I don't see this as a problem at all; you shouldn't be putting so much strain on your connector that you need screws to hold it in. If you have a long, heavy cable hanging from the back of your TV, hold it in place with a cable clip so that there is no strain on the connector at all. I much prefer small, friction-based connectors to the old VGA and DVI style connectors.
The only claim the paper makes is that the data "suggests an association", and only under very specific conditions. Forget the jackknife or sample size, you seem to have trouble understanding something as simple as the meaning of the word "suggests" in scientific writing.
Wheres the bootstrap/jackknife
I hope you can figure out yourself why that is a stupid question. In any case, even if resampling methods were appropriate, many reviewers still expect the traditional parametric tests, and that's why authors use them.
Readers and reviews should have all the data needed to be able to find any mistakes.
That is true in mathematics, it is not true in experimental sciences.
I do this for a living at university. I get good reviews on my papers and I frequently consult on the statistics in other literature. A lot of Medical research is sloppy at best
Yes, you're absolutely right: most scientific papers and reviews are sloppy... because of people like you. You confuse the trappings of formality (raw data, specific statistical methods) with sound reasoning. As an editor, reviews from people like you are nearly useless to me.
Why then does Adobe have Photoshop for Macs but doesn't have it for Linux? The same can be said of many commercial programs.
Because there is a market for their products on OS X, there isn't on Linux.
Adobe was even developing for MacOS and MFC, both of which are far worse than Cocoa; they really don't give a damn how bad a platform is if there is a market.
Who said anything about "dumbing down" the language? The problem with C/C++ is not that it has pointers and other low-level features, it is that those facilities are very, very poorly designed.
It is, in fact, C that is the "dumbed down" language: originally, it had a dumbed down type system because more wasn't possible on a PDP-11 at the time. But the dumbed down type system attracted dumb users, and the rest is history.
The practice is a long history of precisely that sneaking happening
Sure. The mafia also has a long history. That doesn't mean one can't criticize them for what they do.
Apple, in particular, deserves to be criticized sharply for these kinds of patents (they file junk like this regularly) because their marketing department is creating the false impression that Apple is "innovative".
I would suspect that it would be easier to run automated programs for finding buffer over-runs, etc, rather than phishing through thousands of lines of code looking for a non-obvious vulnerability
Yes. It's called "not programming in C/C++".
Buffer overruns are a peculiarity of C/C++ (and a small minority of other, badly designed languages).
Lebedev didn't apply for a patent, Apple did. By doing so, Apple is falsely claiming that they invented the technology, and they are trying to exclude others from using it. That is both dishonest and wrong.
Perhaps Apple patented this as a countermeasure against someone who would try to claim this as an original idea.
That's not how patents work. You can only legally patent things that you actually invented. You aren't patent things that you know to be invalid simply because you think you might be able to sneak them past the examiner.
A differently-worded patent on a new product is better than no patent at all. At least that's my opinion.
Ah, the criminal mind at work: you don't care whether it's legal, you don't care whether it's ethical, you just care whether you can get away with it. Say, do you work for the mafia?
IANAL, but it seems that Art. Lebedev Studio could just negotiate a fat licensing fee for the technology/idea with Apple and both would win from the collaboration...?
Only if Lebedev Studio has a patent. If they don't, they can't force Apple to do anything. They might ask for the patent to be re-examined, but they can't get money from Apple for that.
If the patent issues, Apple, on the other hand, can sue Lebedev. Then, Lebedev can defend itself claiming prior art, but that's all.
I haven't read the whole patent on the Apple keyboard, but it seems to me that there is at least one significant difference between the Lebedev device and the Apple concept, and that is that the keyboard would change dynamically, in real time, i.e. to present contextual controls based on what you are working on.
Nonsense. Dynamic update of the layout is clearly anticipated by the Optimus keyboard, since gaming is one stated application.
Furthermore, you have to read the claims, and many of Apple's claims apply to ("read on") the Optimus keyboard as it is, even if the Optimus keyboard were fully static.
This patent just shows again how arrogant Apple is: either they don't know what's going on in the world, or they don't care.
I'm not calling Cocoa "commercial" (it's proprietary). I'm saying that if you're using OS X, don't bother with open source software for it, pay for the commercial apps.
I've never been much into cross-platform GUI programming
Neither have I, which is why I mostly use Linux: that's where the good open source software runs; open source sofware ported to OS X just isn't worth it.
Nice thing about developing for OS X is that it is free (well, you don't pay extra for it).
Only if you don't value your time: OS X programming is both harder than Linux programming and the skills you need to acquire are not useful for much else.
I think you are a good example for why this is a good requirement: you really should not be providing forensic services without being familiar without satisfying at least the requirements for a PI.
I use macports to maintain a Ruby on Rails dev environment, personally. I'd hate to have to compile and maintain all that stuff manually.
Yeah, but even RoR is pretty self-contained and simple; it hardly uses any additional shared libraries, for example, and doesn't depend much on the OS.
Try getting something like wxRuby, wxPython, TkInter, or SciPy up and running on MacOS; it's a PITA no matter what you try, it never really quite works 100%, and the next Apple upgrade can break it all.
On Linux, that stuff is packaged up consistently, and the OS libraries are always in sync with the packages.
OS X is usable for open source web frameworks, but for desktop apps, I find it better to either stick with the OS X specific commercial stuff or just use Linux.
You won't need a license to go digging through your own hard drive. What you do need a license for is to go digging through other people's hard drives. That seems like a good thing. Among other things, it probably means that Best Buy can't go on fishing expeditions through your hard drives anymore.
Fink is a PITA: it's slow, many packages are missing, the dependencies aren't quite right, and the apps often don't work quite right in a Mac environment; I've given up on it on my Mac. Ditto for Darwin ports.
Use a Mac for what Macs are good for and use Linux for what Linux is good for; anything in Fink, you're better off running on a Linux machine.
I think before any high-tech solutions, a much simpler thing to do would be better training. Many doctors don't seem to be particularly careful about what they touch.
I think doctors should be required to observe each other during training, spot potentially unsafe practices, and give each other feedback.
Furthermore, there should be random checks for the sterility of gloves and instruments.
I think it's rather the opposite: the Mac is a good stepping stone on the way to Linux: when people are tired of getting nickled and dimed and DRM'ed by Apple and add-on products, when they're tired of the limitations of the Mac, they can move up to Linux.
As an X11 developer on OS X, I'd like to know what you're talking about.
X11 on OS X falls short in areas such as window management, keyboard handling, drag-and-drop, cut-and-paste, desktop integration, lack of modern toolkits, etc. X11 on OS X could be as well integrated as Cocoa and Carbon, but Apple chooses to keep X11 something that looks and feels foreign, almost certainly in order to get developers to move to their proprietary APIs.
The Data is missing. Only summary statistics are presented.
This is the norm for scientific papers; you can get the raw data by writing to the authors or the journal.
If you know your statistics you will know why this is a problem.
I know my statistics, and this is not a problem: the authors give you sample size, mean, standard deviation, F values, and significance levels. That is sufficient.
The group size give problems for statistical power in the result.
The statistical power is clear from the data they provide, and the group size is actually fairly large for these kinds of experiments.
Furthermore, the authors are cautious in their claims, since they only claim that the data "suggests an association"; to support that statement, they didn't even have to have statistically significant results at the level they have.
Your kind of criticism is what I mean by "scientific illiteracy": you miss the technical details of the data analysis, and you don't understand what the conclusions are actually saying.
Markov chains are the original, mathematical theory. This is just an application of Markov chains to tracking disease transmission, a fairly common method that long pre-dates Google. Google's page rank algorithm is another application of Markov chains to citation ranking (and, as it turns out, it wasn't the first time that it was applied to that either).
Whitworth is confusing two issues: digital models of physics and simulations.
Digital models of physics have been explored for decades under various names. Yes, they "explain" quantum mechanics and relativity, but unfortunately not yet in any useful or interesting sense.
Saying that the universe is a "virtual reality simulation", however, is something very different: it doesn't just mean that it's a digital system, it implies that there is an intelligence running the simulation inside another physical reality. That is not testable; we might still find out somehow (like if the person running the simulator communicates with us and proves his identity by fundamentally altering physical laws), but there are no obvious experiments we can use to distinguish digital physics from digital physics within a simulation.
I think all interactions with government and all business-related conversations should be legally recordable by any participant without the need for consent or notification.
In order to make these kinds of recordings stick, it might also be a good idea to have cameras and audio devices that actually cryptographically sign the recordings. Adding that to existing cameras or recorders is not hard, and it doesn't interfere with their normal operations. It just means that you can tell whether a recording is original or has been modified in some way.
Slashdot Mirror
Then they should be developing a Linux version. Adobe had already developed versions for Unix, both SGI IRIX and Sun Solaris.
Adobe developed Photoshop for UNIX and it flopped. That's probably why they think that Photoshop for Linux is a bad idea in the first place.
Chemical storage of the energy avoids that problem. As such, there are various forms of chemical energy storage,
Yes, but using CO2 as a battery seems like one of the worst solutions.
If you want to store large amounts of electricity, pumping water, flywheels, and methods like that work quite well.
I don't see this as a problem at all; you shouldn't be putting so much strain on your connector that you need screws to hold it in. If you have a long, heavy cable hanging from the back of your TV, hold it in place with a cable clip so that there is no strain on the connector at all. I much prefer small, friction-based connectors to the old VGA and DVI style connectors.
or not making the claim in the first place.
The only claim the paper makes is that the data "suggests an association", and only under very specific conditions. Forget the jackknife or sample size, you seem to have trouble understanding something as simple as the meaning of the word "suggests" in scientific writing.
Wheres the bootstrap/jackknife
I hope you can figure out yourself why that is a stupid question. In any case, even if resampling methods were appropriate, many reviewers still expect the traditional parametric tests, and that's why authors use them.
Readers and reviews should have all the data needed to be able to find any mistakes.
That is true in mathematics, it is not true in experimental sciences.
I do this for a living at university. I get good reviews on my papers and I frequently consult on the statistics in other literature. A lot of Medical research is sloppy at best
Yes, you're absolutely right: most scientific papers and reviews are sloppy... because of people like you. You confuse the trappings of formality (raw data, specific statistical methods) with sound reasoning. As an editor, reviews from people like you are nearly useless to me.
Why then does Adobe have Photoshop for Macs but doesn't have it for Linux? The same can be said of many commercial programs.
Because there is a market for their products on OS X, there isn't on Linux.
Adobe was even developing for MacOS and MFC, both of which are far worse than Cocoa; they really don't give a damn how bad a platform is if there is a market.
Who said anything about "dumbing down" the language? The problem with C/C++ is not that it has pointers and other low-level features, it is that those facilities are very, very poorly designed.
It is, in fact, C that is the "dumbed down" language: originally, it had a dumbed down type system because more wasn't possible on a PDP-11 at the time. But the dumbed down type system attracted dumb users, and the rest is history.
The practice is a long history of precisely that sneaking happening
Sure. The mafia also has a long history. That doesn't mean one can't criticize them for what they do.
Apple, in particular, deserves to be criticized sharply for these kinds of patents (they file junk like this regularly) because their marketing department is creating the false impression that Apple is "innovative".
I would suspect that it would be easier to run automated programs for finding buffer over-runs, etc, rather than phishing through thousands of lines of code looking for a non-obvious vulnerability
Yes. It's called "not programming in C/C++".
Buffer overruns are a peculiarity of C/C++ (and a small minority of other, badly designed languages).
Lebedev didn't apply for a patent, Apple did. By doing so, Apple is falsely claiming that they invented the technology, and they are trying to exclude others from using it. That is both dishonest and wrong.
Perhaps Apple patented this as a countermeasure against someone who would try to claim this as an original idea.
That's not how patents work. You can only legally patent things that you actually invented. You aren't patent things that you know to be invalid simply because you think you might be able to sneak them past the examiner.
A differently-worded patent on a new product is better than no patent at all. At least that's my opinion.
Ah, the criminal mind at work: you don't care whether it's legal, you don't care whether it's ethical, you just care whether you can get away with it. Say, do you work for the mafia?
IANAL, but it seems that Art. Lebedev Studio could just negotiate a fat licensing fee for the technology/idea with Apple and both would win from the collaboration...?
Only if Lebedev Studio has a patent. If they don't, they can't force Apple to do anything. They might ask for the patent to be re-examined, but they can't get money from Apple for that.
If the patent issues, Apple, on the other hand, can sue Lebedev. Then, Lebedev can defend itself claiming prior art, but that's all.
I haven't read the whole patent on the Apple keyboard, but it seems to me that there is at least one significant difference between the Lebedev device and the Apple concept, and that is that the keyboard would change dynamically, in real time, i.e. to present contextual controls based on what you are working on.
Nonsense. Dynamic update of the layout is clearly anticipated by the Optimus keyboard, since gaming is one stated application.
Furthermore, you have to read the claims, and many of Apple's claims apply to ("read on") the Optimus keyboard as it is, even if the Optimus keyboard were fully static.
This patent just shows again how arrogant Apple is: either they don't know what's going on in the world, or they don't care.
Maybe there's something unique and non-obvious about their method of implementing the "dynamic keyboard" idea.
Why don't you just stop guessing and read the fscking patent? It's on the USPTO web site.
I did. Apple didn't try to patent anything new, they really are just patenting the Optimus keyboard exactly.
That is wrong. It may, in fact, be fraudulent because... how out of it can Apple engineers be?
Why call Cocoa "commercial?"
I'm not calling Cocoa "commercial" (it's proprietary). I'm saying that if you're using OS X, don't bother with open source software for it, pay for the commercial apps.
I've never been much into cross-platform GUI programming
Neither have I, which is why I mostly use Linux: that's where the good open source software runs; open source sofware ported to OS X just isn't worth it.
Nice thing about developing for OS X is that it is free (well, you don't pay extra for it).
Only if you don't value your time: OS X programming is both harder than Linux programming and the skills you need to acquire are not useful for much else.
I think you are a good example for why this is a good requirement: you really should not be providing forensic services without being familiar without satisfying at least the requirements for a PI.
Really?
Really.
http://www.google.com/search?q=pagerank+markov-chain
If so, the Pagerank patent would have been challenged.
Why would anybody bother? Google hasn't sued anybody, and pure page rank doesn't work all that well anyway.
I use macports to maintain a Ruby on Rails dev environment, personally. I'd hate to have to compile and maintain all that stuff manually.
Yeah, but even RoR is pretty self-contained and simple; it hardly uses any additional shared libraries, for example, and doesn't depend much on the OS.
Try getting something like wxRuby, wxPython, TkInter, or SciPy up and running on MacOS; it's a PITA no matter what you try, it never really quite works 100%, and the next Apple upgrade can break it all.
On Linux, that stuff is packaged up consistently, and the OS libraries are always in sync with the packages.
OS X is usable for open source web frameworks, but for desktop apps, I find it better to either stick with the OS X specific commercial stuff or just use Linux.
You won't need a license to go digging through your own hard drive. What you do need a license for is to go digging through other people's hard drives. That seems like a good thing. Among other things, it probably means that Best Buy can't go on fishing expeditions through your hard drives anymore.
Fink is a PITA: it's slow, many packages are missing, the dependencies aren't quite right, and the apps often don't work quite right in a Mac environment; I've given up on it on my Mac. Ditto for Darwin ports.
Use a Mac for what Macs are good for and use Linux for what Linux is good for; anything in Fink, you're better off running on a Linux machine.
I think before any high-tech solutions, a much simpler thing to do would be better training. Many doctors don't seem to be particularly careful about what they touch.
I think doctors should be required to observe each other during training, spot potentially unsafe practices, and give each other feedback.
Furthermore, there should be random checks for the sterility of gloves and instruments.
I think it's rather the opposite: the Mac is a good stepping stone on the way to Linux: when people are tired of getting nickled and dimed and DRM'ed by Apple and add-on products, when they're tired of the limitations of the Mac, they can move up to Linux.
As an X11 developer on OS X, I'd like to know what you're talking about.
X11 on OS X falls short in areas such as window management, keyboard handling, drag-and-drop, cut-and-paste, desktop integration, lack of modern toolkits, etc. X11 on OS X could be as well integrated as Cocoa and Carbon, but Apple chooses to keep X11 something that looks and feels foreign, almost certainly in order to get developers to move to their proprietary APIs.
The Data is missing. Only summary statistics are presented.
This is the norm for scientific papers; you can get the raw data by writing to the authors or the journal.
If you know your statistics you will know why this is a problem.
I know my statistics, and this is not a problem: the authors give you sample size, mean, standard deviation, F values, and significance levels. That is sufficient.
The group size give problems for statistical power in the result.
The statistical power is clear from the data they provide, and the group size is actually fairly large for these kinds of experiments.
Furthermore, the authors are cautious in their claims, since they only claim that the data "suggests an association"; to support that statement, they didn't even have to have statistically significant results at the level they have.
Your kind of criticism is what I mean by "scientific illiteracy": you miss the technical details of the data analysis, and you don't understand what the conclusions are actually saying.
Markov chains are the original, mathematical theory. This is just an application of Markov chains to tracking disease transmission, a fairly common method that long pre-dates Google. Google's page rank algorithm is another application of Markov chains to citation ranking (and, as it turns out, it wasn't the first time that it was applied to that either).
Whitworth is confusing two issues: digital models of physics and simulations.
Digital models of physics have been explored for decades under various names. Yes, they "explain" quantum mechanics and relativity, but unfortunately not yet in any useful or interesting sense.
Saying that the universe is a "virtual reality simulation", however, is something very different: it doesn't just mean that it's a digital system, it implies that there is an intelligence running the simulation inside another physical reality. That is not testable; we might still find out somehow (like if the person running the simulator communicates with us and proves his identity by fundamentally altering physical laws), but there are no obvious experiments we can use to distinguish digital physics from digital physics within a simulation.
I think all interactions with government and all business-related conversations should be legally recordable by any participant without the need for consent or notification.
In order to make these kinds of recordings stick, it might also be a good idea to have cameras and audio devices that actually cryptographically sign the recordings. Adding that to existing cameras or recorders is not hard, and it doesn't interfere with their normal operations. It just means that you can tell whether a recording is original or has been modified in some way.