It doesn't really "poison" the filters, because there are just wayyyyyy too many posible words for this to work. Bayesian filters assign a huge probability of spamminess to every word in a spam email and an exceedingly low prbability of spamminess to every word in a non-spam email during training. If a word appears in both, it just averages out. Over time a given word will appear only once in a spam email with a bunch of random words, and many times in non-spam emails, and therefore after some time (or even pre-emptively) the good words will be recognized as good. The more training, the better; poisoning has little chance of success as long as there's at the number of good and bad emails going in are within an order of magnitude of each other.
That's not to say the technique doesn't help the spammers in the short run; it probably gets past less sophisticated and trained filters.
The interesting thing about his comments about desktop Linux are that he's making them at all. He used to have a position of "Linus is what it is, I don't care where it goes, it's just fun to watch." He's not doing that so much now that it appears to be actually getting the places people imagined it would go 5-10 years ago. To make a specific claim, even one as flexible as that, is out of character for him and shows that he's starting to become interested in seeing his work succeed commercially (other than in the areas he works on directly).
If they have the common good in their interests, they really should patent this process, then place the patent in the commons. (I'm not sure how this process works with patents, but I'm confident it can be done somehow.) To do so would be to protect future generations from patent mischief with this application.
An organization that purports to be able to tell us whether they're useful or not. I'll grant you that "sweat like a pig" certainly implies something inaccurate, but a lot of people like me actually know this about pigs and use the term anyway because it's an interesting turn of phrase. (How hard do you have to be working to be a pig and be sweaty?) Let's look at some others:
Metrosexual? Bling-Bling? Well, I would never use them in a sentence, but if some people find one of them useful, then let them use it. Like many entries in the list, they are words that annoy the committtee them because they're new, not because they're useless or overused. If it really is useless, it will soon cease to be used. That's how language works.
Punked? Actually a useful word, I think. There's no word for "play a elaborate prank on", yet. I think this one annoys them only because it comes from pop culture.
Companion animals: Their smarmy comment "They're called PETS." is probably downright offensive to people who object to the negative connotation of "pet". I call my dogs pets, but I love them, and I can certainly see them as companions.
Smoking gun: "came to us from Iraq". Uh, no it didn't. This term has probably been around for at least decades. Overused? Maybe this year, but that's certainly not cause for banning it evermore.
They're trying to be funny, but they're just demonstrating a failure to understand language evolution.
I just realized that the situation with the entertainers/manual laborers is a bit more complicated than what I described. It revolves around this question: Am I purchasing all the items they need to do their jobs?
This includes 150 costumes (except possibly for the ladies dancing; we'll assume she's at least wearing a g-string), plus 36 milking stools, 36 buckets and enough bottles to hold however much milk is produced. (The song doesn't say what you're going to do with the milk produced, but I'd definitely want to keep it. It's the one thing you're getting for free out of all this.) There's bound to be some sheet music and music stands too; it'd be pretty terrible to have all those pipers playing improv. What about permits to have all these people perform for almost a week? Cleaning costs....
"On the first day of christmas my true love gave to me: X" "On the second day of christmas my true love gave to me: Y"
etc. and each time X or Y includes all the previous items. The only obvious conclusion is that you're getting 12 partridges in 12 pear trees, 22 turtledoves, etc.
The total of the items includes combined items though, which are necessarily counted as separate. For example, you might be able to find a store that sells partridges, but probably not one that puts a pear tree in the same box.
So we've got 12 pear trees, the 142 birds, 40 rings, 40 cows, 76 women, 30 lords (is that more expensive than "30 guys jumping around in fancy dress"?), 22 pipers, 22 pipes, 12 drummers, 12 drums and 12 sets of drumsticks (they could be bongos, but traditionally they are not depicted as such). You might argue that the various musicians bring their instruments and then take them home with them when the gig is done. One assumes this is not slavery, so you have to figure salary for however many days the individuals in question are employed (and figure for the sake of simplicity that they are all let go at the end of the 12th day).
So, depending on whether or not you get to keep the musical instruments after christmas, you have purchased either 234 or 280 items and paid salaries for 150 man-days of entertainment.
"Dear Mr. BS: . . . A SYN-flood attack probably consumes 1 Kbps or less. Everybody else in the known universe can communicate with all of your externally-visible machines except www.sco.com. If the (alleged) attack on www.sco.com has affected any other machines, your network is very poorly administered. I suggest you avail yourself of the vast array of of volunteer expertise that is ready to help any user of a Linux system.
"Even you."
I'm sure they're just lining up. It's the opportunity of a lifetime.. to help SCO secure their internal systems so nobody else can log in and wreak havoc through, say, a backdoor placed there by the idiots currently running the SCO network on the advice of their helpful friends.
There's yet another type of forking that can occur which I call "forking up". This is where a small but important module is developed by someone who needs it, and the module later gets adopted into a project with broader scope and larger visibility. This is in fact the norm when projects submit code for inclusion into the standard distribution of Python or Perl. If Python adopts the module, sometimes independent development on it dies, but many times it does not--pybsddb and pyxml are two examples that spring to mind. Other projects do this as well... Twisted and Zope both get a lot of contributions from smaller projects that want to be part of a larger effort.
This is yet another case that supports parent's argument that resources are not being divided.. resources that work on fixing bugs in the higher-visibility project would never have worked on the smaller project's codebase at all. The smaller project can still benefit from porting bugfixes in stdlib over to their tree.
Companies typically defend themselves against this kind of attack by having patents on lots of stuff. Surely there is lots of code in Linux that is eligible for a patent, and can be assigned to some holding company. Such an entity could then offer to do patent trades whenever MS or some other big software bully decides to swing the patent club.
(BTW, don't bother to tell me that this patent only covers embedded devices formatted with the fs, because Microsoft could change its mind.)
Good one. Knowing how to program isn't a skill anyone can just acquire by being exposed to computers. After all, implementation hiding is one of the principles of interface design. So people exposed to the interfaces of software are being shielded from how it works internally, and they aren't going to absorb that knowledge. And even if they were exposed to the guts all day long, programming has a conceptual foundation in mathematics and needs a great deal of patience and practice--practice developing your memory to hold long logic chains mentally until they come to fruition in the software, practice knowing what patterns work and what don't, practice building disciplined habits.
That said, I've always argued that anyone can become a computer programmer. But the skills are not the sort of thing you pick up just by using a computer all your life. You have to seek out training beyond that.
I do agree with his point about the service model of software development. It's just not going to be anywhere near as hard as he claims it will be to get a job that way.
Personally I lean toward the "SCO Upper Management is a bunch of fucking loonies" theory. Here's an article that puts concrete numbers on the effect SCO is having on linux adoption.
I am a software developer. "Use" to me, means "distribute as part of a software package." Not "execute the runtime." As I already stated, if I distribute a relocatable library as part of a software package, I am bound by the terms of the GPL. This seriously hurts the software's reuse.
Some developers like the license. Some of us find out just how troublesome it can be, and change our minds.
My earlier projects were distributed under the GPL or LGPL. Then I got a job. I have to develop software for an organization that doesn't have the ability to release source code; we are a subsidiary of a larger organization that technically owns the copyright of everything I produce at work.
If I want to create software that uses GPL'd libraries, even if the software I create is a completely separate source tree and I don't change a single line of code in the distributable source of the library, the software I create has to be GPL. This is simply unacceptable; I don't have the time to get permission to do that on every project from my boss, let alone from the head of the parent organization and their lawyers. Eventually, every one of my creations at work may become OSS (and indeed, one of them already has) but I can't sit there and wait for that to happen before I start.
The solution is to use MIT-ish licensed software instead. Therefore, I license my own projects with the MIT license, so I can accept contributions to them and still be able to use them at work. Had I ever accepted contributions to my GPL projects without copyright assignment, pieces of those projects would be GPL *and* copyright to someone else, and I wouldn't be able to use my own projects at work!
I read about the outage first on *slashdot*. You can't tell me the "Internet" was knocked out. It's "parts of the Internet that did not have power" that were knocked out. I mean come on, do they expect the public sewage system to work when there's no water?
In my state and locality, there are laws that say that negligence in protecting your possessions means you share the guilt for crimes committed using those possessions. For example, there's a well-established body of law that if you leave your keys in the ignition of a car and someone steals it and runs someone down with it, you are liable for negligence. It has nothing to do with whether you "implicitly" did anything--what you actually did was fail to protect people by properly securing the things you're responsible for securing.
Law is eternally a gray area, but there is a strong argument to be made that people running wide-open are liable for negligence when crimes are committed on those connections.
Most interesting is the assertion that the decision by Red Hat to end support for its free distribution and Novell's aquisition of SUSE marks not only the death of free software, but actually is a validation of Microsoft's business model.
OSS is not a business model. It's a bunch of different things: a community, a way of developing software, a way of distributing software, a way of thinking about information. But not a business model. Business models can be built on top of OSS, but OSS doesn't care. If those business models crumble--and indeed, many will--OSS will remain, to build on again.
I wouldn't tear apart a single one of your line items. I'm someone migrating our small office to a Linux backend and, in the future, a Linux desktop as well. Our customer-facing services are being rewritten in Python to run on Linux as well.
Your list is not only accurate but pretty complete. With these things, almost any workplace can use Linux.
Here's the thing: All of those capabilities exist in the Linux world, but they are not all integrated seamlessly into any single system. They can be, of course, and such is the power of Linux, but doing so basically means rolling your own distro. Rolling your own distro is labor, and labor costs money, and there goes your free.
I honestly think Debian has the best chance of integrating all these features before any other distro, because Debian is focused on integrating packages with each other. "Integrating" in Debian just means creating a.deb for each of those bullets. Most already exist in Debian testing/unstable or the 3rd-party archives: openoffice.org, kernel drivers for some very odd hardware configurations, Mozilla Firebird and Thunderbird and their respective plugins, and Wine. The only thing there that can't be handled by a well-crafted deb is ISV support; that will always have to be purchased, but I assure you it is possible to pay money for this service.
Once the debs exist, the Debian-based system becomes bulletproof and idiotproof. Then you discover that your maintenance costs on the whole installation go to practically zero.
Does this mean we can eliminate that retarded law that says banks can't accept a cell phone line on an account application? This law was written as part of our absurd War on Some Drugs, and is just another example of the government getting carried away with "protecting" us by taking away our rights.
Slashdot Mirror
It doesn't really "poison" the filters, because there are just wayyyyyy too many posible words for this to work. Bayesian filters assign a huge probability of spamminess to every word in a spam email and an exceedingly low prbability of spamminess to every word in a non-spam email during training. If a word appears in both, it just averages out. Over time a given word will appear only once in a spam email with a bunch of random words, and many times in non-spam emails, and therefore after some time (or even pre-emptively) the good words will be recognized as good. The more training, the better; poisoning has little chance of success as long as there's at the number of good and bad emails going in are within an order of magnitude of each other.
That's not to say the technique doesn't help the spammers in the short run; it probably gets past less sophisticated and trained filters.
Still Control Panel > User Accounts. Or Ctrl-Alt-Delete > Change Password if you prefer. Windows usability sucks, but don't make shit up.
The interesting thing about his comments about desktop Linux are that he's making them at all. He used to have a position of "Linus is what it is, I don't care where it goes, it's just fun to watch." He's not doing that so much now that it appears to be actually getting the places people imagined it would go 5-10 years ago. To make a specific claim, even one as flexible as that, is out of character for him and shows that he's starting to become interested in seeing his work succeed commercially (other than in the areas he works on directly).
They also need IBM's code to produce the evidence.
So they can recopy it, change the comments from "(C) IBM 1973" to "(C) SCO 1972" and send it back.
If they have the common good in their interests, they really should patent this process, then place the patent in the commons. (I'm not sure how this process works with patents, but I'm confident it can be done somehow.) To do so would be to protect future generations from patent mischief with this application.
It's not entirely a misnomer if there's an NFS server in it. (I'm still downloading it, haven't tried it yet.)
I have a real problem with any plan which involves hiding a problem away and hoping that a future generation will figure out how to deal with it.
Obviously, what we can do with the waste is shoot it into space. Problem solved.
How many lossy re-encodings would it take before you have nothing but gray noise? 10? 100? 1000? Someone should set up a script and find out.
Metrosexual? Bling-Bling? Well, I would never use them in a sentence, but if some people find one of them useful, then let them use it. Like many entries in the list, they are words that annoy the committtee them because they're new, not because they're useless or overused. If it really is useless, it will soon cease to be used. That's how language works.
Punked? Actually a useful word, I think. There's no word for "play a elaborate prank on", yet. I think this one annoys them only because it comes from pop culture.
Companion animals: Their smarmy comment "They're called PETS." is probably downright offensive to people who object to the negative connotation of "pet". I call my dogs pets, but I love them, and I can certainly see them as companions.
Smoking gun: "came to us from Iraq". Uh, no it didn't. This term has probably been around for at least decades. Overused? Maybe this year, but that's certainly not cause for banning it evermore.
They're trying to be funny, but they're just demonstrating a failure to understand language evolution.
when i visited the page mozilla firebird cpu spiked up to 99 quite quickly, and quickly fell to 0/1-ish when i closed the tab.
I just realized that the situation with the entertainers/manual laborers is a bit more complicated than what I described. It revolves around this question: Am I purchasing all the items they need to do their jobs?
This includes 150 costumes (except possibly for the ladies dancing; we'll assume she's at least wearing a g-string), plus 36 milking stools, 36 buckets and enough bottles to hold however much milk is produced. (The song doesn't say what you're going to do with the milk produced, but I'd definitely want to keep it. It's the one thing you're getting for free out of all this.) There's bound to be some sheet music and music stands too; it'd be pretty terrible to have all those pipers playing improv. What about permits to have all these people perform for almost a week? Cleaning costs....
"On the first day of christmas my true love gave to me: X"
"On the second day of christmas my true love gave to me: Y"
etc. and each time X or Y includes all the previous items. The only obvious conclusion is that you're getting 12 partridges in 12 pear trees, 22 turtledoves, etc.
The total of the items includes combined items though, which are necessarily counted as separate. For example, you might be able to find a store that sells partridges, but probably not one that puts a pear tree in the same box.
So we've got 12 pear trees, the 142 birds, 40 rings, 40 cows, 76 women, 30 lords (is that more expensive than "30 guys jumping around in fancy dress"?), 22 pipers, 22 pipes, 12 drummers, 12 drums and 12 sets of drumsticks (they could be bongos, but traditionally they are not depicted as such). You might argue that the various musicians bring their instruments and then take them home with them when the gig is done. One assumes this is not slavery, so you have to figure salary for however many days the individuals in question are employed (and figure for the sake of simplicity that they are all let go at the end of the 12th day).
So, depending on whether or not you get to keep the musical instruments after christmas, you have purchased either 234 or 280 items and paid salaries for 150 man-days of entertainment.
I'm sure they're just lining up. It's the opportunity of a lifetime.. to help SCO secure their internal systems so nobody else can log in and wreak havoc through, say, a backdoor placed there by the idiots currently running the SCO network on the advice of their helpful friends.
There's yet another type of forking that can occur which I call "forking up". This is where a small but important module is developed by someone who needs it, and the module later gets adopted into a project with broader scope and larger visibility. This is in fact the norm when projects submit code for inclusion into the standard distribution of Python or Perl. If Python adopts the module, sometimes independent development on it dies, but many times it does not--pybsddb and pyxml are two examples that spring to mind. Other projects do this as well... Twisted and Zope both get a lot of contributions from smaller projects that want to be part of a larger effort.
This is yet another case that supports parent's argument that resources are not being divided.. resources that work on fixing bugs in the higher-visibility project would never have worked on the smaller project's codebase at all. The smaller project can still benefit from porting bugfixes in stdlib over to their tree.
Everybody wins.
Companies typically defend themselves against this kind of attack by having patents on lots of stuff. Surely there is lots of code in Linux that is eligible for a patent, and can be assigned to some holding company. Such an entity could then offer to do patent trades whenever MS or some other big software bully decides to swing the patent club.
(BTW, don't bother to tell me that this patent only covers embedded devices formatted with the fs, because Microsoft could change its mind.)
Good one. Knowing how to program isn't a skill anyone can just acquire by being exposed to computers. After all, implementation hiding is one of the principles of interface design. So people exposed to the interfaces of software are being shielded from how it works internally, and they aren't going to absorb that knowledge. And even if they were exposed to the guts all day long, programming has a conceptual foundation in mathematics and needs a great deal of patience and practice--practice developing your memory to hold long logic chains mentally until they come to fruition in the software, practice knowing what patterns work and what don't, practice building disciplined habits.
That said, I've always argued that anyone can become a computer programmer. But the skills are not the sort of thing you pick up just by using a computer all your life. You have to seek out training beyond that.
I do agree with his point about the service model of software development. It's just not going to be anywhere near as hard as he claims it will be to get a job that way.
Personally I lean toward the "SCO Upper Management is a bunch of fucking loonies" theory. Here's an article that puts concrete numbers on the effect SCO is having on linux adoption.
4 .a sp
http://blogs.ittoolbox.com/linux/archives/00019
16% may be a big number or a small number, depending on your point of view, but it's not what I'd call a palpable blow to Linux.
I am a software developer. "Use" to me, means "distribute as part of a software package." Not "execute the runtime." As I already stated, if I distribute a relocatable library as part of a software package, I am bound by the terms of the GPL. This seriously hurts the software's reuse.
Some developers like the license. Some of us find out just how troublesome it can be, and change our minds.
My earlier projects were distributed under the GPL or LGPL. Then I got a job. I have to develop software for an organization that doesn't have the ability to release source code; we are a subsidiary of a larger organization that technically owns the copyright of everything I produce at work.
If I want to create software that uses GPL'd libraries, even if the software I create is a completely separate source tree and I don't change a single line of code in the distributable source of the library, the software I create has to be GPL. This is simply unacceptable; I don't have the time to get permission to do that on every project from my boss, let alone from the head of the parent organization and their lawyers. Eventually, every one of my creations at work may become OSS (and indeed, one of them already has) but I can't sit there and wait for that to happen before I start.
The solution is to use MIT-ish licensed software instead. Therefore, I license my own projects with the MIT license, so I can accept contributions to them and still be able to use them at work. Had I ever accepted contributions to my GPL projects without copyright assignment, pieces of those projects would be GPL *and* copyright to someone else, and I wouldn't be able to use my own projects at work!
Appropriate term from the gay community which is otherwise innocuous to slip by (think about goatse.cx if you can't figure it out)...
Too bad it's also confusing.
I read about the outage first on *slashdot*. You can't tell me the "Internet" was knocked out. It's "parts of the Internet that did not have power" that were knocked out. I mean come on, do they expect the public sewage system to work when there's no water?
In my state and locality, there are laws that say that negligence in protecting your possessions means you share the guilt for crimes committed using those possessions. For example, there's a well-established body of law that if you leave your keys in the ignition of a car and someone steals it and runs someone down with it, you are liable for negligence. It has nothing to do with whether you "implicitly" did anything--what you actually did was fail to protect people by properly securing the things you're responsible for securing.
Law is eternally a gray area, but there is a strong argument to be made that people running wide-open are liable for negligence when crimes are committed on those connections.
Most interesting is the assertion that the decision by Red Hat to end support for its free distribution and Novell's aquisition of SUSE marks not only the death of free software, but actually is a validation of Microsoft's business model.
OSS is not a business model. It's a bunch of different things: a community, a way of developing software, a way of distributing software, a way of thinking about information. But not a business model. Business models can be built on top of OSS, but OSS doesn't care. If those business models crumble--and indeed, many will--OSS will remain, to build on again.
I wouldn't tear apart a single one of your line items. I'm someone migrating our small office to a Linux backend and, in the future, a Linux desktop as well. Our customer-facing services are being rewritten in Python to run on Linux as well.
.deb for each of those bullets. Most already exist in Debian testing/unstable or the 3rd-party archives: openoffice.org, kernel drivers for some very odd hardware configurations, Mozilla Firebird and Thunderbird and their respective plugins, and Wine. The only thing there that can't be handled by a well-crafted deb is ISV support; that will always have to be purchased, but I assure you it is possible to pay money for this service.
Your list is not only accurate but pretty complete. With these things, almost any workplace can use Linux.
Here's the thing: All of those capabilities exist in the Linux world, but they are not all integrated seamlessly into any single system. They can be, of course, and such is the power of Linux, but doing so basically means rolling your own distro. Rolling your own distro is labor, and labor costs money, and there goes your free.
I honestly think Debian has the best chance of integrating all these features before any other distro, because Debian is focused on integrating packages with each other. "Integrating" in Debian just means creating a
Once the debs exist, the Debian-based system becomes bulletproof and idiotproof. Then you discover that your maintenance costs on the whole installation go to practically zero.
Does this mean we can eliminate that retarded law that says banks can't accept a cell phone line on an account application? This law was written as part of our absurd War on Some Drugs, and is just another example of the government getting carried away with "protecting" us by taking away our rights.