At some point, even with fully open, GPL-compliant software, there is some point you just have to trust someone else to not jack you.
Absolutely. I use computers with closed-source chip design and BIOS. And even though I can read source code, I obviously have not inspected every line of code on my Linux system. And even with scrutinized code, my compiler could be compromised, as you noted.
But, as with all security and trust questions, the objective isn't to achieve perfect security/trust (which is impossible) but to achieve a level of security/trust that is as high as possible (or as high as the application demands).
The more open the code is, and the more people who look at it, the more trust I will put in it. I don't have to inspect the code myself every time. If a trusted third-party trusts the code (e.g. Ubuntu maintainers put it into their repositories), that will increase my trust in it. Perfect trust is impossible. But, all other things being equal, I have more reasons to trust open and scrutinized code than I do to trust closed proprietary code that only a handful of people have ever looked over.
Agreed. The statement in the summary "...the project managed to show how trusting the local law enforcement agencies really were..." infuriates me. Police are not supposed to be harassing people left and right, trying to uncover illegal or just unsanctioned activities. The police were friendly, waved, and didn't bother to investigate something that by all rights did not look overtly illegal. They acted appropriately.
I would much prefer that law enforcement err on the side of trust and friendliness. This probably means that some fraction of illegal actions will go undetected and unpunished (note that only a small fraction of those illegal actions are truly dangerous and unethical)... but that is the 'price' of freedom.
Again, I applaud the police for not flipping out when they see people engaging in activities that they don't exactly understand (but for which there is no evidence of illegal action).
I agree that people shouldn't mark as spam things they voluntarily signed up for (unless attempts to remove oneself from the list fail).
However, I think this also points out a way in which email could be made better. There should really be a standardized way to unsubscribe from mailing lists, so that every mail client automatically shows an "unsubscribe" button inside any mailing list email. The problem with current unsubscribe methods is that they require too much effort (even clicking a few links is "too much effort" in comparison to the "spam" button... moreover many sites make you go through numerous confusing web-forms). Also, an integrated "unsubscribe" button in an email client would send the "please unsubscribe" signal, and simultaneously add the address to a personal blacklist (but not add it to the spam detection list).
If you make it easy for people to use, then they will. The present problem arises largely from people's laziness. But you can't prevent people from being lazy, so instead the tools should adapt to people's common usage.
For example, a lot of research is being done now utilizing ultra-high pressure water as a replacement for organic solvents, for greener chemistry.
I think you mean ultra-high pressure carbon dioxide, not water. Supercritical CO2 is indeed an interesting area of research, as it can be used to replace dangerous organic solvents, making industrial chemistry safer and greener.
And I agree that there is likely a rich unexplored landscape of interesting chemistry beyond standard temperatures and pressures.
I would also argue that it wasn't ever really a dogma of chemistry. It was more of a useful approximation.
Chemists have long known that all the electrons contribute in some way to interactions. However it is a very useful approximation to say that only the outer electrons contribute significantly to bonding interactions. The fact is that they greatly dominate all such interactions, making the approximation useful both conceptually and computationally.
But, we all know that strictly when two atoms interact it should be modeled by taking into account the many-body problem formed by all the electrons and nucleons in both atoms. This new result is interesting in that they demonstrate a case where the contribution from the inner electrons to the final bonding and physical properties is much higher than for other systems. But I don't see how this violates any previously-held scientific principle.
Another example I like is Blender (the open-source 3D modeling platform). Blender was originally a closed-source commercial product. When the company went bankrupt, the creditors agreed to release the code under the GPL for a one-time payment of $100,000. A donation campaign was started and raised the required funds. So now Blender is open-source, and has been extended and enhanced remarkably in the years since its release.
The reason I like this example is it shows that you can get paid to write code that becomes open-source. It also shows that the community is indeed willing to pay for free/open-source software. It's probably too much of a 'gamble' to become a common business model (write software, get people interested in it, state a bounty for its open-source release)... but it's an option more companies should consider when decommissioning software.
The community will pay for open-source, if it's something good.
Like it or not, copyright doesn't apply to bit-streams, but rather to particular instantiations of ideas (and derivatives thereof). No one can copyright a number. But in a particular context, a certain bit-stream can be considered protected by copyright.
This whole "you can't copyright a number!" is a red herring. No one seriously claims that particular numbers are copyrighted. But in a certain context, a particular chunk of data (a number!) can be reasonably shown to be a copy (or derivative) of a particular copyrighted work. If the same number appears in a totally unrelated context, and it's apparent that it is not being used to distribute a copyrighted work, then no court would find that instance of the number to be infringing.
Another way to say this is that copyright law is more concerned with the action of copyright violation (distributing a copy or derivative of a work without authorization), and is not concerned with maintaining a catalog of copyrighted bit-streams.
In fact there is some research that suggests for certain kinds of decisions, more thought is actually counter-productive. That is, initial "gut" decisions are sometimes more optimal than carefully-considered ones (where "optimal" is measured by longer-term happiness/regret of decision). (For instance, check this writeup of this paper, or the associated Slashdot submission.)
The point is that while thinking long and hard about some problems can be helpful (e.g. designing something complex and technical), for other kinds of problems, added thought can hinder (e.g. when there are many confounding unknowns).
I don't get your comment.
Why would we care if a utility site like Sourceforge is blocked? Who is "we"? If "we" is the Slashdot readership, then clearly some of "us" care, since some portion of Slashdot readers are in China, and/or use/contribute to Sourceforge.
It saves us the hassle of providing the network traffic to that country What? So Sourceforge should be happy that an entire country can't access their services, because that reduces network traffic? By that logic, Sourceforge should just shut down completely: then the network traffic would be really reduced! In reality, Sourceforge wants people to go to the site, and so having an entire country blocked is bad for them.
do we really want to answer 'Chinese' informed questions? I'm sorry if I've misinterpreted this question, but on the face of it, it seems racist.
I wonder how much Chinese projects are hosted on Sourceforge "Chinese projects"? You do realize that Chinese people use the same kinds of software as everyone else, right? And that Chinese coders can (and do) contribute to the same open-source projects as everyone else, right?
I guess I don't understand your comment because you seem to be saying "good riddance!"... but why should the open-source community be happy that a government firewall is fracturing the community?
Perhaps you're thinking of wine the wrong way. It is, first and foremost, a windows-compatible API for porting applications to posix. Actually that brings up a question I'd like to ask the Wine developers:
As I understand it, Wine was originally intended to be both (1) a set of libraries that Windows developers could recompile their code against to run on other operating systems; and (2) a compatibility layer to run unmodified Windows binaries on other operating systems. Which one was the "primary" intent of Wine originally?
Also, nowadays, it seems that the vast majority of people use Wine in mode (2). Few developers have used the Wine libraries to recompile their code. Is this a fair assessment? If so, how does this affect the way you develop the Wine codebase? Do you see this changing in the future?
Microsoft doesn't have any grounds for suing. The codebase was written from scratch (so no copyright issues), and if Wine infringes on Microsoft patents, then so does OS X, Linux, BSD, etc. Would you say that all of those are wasted projects, since they are going to be sued and shutdown?
Also, even if Wine suddenly disappeared tomorrow, it still would not have been a waste. It has taken 15 years for Wine to get to where it is now, but it was being actively used during those 15 years. Tens of thousands of people have been successfully using Wine to get their work done for over a decade. That's a success right there. Moreover, the developers no doubt have found Wine very useful over the years... hence why they continued working on it.
If Wine is a "waste", then so is every long-term software project.
I would argue that Wine is much more convenient than virtualization... when it works, that is.
When you run an app in Wine, it integrates (more or less) with your current desktop environment. It immediately has access to the same folder hierarchy. It also performs better (loading the wine libraries seems to have a lower overhead than loading a VM and an OS).*
The only downside to Wine is that not every app runs, and some apps run but are a bit buggy.
So I would say that Wine wins for convenience, whereas virtualization wins on "robustness": any app that runs on Windows will run on Windows in a VM. This is why I use both Wine and virtualization on my system: for most apps, I can just use Wine and it's treated like just another application. For those that don't work well in Wine, I can always open up the VM.
([*] Another aspect of performance to consider is things like hardware acceleration. Most VMs don't take advantage of 3D acceleration, whereas Wine in principle can.)
it's such a rambling mess it's hard to know where to start picking it apart. Agreed. I want to do a line-by-line rebuttal... but I fear that would be a waste of time.
The article does not make a compelling point. It keeps saying that we can give up on models (and science), because now we just have lots of data, and "correlation is enough." What utter BS. Establishing a correlation is not enough. Even if it is predictive for the given trend, it doesn't allow us to generalize to new domains the way a well-established scientific model does. If an engineer is designing a totally new device, that goes above and beyond what any established device has done, what data can he draw upon? If there is no mountain of data, he must rely on the tried-and-true techniques of engineering/science: use our best models, and predict how the new device/system will behave.
The article actually makes this point perfectly clear when it says:
Venter can tell you almost nothing about the species he found.
Indeed. Merely having tons of data doesn't actually give you insight into what you have measured. You must distill the data, pull out trends, and construct models. I just don't see how have mountains of data about a species, but still being unable to answer simple questions about it, is superior to conventional science (which can answer questions about the things it has discovered).
A deluge of data and data-mining techniques is a boon to science. But I don't see the benefit of giving up on the remarkably successful strategy of constructing models to explain the phenomena we've observed. I somehow doubt that having 20 petabytes of data on electron-electron interactions is more useful than having a concise theory of quantum mechanics.
Yup. That myth has been thoroughly debunked, yet it still gets repeated.
The article is full of meaningless or incorrect statements. Like:
Royall is part of a group of scientists who think that if you wait long enough, perhaps billions of years, all glass will eventually crystallize into a true solid. In other words, glass is not in an equilibrium state, (although it appears that way to us during our limited lifetimes).
As a researcher in the field, I can assure you that this isn't a controversial statement. We all agree that glasses are not at equilibrium. We all agree that the low-energy state for glasses is to crystallize, and that (in principle), if you wait long enough they will crystallize. The questions revolve around details like "how far from equilibrium?", "what are the implications of being non-equilibrium (e.g. on phase transitions)?", "what are the kinetics and dynamics?", "how long would it ~actually~ take for a given amount of change/flow/reconstruction/etc.?"...
Also, equating "equilibrium" with "being a solid" is total nonsense. (Solids, liquids, and gases can all be at equilibrium or far from equilibrium...)
In short, don't waste your time with this ridiculously hyped review of some otherwise interesting (but not revolutionary) science.
Yes, in theory. Just as the sun will rise tomorrow "in theory." And if I repeatedly shoot someone in the head, they will die, "in theory." And reality exists, "in theory."
Provability only exists in mathematics. For everything else, from decisions about what to buy at the supermarket, to designs of scientific experiments, we humans must use mental models that rely upon fundamental assumptions about how the universe operates (e.g. that past experiences allow us to make meaningful predictions). In other words, every action we take must be informed by some sort of "theory." The question then becomes "how robust is this model/theory?", "how much can I trust the predictions?", "what is the range of the possible outcomes?", "what are the consequences of errors in the assumptions/model/theory?", and so on.
If you have a specific problem with one of the assumptions, logic, modeling, mathematics, data acquisition, or analysis, then point it out in detail. But saying, "that's just a theory" is not useful. Everything we do is based on theories.
After all, the opposite is also a theory: Not turning on the LHC won't cause the destruction of the Earth... in theory.
...none of their clips are available... (short, low-quality, low resolution) videos... By the way, in addition to short clips from shows, Hulu has full-length TV shows and even movies on the site. The quality/resolution isn't great (although comparable to conventional TV, I guess). They are also starting to phase-in "HD" versions of some of the shows/movies. The resolution is good (1280x720)... but it suffers from the usual drawbacks of buffered video (either it's jerky or you have to wait for it to buffer a bit...).
I'm not saying the Hulu viewing experience is fantastic... just making it clear that they are have lots of full-length content on the site.
I agree with you, by the way, that Hulu is silly to ignore the international market. This may result from their various licensing deals (in which case the content owners should get a clue).
Access denied: remote loader detected.
This request has been identified as coming from a remote-loading website. This is not Wikipedia, please update your bookmarks. Access Wikipedia only through *.wikipedia.org.
A remote loader is a website that loads content from another site on each request. The content is typically filtered, framed with ads, and then displayed to the user.
The remote loader either:
Pretends to be the source website,perhaps using a deceptive domain name; or
Converts all instances of the name of the source website to some other name.
We consider remote loading websites to be an unfair drain on our server resources, and so they are systematically blocked, as this one has been.
So, obviously this site was fetching Wikipedia content in real-time, and sticking in ads and whatnot (rather than using their own local copy of the Wikipedia database). This is obviously a silly drain on Wikipedia's servers.
Moreover, this is a stupid way to design it, since it's trivial for Wikipedia to detect what you're doing, and serve a custom error page, as they have done. In short, why did these people assume Wikipedia was going to let them continue infringing their trademark and taxing their servers?
Except that #2 (and therefore #3) is rather unlikely.
If this new site doesn't provide anything above-and-beyond what Wikipedia provides, then few people will link to it, and its PageRank will be low. Without ranking high on Google, no one will find the site, and their ad revenue will be pathetic.
So, I don't really understand their business model here. Unless they offer some "value added" over the normal Wikipedia (quicker load times, vetted articles, better search, etc.), then they can't hope to attract eyeballs to their adds.
Forking is fine. A crappy fork, however, won't attract interest, and won't last long.
I've always wanted a realtime graphics engine based on something like the POV-ray ray-tracer (or other procedural modeling). The POV-ray syntax is all "exact". Rather than approximating shapes using subdivision into triangles, exact shapes are created by specifying things like "spheres" or "cylinder" or unions, intersections, and differences thereof. More complex objects can be specified by arbitrary mathematical equations, and complex sequences of operations (e.g. take a spline, sweep it along a path, intersect it with another shape, apply a certain matrix transform,...). Having done some modeling both ways, I much prefer the "exactness" of procedural definitions, rather than approximation. (I inevitably wish I could go back and add resolution to a triangulation, but that isn't easy to do properly.)
The neat thing is that the resulting objects (if properly defined) have "infinite" detail. The roughness on a surface, for instance, can be based on a noise function, so you can zoom into it without ever seeing triangulation or other artifacts.
The obvious downside is that the computation here is intensive. Objects can be arbitrarily complicated. Calculating the intersection of a ray with a mathematically-defined surface involves very complex calculations. Rendering POV-ray scenes on modern hardware, for instance, can take minutes to days (depending on complexity).
One upside is that the rendering can be tuned to available resources. On older hardware, the number of light-sources (or the intersection accuracy, etc.) can be reduced. This would mean that video game graphics would get arbitrarily "better and better" on newer hardware, without any need for someone to change the code. Having said all this... I think our hardware is not yet powerful enough to make this kind of thing practical. (There are some neat examples that have been coded, but as a general technique we're not there yet.)
His analysis is in many ways good... but seems ridiculously idealistic. He emphasises:
Where do we turn when we need enhancements to Internet protocols and the applications that use them? Not to Congress, and not to the FCC.... Engineers solve engineering problems.
(Emphasis in original.)
Probably most of us agree with that statement in principle. The problem is that the various players in this (users, content providers, and network operators) do not have their objectives aligned. Thus, the engineers for the network operator will come up with a solution (e.g. throttling) that solves the network company's problem (users using too much of the bandwidth they (over)sold), but the engineers working for the users (e.g. people writing P2P apps) will engineer for a different objective (maximum transfer rates), and will even engineer workarounds to the 'solutions' being implemented by the network.
The problem is thus that everyone is engineering in a fundamentally adversarial way, and this will continue so long as the objectives of all parties are not aligned. Ideally, legislation would help enforce this alignment: for instance, by legally mandating an objective (e.g. requiring ISPs to be transparent in their throttling and associated advertising), or funding an objective (e.g. "high-speed access for everyone"), or by just making illegal one of the adversarial actions (e.g. source-specific throttling).
This is not purely an engineering question. The networks have control of one of the limited resources in this game (the network of cables already underground; and the rights required to lay/replace cables), and this imbalance in power may require laws to prevent abuse. It's not easy to create (or enforce) the laws... and ideally the laws would be informed by the expertise of engineers (and afford ways for smarter future solutions to be implemented)... but suggesting that we should just let everyone 'engineer' the solution misses the mark. Whose engineers? Optimizing for what goal? Working under what incentives?
Put more simply: engineering is always bound by laws.
as this contest shows, it is possible to include backdoor behavior "in the source for everyone to see" without it being discovered.
That's one way to look at it.
Another way to look at it is that this is a (somewhat whimsical) way to test the limits of hiding malicious code in open-source code. This contest, in a sense, is part of the transparency and security of the open-source method. Everyone knows that you can quite easily hide malicious code in a closed-sourced project. But this contest gives the open-source community a chance to see the limits of open-source.
The participants in the contest (including entrants, judges, and interested observers) will all learn valuable lessons about code security, and will have a better idea of how to spot malicious code in plain site, and how to structure programs to avoid malicious or erroneous code contributions.
So, far from showing that the OSS method can't help in avoiding backdoors, I view this as part of the process of avoiding backdoors.
Oh, and note to self, don't download any open source image editing software in the future...
And you really better not install any closed-source image editing software, since finding malicious code in that case is a thousand times harder.
The promoters don't have much choice. Either they:
Send unsolicited promotional copies to various parties. But unsolicited merchandise through the mail is presumptively a gift. So the people will then own the promotional copy, and have the right (as this recent ruling affirms) to resell the item. It would be fraudulent to mail things that say "you must pay for this!" or even "you must return this!" (thereby forcing people to pay postage for things they never requested).
First send out a letter asking the recipient to sign a contract, where they agree that upon receiving the promotional item, they will return it, and agree not to resell or give it to others. This would work, but would be onerous for both parties. In particular, the vast majority of people wouldn't bother with that kind of hassle.
The promoters need to make it easy for the intended recipients to get the promo and to hear it. So I doubt they will use the second option (although I suppose a large company could try to get many outlets to sign a blanket contract that will cover all promos thereafter sent to them). In the end, they will just have to accept the fact that the promotional material can be transferred, bought, and sold, like anything else.
Slashdot Mirror
At some point, even with fully open, GPL-compliant software, there is some point you just have to trust someone else to not jack you.
Absolutely. I use computers with closed-source chip design and BIOS. And even though I can read source code, I obviously have not inspected every line of code on my Linux system. And even with scrutinized code, my compiler could be compromised, as you noted.
But, as with all security and trust questions, the objective isn't to achieve perfect security/trust (which is impossible) but to achieve a level of security/trust that is as high as possible (or as high as the application demands).
The more open the code is, and the more people who look at it, the more trust I will put in it. I don't have to inspect the code myself every time. If a trusted third-party trusts the code (e.g. Ubuntu maintainers put it into their repositories), that will increase my trust in it. Perfect trust is impossible. But, all other things being equal, I have more reasons to trust open and scrutinized code than I do to trust closed proprietary code that only a handful of people have ever looked over.
Agreed. The statement in the summary "...the project managed to show how trusting the local law enforcement agencies really were..." infuriates me. Police are not supposed to be harassing people left and right, trying to uncover illegal or just unsanctioned activities. The police were friendly, waved, and didn't bother to investigate something that by all rights did not look overtly illegal. They acted appropriately.
I would much prefer that law enforcement err on the side of trust and friendliness. This probably means that some fraction of illegal actions will go undetected and unpunished (note that only a small fraction of those illegal actions are truly dangerous and unethical)... but that is the 'price' of freedom.
Again, I applaud the police for not flipping out when they see people engaging in activities that they don't exactly understand (but for which there is no evidence of illegal action).
I agree that people shouldn't mark as spam things they voluntarily signed up for (unless attempts to remove oneself from the list fail).
However, I think this also points out a way in which email could be made better. There should really be a standardized way to unsubscribe from mailing lists, so that every mail client automatically shows an "unsubscribe" button inside any mailing list email. The problem with current unsubscribe methods is that they require too much effort (even clicking a few links is "too much effort" in comparison to the "spam" button... moreover many sites make you go through numerous confusing web-forms). Also, an integrated "unsubscribe" button in an email client would send the "please unsubscribe" signal, and simultaneously add the address to a personal blacklist (but not add it to the spam detection list).
If you make it easy for people to use, then they will. The present problem arises largely from people's laziness. But you can't prevent people from being lazy, so instead the tools should adapt to people's common usage.
For example, a lot of research is being done now utilizing ultra-high pressure water as a replacement for organic solvents, for greener chemistry.
I think you mean ultra-high pressure carbon dioxide, not water. Supercritical CO2 is indeed an interesting area of research, as it can be used to replace dangerous organic solvents, making industrial chemistry safer and greener.
And I agree that there is likely a rich unexplored landscape of interesting chemistry beyond standard temperatures and pressures.
I would also argue that it wasn't ever really a dogma of chemistry. It was more of a useful approximation.
Chemists have long known that all the electrons contribute in some way to interactions. However it is a very useful approximation to say that only the outer electrons contribute significantly to bonding interactions. The fact is that they greatly dominate all such interactions, making the approximation useful both conceptually and computationally.
But, we all know that strictly when two atoms interact it should be modeled by taking into account the many-body problem formed by all the electrons and nucleons in both atoms. This new result is interesting in that they demonstrate a case where the contribution from the inner electrons to the final bonding and physical properties is much higher than for other systems. But I don't see how this violates any previously-held scientific principle.
Another example I like is Blender (the open-source 3D modeling platform). Blender was originally a closed-source commercial product. When the company went bankrupt, the creditors agreed to release the code under the GPL for a one-time payment of $100,000. A donation campaign was started and raised the required funds. So now Blender is open-source, and has been extended and enhanced remarkably in the years since its release.
The reason I like this example is it shows that you can get paid to write code that becomes open-source. It also shows that the community is indeed willing to pay for free/open-source software. It's probably too much of a 'gamble' to become a common business model (write software, get people interested in it, state a bounty for its open-source release)... but it's an option more companies should consider when decommissioning software.
The community will pay for open-source, if it's something good.
You're quite right.
Like it or not, copyright doesn't apply to bit-streams, but rather to particular instantiations of ideas (and derivatives thereof). No one can copyright a number. But in a particular context, a certain bit-stream can be considered protected by copyright.
This whole "you can't copyright a number!" is a red herring. No one seriously claims that particular numbers are copyrighted. But in a certain context, a particular chunk of data (a number!) can be reasonably shown to be a copy (or derivative) of a particular copyrighted work. If the same number appears in a totally unrelated context, and it's apparent that it is not being used to distribute a copyrighted work, then no court would find that instance of the number to be infringing.
Another way to say this is that copyright law is more concerned with the action of copyright violation (distributing a copy or derivative of a work without authorization), and is not concerned with maintaining a catalog of copyrighted bit-streams.
How about:
Not a solution to defeat ISPs attempts to control what's going through the government-funded, monopoly-protected, public-land-using network.
You're right, facts do change the interpretation.
The group is called "I Text Message People While Driving and I Haven't Crashed Yet!"
:)
Seems like the police should keep an eye on people who suddenly leave the group.
In fact there is some research that suggests for certain kinds of decisions, more thought is actually counter-productive. That is, initial "gut" decisions are sometimes more optimal than carefully-considered ones (where "optimal" is measured by longer-term happiness/regret of decision). (For instance, check this writeup of this paper, or the associated Slashdot submission.)
The point is that while thinking long and hard about some problems can be helpful (e.g. designing something complex and technical), for other kinds of problems, added thought can hinder (e.g. when there are many confounding unknowns).
I guess I don't understand your comment because you seem to be saying "good riddance!"... but why should the open-source community be happy that a government firewall is fracturing the community?
As I understand it, Wine was originally intended to be both (1) a set of libraries that Windows developers could recompile their code against to run on other operating systems; and (2) a compatibility layer to run unmodified Windows binaries on other operating systems. Which one was the "primary" intent of Wine originally?
Also, nowadays, it seems that the vast majority of people use Wine in mode (2). Few developers have used the Wine libraries to recompile their code. Is this a fair assessment? If so, how does this affect the way you develop the Wine codebase? Do you see this changing in the future?
Microsoft doesn't have any grounds for suing. The codebase was written from scratch (so no copyright issues), and if Wine infringes on Microsoft patents, then so does OS X, Linux, BSD, etc. Would you say that all of those are wasted projects, since they are going to be sued and shutdown?
Also, even if Wine suddenly disappeared tomorrow, it still would not have been a waste. It has taken 15 years for Wine to get to where it is now, but it was being actively used during those 15 years. Tens of thousands of people have been successfully using Wine to get their work done for over a decade. That's a success right there. Moreover, the developers no doubt have found Wine very useful over the years... hence why they continued working on it.
If Wine is a "waste", then so is every long-term software project.
I would argue that Wine is much more convenient than virtualization... when it works, that is.
When you run an app in Wine, it integrates (more or less) with your current desktop environment. It immediately has access to the same folder hierarchy. It also performs better (loading the wine libraries seems to have a lower overhead than loading a VM and an OS).*
The only downside to Wine is that not every app runs, and some apps run but are a bit buggy.
So I would say that Wine wins for convenience, whereas virtualization wins on "robustness": any app that runs on Windows will run on Windows in a VM. This is why I use both Wine and virtualization on my system: for most apps, I can just use Wine and it's treated like just another application. For those that don't work well in Wine, I can always open up the VM.
([*] Another aspect of performance to consider is things like hardware acceleration. Most VMs don't take advantage of 3D acceleration, whereas Wine in principle can.)
The article does not make a compelling point. It keeps saying that we can give up on models (and science), because now we just have lots of data, and "correlation is enough." What utter BS. Establishing a correlation is not enough. Even if it is predictive for the given trend, it doesn't allow us to generalize to new domains the way a well-established scientific model does. If an engineer is designing a totally new device, that goes above and beyond what any established device has done, what data can he draw upon? If there is no mountain of data, he must rely on the tried-and-true techniques of engineering/science: use our best models, and predict how the new device/system will behave.
The article actually makes this point perfectly clear when it says: Indeed. Merely having tons of data doesn't actually give you insight into what you have measured. You must distill the data, pull out trends, and construct models. I just don't see how have mountains of data about a species, but still being unable to answer simple questions about it, is superior to conventional science (which can answer questions about the things it has discovered).
A deluge of data and data-mining techniques is a boon to science. But I don't see the benefit of giving up on the remarkably successful strategy of constructing models to explain the phenomena we've observed. I somehow doubt that having 20 petabytes of data on electron-electron interactions is more useful than having a concise theory of quantum mechanics.
The article is full of meaningless or incorrect statements. Like: As a researcher in the field, I can assure you that this isn't a controversial statement. We all agree that glasses are not at equilibrium. We all agree that the low-energy state for glasses is to crystallize, and that (in principle), if you wait long enough they will crystallize. The questions revolve around details like "how far from equilibrium?", "what are the implications of being non-equilibrium (e.g. on phase transitions)?", "what are the kinetics and dynamics?", "how long would it ~actually~ take for a given amount of change/flow/reconstruction/etc.?"...
Also, equating "equilibrium" with "being a solid" is total nonsense. (Solids, liquids, and gases can all be at equilibrium or far from equilibrium...)
In short, don't waste your time with this ridiculously hyped review of some otherwise interesting (but not revolutionary) science.
Yes, in theory. Just as the sun will rise tomorrow "in theory." And if I repeatedly shoot someone in the head, they will die, "in theory." And reality exists, "in theory."
Provability only exists in mathematics. For everything else, from decisions about what to buy at the supermarket, to designs of scientific experiments, we humans must use mental models that rely upon fundamental assumptions about how the universe operates (e.g. that past experiences allow us to make meaningful predictions). In other words, every action we take must be informed by some sort of "theory." The question then becomes "how robust is this model/theory?", "how much can I trust the predictions?", "what is the range of the possible outcomes?", "what are the consequences of errors in the assumptions/model/theory?", and so on.
If you have a specific problem with one of the assumptions, logic, modeling, mathematics, data acquisition, or analysis, then point it out in detail. But saying, "that's just a theory" is not useful. Everything we do is based on theories.
After all, the opposite is also a theory: Not turning on the LHC won't cause the destruction of the Earth... in theory.
...none of their clips are availableI'm not saying the Hulu viewing experience is fantastic... just making it clear that they are have lots of full-length content on the site.
I agree with you, by the way, that Hulu is silly to ignore the international market. This may result from their various licensing deals (in which case the content owners should get a clue).
For those of you who prefer video, here's Jan Chipchase's TED talk, which covers similar topics.
Moreover, this is a stupid way to design it, since it's trivial for Wikipedia to detect what you're doing, and serve a custom error page, as they have done. In short, why did these people assume Wikipedia was going to let them continue infringing their trademark and taxing their servers?
Except that #2 (and therefore #3) is rather unlikely.
If this new site doesn't provide anything above-and-beyond what Wikipedia provides, then few people will link to it, and its PageRank will be low. Without ranking high on Google, no one will find the site, and their ad revenue will be pathetic.
So, I don't really understand their business model here. Unless they offer some "value added" over the normal Wikipedia (quicker load times, vetted articles, better search, etc.), then they can't hope to attract eyeballs to their adds.
Forking is fine. A crappy fork, however, won't attract interest, and won't last long.
I've always wanted a realtime graphics engine based on something like the POV-ray ray-tracer (or other procedural modeling). The POV-ray syntax is all "exact". Rather than approximating shapes using subdivision into triangles, exact shapes are created by specifying things like "spheres" or "cylinder" or unions, intersections, and differences thereof. More complex objects can be specified by arbitrary mathematical equations, and complex sequences of operations (e.g. take a spline, sweep it along a path, intersect it with another shape, apply a certain matrix transform, ...). Having done some modeling both ways, I much prefer the "exactness" of procedural definitions, rather than approximation. (I inevitably wish I could go back and add resolution to a triangulation, but that isn't easy to do properly.)
The neat thing is that the resulting objects (if properly defined) have "infinite" detail. The roughness on a surface, for instance, can be based on a noise function, so you can zoom into it without ever seeing triangulation or other artifacts.
The obvious downside is that the computation here is intensive. Objects can be arbitrarily complicated. Calculating the intersection of a ray with a mathematically-defined surface involves very complex calculations. Rendering POV-ray scenes on modern hardware, for instance, can take minutes to days (depending on complexity).
One upside is that the rendering can be tuned to available resources. On older hardware, the number of light-sources (or the intersection accuracy, etc.) can be reduced. This would mean that video game graphics would get arbitrarily "better and better" on newer hardware, without any need for someone to change the code. Having said all this... I think our hardware is not yet powerful enough to make this kind of thing practical. (There are some neat examples that have been coded, but as a general technique we're not there yet.)
Probably most of us agree with that statement in principle. The problem is that the various players in this (users, content providers, and network operators) do not have their objectives aligned. Thus, the engineers for the network operator will come up with a solution (e.g. throttling) that solves the network company's problem (users using too much of the bandwidth they (over)sold), but the engineers working for the users (e.g. people writing P2P apps) will engineer for a different objective (maximum transfer rates), and will even engineer workarounds to the 'solutions' being implemented by the network.
The problem is thus that everyone is engineering in a fundamentally adversarial way, and this will continue so long as the objectives of all parties are not aligned. Ideally, legislation would help enforce this alignment: for instance, by legally mandating an objective (e.g. requiring ISPs to be transparent in their throttling and associated advertising), or funding an objective (e.g. "high-speed access for everyone"), or by just making illegal one of the adversarial actions (e.g. source-specific throttling).
This is not purely an engineering question. The networks have control of one of the limited resources in this game (the network of cables already underground; and the rights required to lay/replace cables), and this imbalance in power may require laws to prevent abuse. It's not easy to create (or enforce) the laws... and ideally the laws would be informed by the expertise of engineers (and afford ways for smarter future solutions to be implemented)... but suggesting that we should just let everyone 'engineer' the solution misses the mark. Whose engineers? Optimizing for what goal? Working under what incentives?
Put more simply: engineering is always bound by laws.
Another way to look at it is that this is a (somewhat whimsical) way to test the limits of hiding malicious code in open-source code. This contest, in a sense, is part of the transparency and security of the open-source method. Everyone knows that you can quite easily hide malicious code in a closed-sourced project. But this contest gives the open-source community a chance to see the limits of open-source.
The participants in the contest (including entrants, judges, and interested observers) will all learn valuable lessons about code security, and will have a better idea of how to spot malicious code in plain site, and how to structure programs to avoid malicious or erroneous code contributions.
So, far from showing that the OSS method can't help in avoiding backdoors, I view this as part of the process of avoiding backdoors. And you really better not install any closed-source image editing software, since finding malicious code in that case is a thousand times harder.
- Send unsolicited promotional copies to various parties. But unsolicited merchandise through the mail is presumptively a gift. So the people will then own the promotional copy, and have the right (as this recent ruling affirms) to resell the item. It would be fraudulent to mail things that say "you must pay for this!" or even "you must return this!" (thereby forcing people to pay postage for things they never requested).
- First send out a letter asking the recipient to sign a contract, where they agree that upon receiving the promotional item, they will return it, and agree not to resell or give it to others. This would work, but would be onerous for both parties. In particular, the vast majority of people wouldn't bother with that kind of hassle.
The promoters need to make it easy for the intended recipients to get the promo and to hear it. So I doubt they will use the second option (although I suppose a large company could try to get many outlets to sign a blanket contract that will cover all promos thereafter sent to them). In the end, they will just have to accept the fact that the promotional material can be transferred, bought, and sold, like anything else.