of an add-on package, because someone decided to swap the order of the longitude and latitude arguments to a function for no particular reason.
That's not to say that R is not usable, but in my opinion is is unsuitable for any kind of programming, and perhaps unsuitable for programmers. R is a powerful, extensible system for statistical analysis, with a command line interface. If you consider your text files with R code as reference notes rather than as source files, and if you use R interactively, copy-pasting lines from your notes and checking after every couple of operations that it's actually doing what you think it is, then you can do useful things with R. Looking around me, that is in fact how most people use it, and what I've taken to doing as well, although I can't resist attempting to automate things here and there.
Comparing R to Matlab, in my eyes there's no contest in terms of ease of use. The Matlab help files are professionally written and tell you what you need to know in enough detail to be useful, and that difference alone makes it a lot better. The language itself is also a bit more sensible, at least to me, being designed as an easier-to-use alternative to FORTRAN, where R is based on LISP. I was originally considering moving my course to R from Matlab, since I don't like to teach proprietary software, but now that I have some experience with R I'm pretty sure
Well, it doesn't have to be technically the exact same set of bits, or even an exactly identical interface. What he seems to be getting at is that he wants to shift from a model of personal computing where you own a personal computer that you use in the privacy of your home and have full control of, to a model where you consume a set of interlinked services that are provided partially by devices you lease (e.g. a smart phone), partially by devices you own (tablet), and partially by servers run by third parties, with ultimate control of your actions and your data mostly in the hands of the service providers. This, of course, is the antithesis of the idea of Free Software.
Further down, he states "If your way of seeing the world IS genuinely more productive, effective, efficient, insightful and usable, then you should be confident that you will win in the long term" which is either naive or disingenuous, and incidentally a nice example of why RMS dislikes the term Open Source so much. Looking at the current state of GNU/Linux on the desktop and comparing that to Windows XP and Windows 7, I'm not so sure that GNU/Linux is any better (disclaimer, I have been running GNU/Linux exclusively since 1999, so I'm comparing between my machines and my coworkers'). But that is not the point, and that is not what the question was about. The question was about freedom, about controlling the computation and communication that is done on your behalf by your equipment. The fact that he sidesteps this question in much the same way that Jono Bacon sidestepped RMS's criticism of the new integrated Amazon search engine says a lot.
So here's one more attempt to get through to the fine folks at Canonical: the question is not about whether I want to be able to buy things from Amazon easily. The question is where else in the OS you are sending things that I type into my personal computer to some server on the Internet without me knowing about it, and how I can trust you to not do that without my knowledge or permission in the future? That is what Free Software is about. Unfortunately, Canonical seems to have a very different agenda.
I'd say this is like a cat leaping into the air to catch a bird, or an Osprey diving down into the water to catch a fish. Seems like many species happily leave their domain temporarily if there's food to be had. Still, interesting that these fish have picked up the idea (maybe it's the "cat" in their name?), and anything that gets rid of pigeons is a good thing :-).
There's another issue with the UN, which is that the International Telecommunications Union, the UN organisation that would take over control of the Internet, is run by the old national telephone companies. They're extremely conservative and would be very happy to grab control and kill all innovation, or at least slow it down a lot. See Mother Earth, Mother Board for how control of the undersea cable system that ties together most of the 'net was wrested from them (and because it's an excellent read). Imagine being only allowed to connect "approved equipment" to the Internet, the way it used to be with phone networks until not that long ago. It's bureaucracy that we should be afraid of, maybe more so than censorship (after all, the US has been seizing domain names as well).
Also, don't forget to go by Egan's homepage to play some quantum soccer. It has a collection of Java applets demonstrating various things from his books, and some free stories too. Seriously geeky!
And KDE has dedicated releases to deceased contributors as well. Why not? A small note in the release notes and perhaps in the About box would be the most tasteful option in my opinion, but it's your grandmother, so you decide.
Personally I am shocked that Microsoft is taking this action. It's the number-one most consumer-friendly thing I've ever heard of them doing -- literally.
It's pretty obvious to me: Google and Facebook, two very big Microsoft competitors, get most of their revenues from online advertising. Microsoft is doing this to damage them, not to promote their user's privacy (remember Windows Media Player phoning home every file played?). The only goal here is to make the services offered by Google and Facebook look more expensive by emphasising that while they seem free, you're paying with your privacy. Remember the whole Linux Total Cost of Ownership FUD campaign? Same old, except that people don't own licences to Google or Facebook software, so there's no ownership and they've had to approach it a bit differently. Meanwhile of course, everyone still pays Microsoft tax on every new computer they buy whether they want Windows or not, but let's not mention that...
I suspect that for most people, the reason to use a mind map rather than UML would not be UML's complexity. UML scales pretty well: a bunch of boxes with names in them and lines between them to signal what is related to what is a valid UML class diagram already. Instead, I think the problem with UML is that it forces you to think very carefully about what exactly it is that you're going to create. Back in the 1960's and 1970's, people like Donald Knuth and Edsger Dijkstra advocated careful thinking about software, rigid specification, and proving correct any important algorithms. They saw software as a mathematical construct, and the exercise of building software as akin to proving theorems.
Fast forward to the Internet age. Software is everywhere, and rare is the project where the customer can tell you clearly what they want. The small cadre of people who are capable of the precise and abstract thought required to do programming the mathematical way is not by far big enough to write all the software that the world needs, so even if customers could make rigid specifications, most programmers would find them written in an alien language. So we have adopted a biological rather than mathematical approach: specifications are never exact, software is always broken, but it's okay because the software has an immune system (we call it vendor support), which fixes up errors continuously. In such a world, maybe a mind map is as formal a description as you need.
Personally, I used UML to describe a logical data model in my last big project. I was the only one with a formal CS background in the project, but everyone understood the diagrams just fine. I had to explain a few more advanced things to some people, but it was no problem. And we did think everything through very carefully, and so far the whole thing is holding up very well because of that. In my opinion, even in the age of agile development and web technologies, careful thought is still invaluable in software development, and a diagram language that lets you specify a bit more detail when you need it is a very useful tool. I'll stick with UML.
That's not quite right. If the cyclist is 14 or younger, the car driver('s compulsory insurance) is 100% liable for all damages due to the accident. If the cyclist is over 14, they're presumed to have some responsibility for avoiding an accident as well, and the car driver's liability is at least 50% and at most 100%, depending on the circumstances. Note that that is only about the (civil) damages, any criminal negligence or reckless driving charge or some such is dealt with separately.
My girlfriend was hit by a taxi while riding her bike in Amsterdam a couple of years ago. The driver was distracted by the passenger, and was (legally, but it was literally his first time doing so) driving in the tram lane. Road side obstacles and the curvature of the road (it was on a relatively small but high bridge) kept them from seeing each other until it was too late. The prosecutor's office decided not to prosecute the criminal case as it was mostly just the dangerous crossing that was at fault, and the taxi was not speeding. After a couple of letters from my girlfriend's lawyer the car driver's insurance did assume 100% liability, as they were the more guilty party.
In the end, what this law mostly means is that damages from car/bike or car/pedestrian accidents are paid for collectively by car drivers, through their insurance premiums. And it results in cyclists always getting the right of way, generally ignoring traffic lights, and sometimes riding quite recklessly, especially here in Amsterdam. You get used to it...
(My girlfriend recovered well and still doesn't wear a helmet. We never did find out if the windscreen survived.)
The difference is that when Bain did it, it was because opening jobs overseas was the only way to save a company. When Obama did it, it was for.... Actually, I have no idea why Obama would invest in Brazilian oil and not Gulf of Mexico oil. Of course, like I said, Obama shipped jobs overseas with YOUR money against your will. Romney did not.
The Netherlands has quite a substantial amount of natural gas in fields in the north of the country. We've been exploiting those fields, but not as much as we could, and we're buying gas from abroad to fulfil the rest of our need. Essentially, we're treating it partially as a strategic reserve; when fossil fuels run out, we'll have a buffer supply to get us through a transition to other types of energy. Gas production is done by a commercial company, but they pay the government for the privilege, and that money ends up in a special fund for infrastructure investment, i.e. new motorways, train lines, and even some for improvements in education as well IIRC.
Anyway, it may not be a bad idea at all for the US to save some of its domestic supply for later, and get oil from elsewhere while those other countries are still willing to sell. Obama strikes me as smart enough to think ahead strategically. Or else the oil is Brazil is pumped up by an American company and they lobbied for some more subsidies...
That's an interesting case of Prisoner's Dilemma there. If everyone is anonymous, nobody can stalk anybody, but there will be relatively many rude people. If only some people are anonymous (and potentially rude) and others aren't (and are civil), then the civil folks run the risk of being stalked by anonymous rude people. If nobody is anonymous, nobody can stalk anybody, at least not while escaping punishment, and we'll have fewer rude people.
In the Prisoner's Dilemma, there are several solutions to this. One is having an external authority that keeps both prisoner's from talking. Another strategy that works well (in the sense of fostering cooperation) is tit-for-tat with random altruism, where in each round of Prisoner's Dilemma you cooperate if the other person cooperated with you in the previous one, and betray if they betrayed you, except that you occasionally give them a break and cooperate even if they betrayed you. This latter feature keeps the system from getting stuck in endless loops of betrayal. The analogues in the case of Internet anonymity would be a real-name policy, and letting users selectively share their personal information with others. Note that for the user to follow the tit-for-tat strategy, users have to know that it's the same person who betrayed them last time, so you can't have multiple accounts per person (another reason for a real-name policy). Of course, there are also good valid uses of anonymity and anonymous speech, and there should be a place for them as well.
Exactly, and that is why the solution that Traynor chose is the right solution, and it's quite admirable that he did so. But even if all he would consider would be self-interest, this would have been the correct solution. Prisons and other punishment don't solve the problem of crime in a society. They just repress the symptoms, and they deflect and channel feelings of revenge into something more civilised than a lynching. If you actually want to solve the crime problem, you have to prevent people from becoming criminals, because once they are there is very little you can do to turn them into good citizens again. That means keeping a very close eye on 12-18 year olds who are running with the wrong crowd, who live in unstable families, and so on, and intervening where necessary. Keep them on the straight during puberty and adolescence, make sure they finish school and find a job and start a family, that's basically all you have to do (see also the recent find-hackers-girlfriends item). And for the ones for which it's too late, and the ones with mental disorders, well, so far prisons and closed institutions is the best we can do it seems. But if Traynor had had this kid sent to prison, the kid would have been out in a couple years, and would probably have got right back into stalking people or worse. The path he chose to take gives Traynor a much better chance of never having to deal with this again.
Two things:
1) "Wherever Possible" means that they know FOSS doesn't have a solution to every little problem. However, if it comes down to word processors, databases, web browsers, etc then there are numerous FOSS alternatives.
The Netherlands' government has been operating on a "comply or explain" principle for years. All government agencies are required to use open source software and free standards, or else explain why they don't. All the government agencies I've seen in the past couple of years (municipalities, provinces, and a couple other government and semi-government organisations) use Microsoft software everywhere, with the exception of the databases, for which they use Oracle. Spatial planning is done with the proprietary ESRI stack. The only open source is usually the CMS they use for the web site, either an existing one or a home grown one which they open sourced themselves, and they have an ODF plugin installed in Word so that they can fulfil their legal requirement to be able to communicate using the ODF standard. Of course, everyone uses doc and docx.
I think the main reason is that they simply don't employ any real IT staff, just a few technicians who know how to swap out a machine and which phone number to call the supplier on when something breaks. It's difficult to find people who, given a bunch of open source software, can actually fix things themselves, and those people are expensive. Getting external support for FOSS is also not easy unless it's for something extremely mainstream. The FOSS GIS stack is getting quite capable for example, but I think there are only a handful of companies world wide who offer support for the thing, and they're all pretty small and on the other side of the world from here. So ESRI and Oracle Spatial it is.
So, which the initiative is great, and all sovereign governments should be using Free software on general principle, I'm afraid that this is not going to change much in France.
And perhaps more importantly, that increased CO2 increases temperatures, which induces coral bleaching, and it increases the oceans' acidity, which compounds the problem. At least 20% of the world's coral reefs have already been destroyed by climate change and other human activity. Since we're not doing anything to mitigate climate change, Google Maps may well end up the only place where our grandchildren will be able to see coral reefs. So kudos to Google and CSS for at least saving the view.
Actually, after the acquisition Sun Microsystems, Inc. and Oracle USA, Inc. were merged to form Oracle America, Inc. So strictly speaking, Oracle is Sun. I wholly agree though that we need to know for how long they knew about this before passing judgement.
There are some interesting parallels here. Google is starting to look more and more like an operating system, with the menu bar at the top and the integration of a lot of their services into a desktop-like interface. And in a way, the "start menu" for this operating system is Google Search (it is after all the one at google.com). So the question then is, are they allowed to bundle other applications with this operating system, or should they allow others to compete with their own applications? In that sense it's similar to the whole Windows/IE bundling case. And in fact, Google could argue just like Microsoft did (although MS made some ridiculous claims about it being technically impossible to remove IE) that the embedded Maps is not a separate service at all, but that Search simply has an embedded viewer for search results that are geographical locations, which happens to be powered by the same technology as Maps.
Of course, what matters legally is the effect the thing has on the markets, not any kind of technical consideration. In that case, Google Search is a near-monopoly in the search market, and it's conceivable that its embedding of Google Maps to display results advantages Google Maps over other mapping services. I'm not sure how you would prove that (and have no idea what the standard of proof would be here), but if it turns out to be the case, then Google could remedy it by offering any other mapping services an open API that they can use to register their mapping service with Google, with Google then giving the user the option to choose a mapping service for showing embedded search results. That would be similar to the IE solution.
As for Google being evil, right now the EU is investigating if there is a crime at all. Antitrust law is a murky thing; there is no exact borderline where a market leader becomes a monopolist and where integrating services or products becomes too big a distortion of the market. So let's wait for the EU opinion first. Then, let's see how Google handles it. Will they work with the regulators to find an acceptable solution and implement it quickly, or will they try to lie, sue and lobby their way out of it like Microsoft did? I'd say that their reaction of a potential complaint constitutes a much better test of their character than just the fact that the EU has decided to investigate something.
As current species move toward the poles, more heat tolerant species are generated at the equator. And life backfills the change with more life. Such is at it has always been. Our dynamic world has never been stable, and should never be.
Heat tolerant species are generated? By whom? And more importantly, how quickly? Plants generally live quite long (annuals excepted obviously) and take quite a while to reproduce, especially in extreme circumstances. That means that evolution goes slowly. Climate change is currently happening extremely quickly. Hand waving doesn't make that problem go away.
The supposition of AGC has ever been that a static climate is a desirable thing, despite the fact that such a thing has never existed in the history of the Earth, and never will.
Humans haven't existed for most of the history of the Earth. Like all species we have a climatic niche that we fit into (hint: deserts, rain forests and polar regions are very sparsely populated). Now we're changing the planet to reduce the amount of space that falls within that climatic niche. Has the planet seen similar circumstances before? Sure, and it was fine. But what's important to us is that our species survives and thrives. Arguing that the planet will be fine is like saying that it doesn't matter when your only car breaks down halfway during your trip because the road will still be there. That's as correct as it is useless.
I can't tell where the factor of 2.6 comes from, it's not in the abstract and the article doesn't explain. I can't access the full article, but the abstract at least does not say how big the effect of BPA is relative to the factors that they controlled for. It may well be that the effect of say caloric intake is much larger still than that of BPA. And as the abstract states at the end, it may be that the obese children have a different mix of foods, something that wasn't controlled for. Perhaps the obese children are obese because their parents are more lazy and set them a bad example for getting some exercise, and the lazy parents also serve them more canned and packaged foods because they're too lazy to cook, so they have a higher BPA level.
It's also interesting to note that there is no significant difference between the second, third and fourth quartiles of urinary BPA, i.e. what they found was that (a bunch of things like ethnicity, sex, income, caloric intake, etc. being equal), if you're in the 25% of people with the lowest urinary BPA, you have about a 10% chance of being obese, while if you're in the 75% of people with highest urinary BPA then you have about a 20% chance of being obese. So perhaps a better way of summarising that result is to say that people with a low level of urinary BPA have, most other things being equal, only half the chance of being obese compared to people with an average or high level of urinary BPA. Perhaps the BPA acts as a kind of enabler, i.e. getting a certain minimum amount of it will change something in your biochemistry, but once you cross that threshold getting more of it doesn't make any more difference? Or perhaps it points to some statistical fluke?
And that the fashion photographer wasn't actually a fashion photographer but a professional skydiver? Say it ain't so!
If the phone can actually do image stabilisation and it's not much worse than what's shown in the ad (regardless of how the ad was produced) then I don't see how this is misleading beyond the fact that it's an advertisement, and thus by definition intentionally misleading. Lies, damned lies, statistics, and advertising...