It's not the guilt by association that makes them unhireable, it's the danger of finding yourself being sued by SCO that makes you not want to hire them.
Considering the ease at which SCO starts lawsuits on their alleged IP, it is not unimagenable that anyone that is tainted with knowledge of SCO's code will prompt a SCO lawsuit once he contributes to your codebase.
This is indeed very important, thank you very much for this link. This completely shatters my belief that piracy equalling copyright infringement is a new use of the word.
In other words, if someone with moderator points scans this post, please mod it up, as this is an interesting tidbit of info. I for one will undoubtedly steal this point (and link) for quick karma-whoring when the topic resurfaces again.
Hey, I can get 'unfit' software for free. The scheme works as long as I contribute 'unfit' software myself, which I do, and which is great fun. I don't think this will be a good idea in construction though.
Funny this. Maybe we should stop fixing up drivers who got hurt in a car accident. It was their own choice to drive, right? Pedestrians deliberately take a risk when crossing the street, no need to help them when it goes wrong, right? Accidents in the house. You took the choice to move around. I shouldn't be paying if you're a blithering idiot that falls down the stairs. Ok, more apt example: if you're a regular at McDonalds, you'll not be treated for cardio-vascular problems.
Face it, life is lethal. If you're a smoker and die of lung-cancer, that saves you from dying of any other illment. Almost any death in our society costs money,
in fact, old age seems to be the most expensive bad habit of all. Why should I, as a smoker, pay for the pensions, medicines and operations of all these old geezers that refuse to die before their 90's. I'm going to be gone way before that. I want a refund!
Personally I don't mind changing the customization settings. Once. But with freaking windows I have to do it time and time again. When changing computers, when I'm at a different location, when upgrading office, all the time. I have to do it in Word, in Excel, in Powerpoint, in Windows itself, in whatever fucking application that has inane defaults. As I can't bring all settings easily with me such as I can do under linux (all files in the.dirs), I just have to live with the defaults as they are as I don't have the time to customize it all the time.
Flip a fair coin. It will come up heads or tails in an unpredictable way. Flip the coin a million times, a billion times, a trillion times and count the average number of times it came up heads. The more often you throw, the closer this value will go to 0.5 (even though it does vary). Given that the coin is fair (and not wearing out after trillion throws:), and the throwing method is 'chaotic', every single coinflip will be unpredictable, yet the end result is very well known and becomes ever more accurate with more throws.
Maybe they did exactly this? The exploit was the ptrace exploit, a local exploit. Maybe an inside job, maybe not. This could however simply mean that it was this limited connected server that was compromised. Maybe all machines inside were compromised, and the ftp server was just one of them. Once such a crack appears inside the citadel, nothing can be trusted anymore.
As some other posters in other threads noticed, the FSF does not have full backups because all backups made after early 2003 can be compromised. The crack happened in March, and what they miss is all the stuff that was uploaded after the crack. Backups from before March are available. In this situation no backup strategy at all would leave you with total security after March. The fact that the site was cracked five months ago is a bit scary though.
Funny you mention Microsoft as an example of non-gamblers. They are gamblers, highly successful ones, but gamblers none the less. Investing in Microsoft is risky: they have a monopoly on desktop machines, but their entire corporation is built upon the increase in value of the MS stock. Once the stock doesn't rise anymore for an extended period of time, only then we will see how much of a real company Microsoft is.
IBM, yes, that's a different story. Any company that has survived for a century or more can be expected not to fall over next decade or so. But Microsoft? Are you willing to put money in the Microsoft stock and not look at it for 10 years?
I think that the reason for the GPL not to be challenged in court is that it's a lose-lose proposition for the challenger:
If they lose the case, they should stop distributing GPL software without source code
If they win the case, the GPL is invalid, normal copyright still applies, and they have to get the author's permission (probably the one they are suing), to distribute the binaries.
Interesting thought about membrames, though I doubt this supposed non-binaryness is the key. I do think however that at some point we will view a workstation as our computer analogue of a cell, which needs its protection through some sort of membrame (call it a firewall, a virus filter or whatever), even though it needs to communicate with other cells. Biology definitely has something to learn CS for defense against hostile actions, that's for sure.
Interesting that you mention Turing as an example on non-biological computing. In my opinion, Church/Turing and Godel, while laying the foundations of computing, actually pointed to a deep link between CS and biology. Their results depend completely on a fundamental aspect of both biology and computing in that code and data are the different sides of the same medal. You can read code and you can execute data, just like DNA, RNA and proteins. They're read, executed and produce new data (or code, depending on the way it is read/executed).
Current day computing with its emphasis on encapsulation of data, divorced from algorithms, structured or OO based programming, are a far cry from the early beginnings of computer science (which predates Watson&Crick). So I would say that rediscovering Turing would bring CS back to biology (and have you ever checked the biological phenomenon of Turing patterns? Relevant to leopard skins and such).
Uhm, no. Quantum computing works with qubits, which are just like normal bits only their oneness can be described by a probability amplitude (instead of > x Volts). What such an amplitude actually means has been beyond anyone's grasp. So they're not three states: 0 or 1 or something inbetween, but all of these together and then some.
Presuming you're not a creationist, there are MILLIONS of generations worth of Darwinism at work in even a simple worm - weeding out the inefficient in times of stress, etc.
Even if you're a creationist, there still will be MILLIONS of generations of natural selection at work. Come to think of it, millions is a slight understatement. Some friends and I actually tried to calculate the number of generations that make up a human two weeks back. Given that most of evolution has been in the single cell stage, us mammals are very slow and therefore insignificant in counting generations. We used an average reproductive cycle of 1 hour (E-coli is known to do it in twenty minutes), and eventually (can't remember the exact calculation, I think we used 5 billion years ago as a starting point for life) ended up with 8 trillion generations to create me (and possibly you as well). That's 10^12 generations to create a person.
Not all of those generations had a strong selection pressure, but still.
Interesting, but understanding other people's assembler has nothing whatsoever to do with understanding DNA. I'm not flaming you here, but suppose that I would confront you with an assembler program where something that at one point was data was suddenly executed, but only after some repair algorithm in some other part of the program made some seemingly nonsensical changes to it?
You can understand assembler because you have a pretty good model of how you other people are taught to write assembler programs whose structure they can keep in their head. This means for instance making a rigorous distinction between algorithm and data. In DNA however, such a distinction is simply not there: any part of the code might function as both.
As there's no intelligent designer working behind the scenes of the DNA-program, it would be one hell of a job to read such a program, as anything that is physically possible can be going on around here. Contrast this with human-written computer code, and you'll notice (as you noticed) that there are plenty of constraints at work that make it easier to understand what is going on.
The computer program equivalent of a DNA-program is a piece of code, when executed in the right environment (let's call it an operating system), will self-unpack and iteratively expand itself to fill the computer (cell) and many computers around it, growing seemingly tangential functions in the process. Imagine a piece of assembler that will copy itself to a fresh machine, start communicating with itself, will create some new assembler (encoded in itself), executes it, repairs it, executes it again, and grow a bit more, until it is ready to create offspring. All this iterative expansion is somehow located on that single initial string, but in all but isolated circumstances (genes), simply poking around and seeing what will happen will crash the entire developmental process. And no, there's no explicit graphical routine in the initial code, it is grown from it.
Last time I checked, the openoffice format was a zipfile containing the document (and other things) in XML-format. The binary part is thus pkzip, which is pretty much available for all time, the open office part is XML. How well that format is documented is another issue.
Another thing is graphics, and I don't know if Python has an interface to the Windows graphics APIs
Check out wxPython, these are wxWindows bindings for python, work very naturally for Windows API literates and binds to several widget sets: windows API, gtk, and quite likely more.
Point taken. Maybe the best way to approach this is to start with a FUD campain that proprietary formats kill people (general practioners data can't be read by hospitals? people can die from this!). Once the point got across that open standards are the only way to save lives, the next step would be to argue that only an open implementation of an open standard can be trusted (.doc format is made open by microsoft? How do we know it can actually work according to these specs? show the implementation!).
3: Open Source Wins.
It seems like a strategy, even the step 2 is filled in,
but who am I kidding? I'm too involved with other things to fight this battle.
No, the preferred order would indeed be to enforce well-documented, open formats for any and all governmental data. This is an easy battle to win because you can scare any government official with the fact that they are currently (using MS) very likely to lose data if they upgrade a few times. This is very strong FUD that we can play out. Once this battle is won, MS needs to support some open format or another, but at the same time, open source software can implement the exact same spec (it's open, right). At that point it will become a battle on merit and the free beer of open source will be a big advantage.
If you do it the other way around (as seems to be advocated most strongly in this story and discussion), the issues of free beer get mixed with free speech (which in this case means readible data), and you'll be fighting both battles at the same time. Bad tactics that is.
That's true, but India does not have a large collection of small software companies that program exclusively for windows for their internal market. In western Europe for instance, many small shops exist that have their own local customer base (including government), and create software based on Windows. Their worth (in terms of tax and employment) is larger than the OSS market.
In any case, I didn't say that I endorse this argument about closed source being economically more important, what I did note was that the creation of open source as an economic activity is controversial (and you only have to read slashdot to figure that one out). If we get rid of the barriers to entry by having the government endorse open standards throughout, this will benefit open source tremendously (as almost by definition, an open source product uses an open standard for storage and communication).
Slashdot Mirror
Considering the ease at which SCO starts lawsuits on their alleged IP, it is not unimagenable that anyone that is tainted with knowledge of SCO's code will prompt a SCO lawsuit once he contributes to your codebase.
In other words, if someone with moderator points scans this post, please mod it up, as this is an interesting tidbit of info. I for one will undoubtedly steal this point (and link) for quick karma-whoring when the topic resurfaces again.
Hey, I can get 'unfit' software for free. The scheme works as long as I contribute 'unfit' software myself, which I do, and which is great fun. I don't think this will be a good idea in construction though.
Face it, life is lethal. If you're a smoker and die of lung-cancer, that saves you from dying of any other illment. Almost any death in our society costs money, in fact, old age seems to be the most expensive bad habit of all. Why should I, as a smoker, pay for the pensions, medicines and operations of all these old geezers that refuse to die before their 90's. I'm going to be gone way before that. I want a refund!
The point? If it's not fit for ANY purpose, why on earth should I pay for it?
Oh, and I'll see your two judges, and raise you eight ballot counters.
Personally I don't mind changing the customization settings. Once. But with freaking windows I have to do it time and time again. When changing computers, when I'm at a different location, when upgrading office, all the time. I have to do it in Word, in Excel, in Powerpoint, in Windows itself, in whatever fucking application that has inane defaults. As I can't bring all settings easily with me such as I can do under linux (all files in the .dirs), I just have to live with the defaults as they are as I don't have the time to customize it all the time.
Flip a fair coin. It will come up heads or tails in an unpredictable way. Flip the coin a million times, a billion times, a trillion times and count the average number of times it came up heads. The more often you throw, the closer this value will go to 0.5 (even though it does vary). Given that the coin is fair (and not wearing out after trillion throws :), and the throwing method is 'chaotic', every single coinflip will be unpredictable, yet the end result is very well known and becomes ever more accurate with more throws.
Maybe they did exactly this? The exploit was the ptrace exploit, a local exploit. Maybe an inside job, maybe not. This could however simply mean that it was this limited connected server that was compromised. Maybe all machines inside were compromised, and the ftp server was just one of them. Once such a crack appears inside the citadel, nothing can be trusted anymore.
As some other posters in other threads noticed, the FSF does not have full backups because all backups made after early 2003 can be compromised. The crack happened in March, and what they miss is all the stuff that was uploaded after the crack. Backups from before March are available. In this situation no backup strategy at all would leave you with total security after March. The fact that the site was cracked five months ago is a bit scary though.
IBM, yes, that's a different story. Any company that has survived for a century or more can be expected not to fall over next decade or so. But Microsoft? Are you willing to put money in the Microsoft stock and not look at it for 10 years?
I think that the reason for the GPL not to be challenged in court is that it's a lose-lose proposition for the challenger:
If they lose the case, they should stop distributing GPL software without source code
If they win the case, the GPL is invalid, normal copyright still applies, and they have to get the author's permission (probably the one they are suing), to distribute the binaries.
Interesting thought about membrames, though I doubt this supposed non-binaryness is the key. I do think however that at some point we will view a workstation as our computer analogue of a cell, which needs its protection through some sort of membrame (call it a firewall, a virus filter or whatever), even though it needs to communicate with other cells. Biology definitely has something to learn CS for defense against hostile actions, that's for sure.
How old is your computer? and the one before that?
-Virus vulnerability (no pun intended)
ok, you got me there. There's no issue of viruses with present day computers.
-Nutrition requirements
Let me dump you and your computer somewhere in a wilderness, which machine will get nutrients faster?
Current day computing with its emphasis on encapsulation of data, divorced from algorithms, structured or OO based programming, are a far cry from the early beginnings of computer science (which predates Watson&Crick). So I would say that rediscovering Turing would bring CS back to biology (and have you ever checked the biological phenomenon of Turing patterns? Relevant to leopard skins and such).
By its very definition, computer pr0n is a merger between biology and computers.
Even if you're a creationist, there still will be MILLIONS of generations of natural selection at work. Come to think of it, millions is a slight understatement. Some friends and I actually tried to calculate the number of generations that make up a human two weeks back. Given that most of evolution has been in the single cell stage, us mammals are very slow and therefore insignificant in counting generations. We used an average reproductive cycle of 1 hour (E-coli is known to do it in twenty minutes), and eventually (can't remember the exact calculation, I think we used 5 billion years ago as a starting point for life) ended up with 8 trillion generations to create me (and possibly you as well). That's 10^12 generations to create a person. Not all of those generations had a strong selection pressure, but still.
The computer program equivalent of a DNA-program is a piece of code, when executed in the right environment (let's call it an operating system), will self-unpack and iteratively expand itself to fill the computer (cell) and many computers around it, growing seemingly tangential functions in the process. Imagine a piece of assembler that will copy itself to a fresh machine, start communicating with itself, will create some new assembler (encoded in itself), executes it, repairs it, executes it again, and grow a bit more, until it is ready to create offspring. All this iterative expansion is somehow located on that single initial string, but in all but isolated circumstances (genes), simply poking around and seeing what will happen will crash the entire developmental process. And no, there's no explicit graphical routine in the initial code, it is grown from it.
Oh yes, you're probably right about the junk DNA.
Last time I checked, the openoffice format was a zipfile containing the document (and other things) in XML-format. The binary part is thus pkzip, which is pretty much available for all time, the open office part is XML. How well that format is documented is another issue.
Check out wxPython, these are wxWindows bindings for python, work very naturally for Windows API literates and binds to several widget sets: windows API, gtk, and quite likely more.
3: Open Source Wins.
It seems like a strategy, even the step 2 is filled in, but who am I kidding? I'm too involved with other things to fight this battle.
If the vendor is compromised, so is the mirror, and so is the md5sum. For proper md5sum checking, you would need a web of trust.
If you do it the other way around (as seems to be advocated most strongly in this story and discussion), the issues of free beer get mixed with free speech (which in this case means readible data), and you'll be fighting both battles at the same time. Bad tactics that is.
I tried to argue this for this same article here.
In any case, I didn't say that I endorse this argument about closed source being economically more important, what I did note was that the creation of open source as an economic activity is controversial (and you only have to read slashdot to figure that one out). If we get rid of the barriers to entry by having the government endorse open standards throughout, this will benefit open source tremendously (as almost by definition, an open source product uses an open standard for storage and communication).