Too many people try to fit themselves into a field that they think they like, only to find that they have no aptitude for it, or that the nuts and bolts of the field aren't interesting to them.
This is true in college as well. In some of my 4000-level CS classes there were a few people who got into CS because they thought "oh, computers would be fun," and it's quite obvious they have NO idea what's going on. They ask questions to which they should have learned the answers in some of the 1000- and 2000-level classes. They may work hard, but hard work is meaningless without understanding.
I had to work with a couple people like this on a group assignment for an OOP class (4000-level at my university), and it was a nightmare; of the 8 or so people in the group, about 3 had any concept of what we were doing, and a couple of the others couldn't program - AT ALL. One of the non-programmers actually spent his days working a programming job. It continues to boggle my mind to this day.
I describe this phenomenon as an "Accountability Void." No one is responsible for spam.
Accountable, no. Responsible, yes. The spammer, regardless of what s/he says, is responsible for their own misuse of the network. They just can't be held accountable since, for the most part, they hide themselves well enough that nobody can find out who they are. The spammer knows s/he's doing something underhanded, otherwise s/he wouldn't have to use fake accounts that get closed in 6 hours.
Accountability and responsibility are two very different animals.
Of course, there's a big responsibility vacuum in society today, and the spammer is just a really good example of it.
my problem here with these guys statements is that they all...makes these swooping opinionated statements without any meat to back them up.
Well, that's the thing about opinions - you can just say "I don't like it", and that's pretty much good enough. Opinions are like assholes, after all - everybody's got one, and they all stink. The one guy said that he didn't like the new start menu. Making a comment like that requires that he know the differences between the new and the old, so it's fairly obvious to me that he's at least tried both of them out. I can't even say that much, because I haven't been around an XP installation for long enough to even know that there is a new start menu.
To be real honest, I can't say I liked the old start menu - the DWYTIM (do what you think I mean) option-hiding just pissed me off to no end. I want to see all the options, not a subset, and especially not a subset that was decided by some programmer who I've never met, based on his assumptions and biases, which just happen to be grossly invalid for me. And as far as unsophisticated users go, I can't see that seeing a subset, which is subject to change by some force you don't know about, could be simpler; I would think the "Aunt Tillie" user might be frightened or confused by this ("Where did my program/file go?! Oh, no! I'm screwed!"). I think I posted a rant on this very topic a while ago, but I'm too lazy to go find the link.
Capitalism should have no problem eliminating overzealous, opressive DRM.
OK, capitalism should have no problem with this, I agree on that much. However, at the rate the "democracy" is going, capitalism won't even get a chance to do its job. The legislators will get told lots of horror stories by the RIAA/MPAA/Microsoft lobbyists with the big suitcases full of cash, and of course they'll believe them. After all, what business that has so much money could possibly be leading them astray? Legislation will be purchased, and you and I will get left out in the cold, wondering why we can't play our new CD in both the stereo in the living room and the boombox in the garage.
I have some faith in the invisible hand of the marketplace. But I have less than no faith in our supposed leaders, who actually seem more interested in adding the new wing on their third mansion than actually doing their jobs, which is actually representing the people.
I once read an interview of Arthur C. Clarke, who had said that almost all good sci-fi is about people. People interacting with other people, or with technology, in a technologically-advanced setting. Look at 2001 - it was (at its core) about the interaction between a man and a human-like computer. Star Wars? Groups of people fighting other groups of people. Most of Clarke's stuff is about how people react and interact with new technology.
I, too, greatly enjoyed "Inner Light", and many of the other character driven eps, like "Face of the Enemy", "Darmok", "Lessons", even "Lower Decks", which showed us a few characters which we would never have seen otherwise. We get to learn about the characters as almost-real people, and to paraphrase Clarke, it's all about the people.
Their products are flawed beyond belief, regardless of the fact that they're sniffable. I've got an opener from these chuckleheads, and it's the biggest piece of crap I've ever seen. Not only do I have to drag the ladder out and reset the thing about weekly, but there's about a 40% chance of the door actually opening on the first button-push. The rest of the time the light just comes on, or the door will move a couple inches, or something equally not-opening-the-door will happen. Perhaps those are other security features they offer with their products...?
...wasn't copyright originally intended to protect the little guy?
Not from the interpretation of the dissenters. I managed to get a dozen or so pages into each of the dissenting opinions (I read about that much of the majority opinion as well), and the point that the dissenters wanted to make was that the main part of the copyright/patent thing was to protect the public. If the works in question never pass into the public domain, who loses? The public. Breyer's dissenting opinion goes into considerable detail on this point (in addition to being the easier to read of the two, IMNSHO). The other dissenter talks a lot about the history of the various copyright extension legislations, and apparently there have been quite a number of Congressional copyright acts which benefitted one company, some of which were from the distant past (lke about 1807 or so). The majority opinion talks about the fact that the Constitution doesn't specifically spell out what a "limited time" is, and the fact that the current copyright terms do have a limit, so the test is fulfilled.
That said, all of the opinions are serious slogs to try to read, so make sure you're very awake when you start...
Having driven on the real Laguna Seca a couple weeks ago, I had to get on GT3 and see how it compared. It was very close! There were some tiny differences: there's a bit of a dip on the inside of turn 6, which didn't quite make it into the game, and turn 5 seemed a little flatter than in reality. And IIRC, the game doesn't have any traces of the old track route, which in reality helps a little for coming into turn 2 (the hairpin). Even some of the scenery was the same: coming over turn 8 (the corkscrew), there are 3 trees at the far edge of the track which you can use to aim through the yet-invisible turn; they're just the same in the game (drive straight to the right-most one from your apex). My line on the real track translated exactly to the game with no changes, except braking and shift points, since the cars were different. A few of the impact walls are in different places, but if you've gotta worry about them, you've got a whole other set of problems.:)
So I think it very unlikely indeed that X Windows was named for MS Windows
They aren't at all, because they aren't called "X Windows". In the X(7) manpage, we find (reformatted slightly to defeat lameness filter):
The X Consortium requests that the following names be used when referring to this software:
X
X Window System
X Version 11
X Window System, Version 11
X11
X Window System is a trademark of X Consortium, Inc.
Now that I'm done being pedantic (sorry man), I'd be curious to see what the W system was all about; W is supposed to be the father of X. I would guess that the W stands for "window(ing)". Anybody know anything about it?
It's pretty likely that since they're working on embedded systems, that they don't include any of the GNU tools. So Linux (the kernel) is the right thing to say.
Yeah, and all the while Jeff "I need some more patents" Bezos, with his poster-child bullshit patent, keeps clamoring for patent reform? Man, I just don't get that guy. I've been boycotting Amazon since that stuff came out too.
...their products ported and tailored to run on it.
If they actually end up doing something like this, they'll probably do something like what they did for IE for Solaris. They didn't bother porting the actual application, they instead wrote an entire Windows-on-Solaris layer which runs underneath IE, and makes it huge and slow. It works, but it's a 40 or 50MB beast. The X server itself doesn't even eat that much!
I loathe RedHat with all my heart, but it's what everyone and their mothers use, so it'll be the most useful for them.
And this attitude is a large part of why the computer industry is the way it is today. "It's what <other_group> uses, so we need to use it too." Wrong. If it's what will get the job done, then use that. Otherwise, there's probably something else out there that will get the job done, and more simply/quickly/intuitively/etc.
I would suggest that for a class such as this, something like Debian would be ideal, in no small part because of the simplicity of 'apt-get'. No awful dependencies, no finding the right file to install, it just happens. Showing that Linux can be simple is a nice goal, don'tcha think?:) And the fact that Debian is pretty much all Free Software might help get the FS message out there too.
Namely, that the books promote witchcraft, Satanism, and ideas that are diametrically opposed to the moral foundation that this country was founded on.
Except for the minor fact that the idea of which you speak is freedom from religious persecution. Last I checked, Satanism, Wicca, and the like were religions. Perhaps they're not religions which you support, but they're religions, and are guaranteed freedom by the first amendment to the constitution to worship in the manner of their choosing. Of course if they start killing sacrificial virgins and the like, the cops might not appreciate it, but by and large, there's nothing you can do about it.
And anyway, this is an old argument; the Religious Reich trotted this out in the early 80's when Dungeons and Dragons was first hitting it really big. Wasn't true then, and it still isn't.
The file/folder metaphor may have probems for newbies but the only real problem is that file (particularly with Unix style file systems) may have more than one name. This is a feature not a bug.
I'm going back and forth on this "save the inode, not the fname" thing, and the author could have something here. It could be pretty cool, since the program could call it anything it wanted, and it wouldn't matter (files can have more than one name, as you said). You wouldn't need to care where it was saved, just that it was accessible to your program. The whole problem is that some programs, when faced with writing the file to disk, write a new file, with a new inode, and rename the old file to fname~ or even unlink the old version. The inode that points to the proper file is now pointing to the old file, or unallocated disk space (depending on how the smart program/OS handles filesystem ref-counting). And then another problem I just though of, what happens when you need to use more than one application to handle the file? Then you're stuck with reimplementing the filesystem, on top of the already-existing filesystem, so the other application can get to the file. Crap.
There's something there, I think, but there are some pretty important problems to overcome before it can happen.
if someone sends something that's still mostly spam... then all the neutral words will be tarnished as well.
The way a filter such as this works is all about word frequency: if a single spam contains a neutralish word, then it becomes more spammy based on the total number of times that word has been seen in all mail. If it's been seen many times in innocent mails, the one-time appearance in a single spam won't taint it much.
If a word is neutral (that is, it signals a ~0.50 probability of the mail being a spam), it's not going to count much toward the final probability anyway. Graham's prototype filter took only the 15 most interesting words into account, where interestingness is abs(word_probability - 0.50). Words that are truly neutral don't actually matter, just those which are a strong indicator one way or the other.
I've written a filter of this type myself, and once it's been trained reasonably well, it's quite accurate. I've looked through my word probabilities list, and some of the results are surprising, but they work. My userid actually signals a 31% chance of the mail it appears in being spam, but by the filtering criteria, that's actually a pretty uninteresting word (i.e. close to 50%, which is absolutely neutral, and completely uninteresting). As the number of spams I use to train approaches the number of legitimate mails I use to train, I expect my userid's spam probability will more closely approach 50%.
MS: Hey, we're going to give you free software, computers, and all sorts of stuff, FREE! SN: Sounds like a deal. Bring forth the booty! MS: Oops, we're not going to give you the free computers, or the free OS, but we will give you this maaahhhvelous office suite for free... SN: Sit and spin, bitches.
It sounded like they were in negotiations of some sort (probably Microsoft wanted to get the "Namibia replaces Linux with Microsoft" headlines, of course), in which Microsoft was promising all sorts of free stuff, and then they came back later and changed the deal for whatever reason. And SchoolNet called them on it.
The technique sounds a bit like an ICMP "source quench" (see RFC 896 for some earlier work on the topic), though interpreted by the intermediate routers, instead of the endpoints. So yeah, that's about the right word for it.
Why do you think they break "kernel module source compatibility" with every patchlevel release?
I saw a response from Linus one time (probably on LKML a couple years ago) about why this is, and it has nothing whatsoever to do with ideology. It's to make driver maintainers verify their code when there is an API change. For a large driver API change, they have to go through everything and make absolutely sure everything will work. Otherwise if the interface is changed just a little bit, the driver maintainer might not go through every API call, and subtle bugs can creep in.
Finally, dachshund has gotten to the kernel of the case! Congress's responsiveness to the private citizen has basically gone away. They don't have much time for you unless you have a lot of money; it's basically a big payola scheme (the US, run by payola - that makes me shudder). If Congress starts making laws that completely pervert the intent of what the Constitutional framers intended, which is exactly what they're doing nowadays, then who can stop them? The President? He's a big rubber stamp; don't count on much intelligence there. The Supreme Court also has the power to limit Congress, and this is exactly correct way to go about it.
The only part that concerns me comes from the notes that whoever took at the proceedings. According to his report, one of the justices said that the court does not do "policy", but only "legality". This seems a bit more policy-like from where I sit.
Slashdot Mirror
Too many people try to fit themselves into a field that they think they like, only to find that they have no aptitude for it, or that the nuts and bolts of the field aren't interesting to them.
This is true in college as well. In some of my 4000-level CS classes there were a few people who got into CS because they thought "oh, computers would be fun," and it's quite obvious they have NO idea what's going on. They ask questions to which they should have learned the answers in some of the 1000- and 2000-level classes. They may work hard, but hard work is meaningless without understanding.
I had to work with a couple people like this on a group assignment for an OOP class (4000-level at my university), and it was a nightmare; of the 8 or so people in the group, about 3 had any concept of what we were doing, and a couple of the others couldn't program - AT ALL. One of the non-programmers actually spent his days working a programming job. It continues to boggle my mind to this day.
I describe this phenomenon as an "Accountability Void." No one is responsible for spam.
Accountable, no. Responsible, yes. The spammer, regardless of what s/he says, is responsible for their own misuse of the network. They just can't be held accountable since, for the most part, they hide themselves well enough that nobody can find out who they are. The spammer knows s/he's doing something underhanded, otherwise s/he wouldn't have to use fake accounts that get closed in 6 hours.
Accountability and responsibility are two very different animals.
Of course, there's a big responsibility vacuum in society today, and the spammer is just a really good example of it.
You might want to check your math on this. By my calculations, 600-700 of 1800 is 33-38%. Therefore being less than 40%.
You were saying about making up statistics...?
my problem here with these guys statements is that they all...makes these swooping opinionated statements without any meat to back them up.
Well, that's the thing about opinions - you can just say "I don't like it", and that's pretty much good enough. Opinions are like assholes, after all - everybody's got one, and they all stink. The one guy said that he didn't like the new start menu. Making a comment like that requires that he know the differences between the new and the old, so it's fairly obvious to me that he's at least tried both of them out. I can't even say that much, because I haven't been around an XP installation for long enough to even know that there is a new start menu.
To be real honest, I can't say I liked the old start menu - the DWYTIM (do what you think I mean) option-hiding just pissed me off to no end. I want to see all the options, not a subset, and especially not a subset that was decided by some programmer who I've never met, based on his assumptions and biases, which just happen to be grossly invalid for me. And as far as unsophisticated users go, I can't see that seeing a subset, which is subject to change by some force you don't know about, could be simpler; I would think the "Aunt Tillie" user might be frightened or confused by this ("Where did my program/file go?! Oh, no! I'm screwed!"). I think I posted a rant on this very topic a while ago, but I'm too lazy to go find the link.
Capitalism should have no problem eliminating overzealous, opressive DRM.
OK, capitalism should have no problem with this, I agree on that much. However, at the rate the "democracy" is going, capitalism won't even get a chance to do its job. The legislators will get told lots of horror stories by the RIAA/MPAA/Microsoft lobbyists with the big suitcases full of cash, and of course they'll believe them. After all, what business that has so much money could possibly be leading them astray? Legislation will be purchased, and you and I will get left out in the cold, wondering why we can't play our new CD in both the stereo in the living room and the boombox in the garage.
I have some faith in the invisible hand of the marketplace. But I have less than no faith in our supposed leaders, who actually seem more interested in adding the new wing on their third mansion than actually doing their jobs, which is actually representing the people.
I once read an interview of Arthur C. Clarke, who had said that almost all good sci-fi is about people. People interacting with other people, or with technology, in a technologically-advanced setting. Look at 2001 - it was (at its core) about the interaction between a man and a human-like computer. Star Wars? Groups of people fighting other groups of people. Most of Clarke's stuff is about how people react and interact with new technology.
I, too, greatly enjoyed "Inner Light", and many of the other character driven eps, like "Face of the Enemy", "Darmok", "Lessons", even "Lower Decks", which showed us a few characters which we would never have seen otherwise. We get to learn about the characters as almost-real people, and to paraphrase Clarke, it's all about the people.
I read a thing on what POSIX stood for one time:
P - portable
O - operating
S - system
I - interface
X - all proper OS's end in X
Heh.
Their products are flawed beyond belief, regardless of the fact that they're sniffable. I've got an opener from these chuckleheads, and it's the biggest piece of crap I've ever seen. Not only do I have to drag the ladder out and reset the thing about weekly, but there's about a 40% chance of the door actually opening on the first button-push. The rest of the time the light just comes on, or the door will move a couple inches, or something equally not-opening-the-door will happen. Perhaps those are other security features they offer with their products...?
Not from the interpretation of the dissenters. I managed to get a dozen or so pages into each of the dissenting opinions (I read about that much of the majority opinion as well), and the point that the dissenters wanted to make was that the main part of the copyright/patent thing was to protect the public. If the works in question never pass into the public domain, who loses? The public. Breyer's dissenting opinion goes into considerable detail on this point (in addition to being the easier to read of the two, IMNSHO). The other dissenter talks a lot about the history of the various copyright extension legislations, and apparently there have been quite a number of Congressional copyright acts which benefitted one company, some of which were from the distant past (lke about 1807 or so). The majority opinion talks about the fact that the Constitution doesn't specifically spell out what a "limited time" is, and the fact that the current copyright terms do have a limit, so the test is fulfilled.
That said, all of the opinions are serious slogs to try to read, so make sure you're very awake when you start...
Having driven on the real Laguna Seca a couple weeks ago, I had to get on GT3 and see how it compared. It was very close! There were some tiny differences: there's a bit of a dip on the inside of turn 6, which didn't quite make it into the game, and turn 5 seemed a little flatter than in reality. And IIRC, the game doesn't have any traces of the old track route, which in reality helps a little for coming into turn 2 (the hairpin). Even some of the scenery was the same: coming over turn 8 (the corkscrew), there are 3 trees at the far edge of the track which you can use to aim through the yet-invisible turn; they're just the same in the game (drive straight to the right-most one from your apex). My line on the real track translated exactly to the game with no changes, except braking and shift points, since the cars were different. A few of the impact walls are in different places, but if you've gotta worry about them, you've got a whole other set of problems. :)
So I think it very unlikely indeed that X Windows was named for MS Windows
They aren't at all, because they aren't called "X Windows". In the X(7) manpage, we find (reformatted slightly to defeat lameness filter):
The X Consortium requests that the following names be used when referring to this software:
X Window System is a trademark of X Consortium, Inc.
Now that I'm done being pedantic (sorry man), I'd be curious to see what the W system was all about; W is supposed to be the father of X. I would guess that the W stands for "window(ing)". Anybody know anything about it?
It's pretty likely that since they're working on embedded systems, that they don't include any of the GNU tools. So Linux (the kernel) is the right thing to say.
Yeah, and all the while Jeff "I need some more patents" Bezos, with his poster-child bullshit patent, keeps clamoring for patent reform? Man, I just don't get that guy. I've been boycotting Amazon since that stuff came out too.
If they actually end up doing something like this, they'll probably do something like what they did for IE for Solaris. They didn't bother porting the actual application, they instead wrote an entire Windows-on-Solaris layer which runs underneath IE, and makes it huge and slow. It works, but it's a 40 or 50MB beast. The X server itself doesn't even eat that much!
I loathe RedHat with all my heart, but it's what everyone and their mothers use, so it'll be the most useful for them.
And this attitude is a large part of why the computer industry is the way it is today. "It's what <other_group> uses, so we need to use it too." Wrong. If it's what will get the job done, then use that. Otherwise, there's probably something else out there that will get the job done, and more simply/quickly/intuitively/etc.
I would suggest that for a class such as this, something like Debian would be ideal, in no small part because of the simplicity of 'apt-get'. No awful dependencies, no finding the right file to install, it just happens. Showing that Linux can be simple is a nice goal, don'tcha think? :) And the fact that Debian is pretty much all Free Software might help get the FS message out there too.
Yeah, I've been trolled, whatever...
Namely, that the books promote witchcraft, Satanism, and ideas that are diametrically opposed to the moral foundation that this country was founded on.
Except for the minor fact that the idea of which you speak is freedom from religious persecution. Last I checked, Satanism, Wicca, and the like were religions. Perhaps they're not religions which you support, but they're religions, and are guaranteed freedom by the first amendment to the constitution to worship in the manner of their choosing. Of course if they start killing sacrificial virgins and the like, the cops might not appreciate it, but by and large, there's nothing you can do about it.
And anyway, this is an old argument; the Religious Reich trotted this out in the early 80's when Dungeons and Dragons was first hitting it really big. Wasn't true then, and it still isn't.
The file/folder metaphor may have probems for newbies but the only real problem is that file (particularly with Unix style file systems) may have more than one name. This is a feature not a bug.
I'm going back and forth on this "save the inode, not the fname" thing, and the author could have something here. It could be pretty cool, since the program could call it anything it wanted, and it wouldn't matter (files can have more than one name, as you said). You wouldn't need to care where it was saved, just that it was accessible to your program. The whole problem is that some programs, when faced with writing the file to disk, write a new file, with a new inode, and rename the old file to fname~ or even unlink the old version. The inode that points to the proper file is now pointing to the old file, or unallocated disk space (depending on how the smart program/OS handles filesystem ref-counting). And then another problem I just though of, what happens when you need to use more than one application to handle the file? Then you're stuck with reimplementing the filesystem, on top of the already-existing filesystem, so the other application can get to the file. Crap.
There's something there, I think, but there are some pretty important problems to overcome before it can happen.
if someone sends something that's still mostly spam... then all the neutral words will be tarnished as well.
The way a filter such as this works is all about word frequency: if a single spam contains a neutralish word, then it becomes more spammy based on the total number of times that word has been seen in all mail. If it's been seen many times in innocent mails, the one-time appearance in a single spam won't taint it much.
If a word is neutral (that is, it signals a ~0.50 probability of the mail being a spam), it's not going to count much toward the final probability anyway. Graham's prototype filter took only the 15 most interesting words into account, where interestingness is abs(word_probability - 0.50). Words that are truly neutral don't actually matter, just those which are a strong indicator one way or the other.
I've written a filter of this type myself, and once it's been trained reasonably well, it's quite accurate. I've looked through my word probabilities list, and some of the results are surprising, but they work. My userid actually signals a 31% chance of the mail it appears in being spam, but by the filtering criteria, that's actually a pretty uninteresting word (i.e. close to 50%, which is absolutely neutral, and completely uninteresting). As the number of spams I use to train approaches the number of legitimate mails I use to train, I expect my userid's spam probability will more closely approach 50%.
The way I read it was like this:
MS: Hey, we're going to give you free software, computers, and all sorts of stuff, FREE!
SN: Sounds like a deal. Bring forth the booty!
MS: Oops, we're not going to give you the free computers, or the free OS, but we will give you this maaahhhvelous office suite for free...
SN: Sit and spin, bitches.
It sounded like they were in negotiations of some sort (probably Microsoft wanted to get the "Namibia replaces Linux with Microsoft" headlines, of course), in which Microsoft was promising all sorts of free stuff, and then they came back later and changed the deal for whatever reason. And SchoolNet called them on it.
I've been getting some spams lately which were trying to sell me software to reduce my spam. Maybe that was a prank too?
Man, I haven't laughed that hard in at least a week. Thanks, whoever!
The technique sounds a bit like an ICMP "source quench" (see RFC 896 for some earlier work on the topic), though interpreted by the intermediate routers, instead of the endpoints. So yeah, that's about the right word for it.
sludg-o, you are officially my hero for this Friday. This comment has made my day. :)
Why do you think they break "kernel module source compatibility" with every patchlevel release?
I saw a response from Linus one time (probably on LKML a couple years ago) about why this is, and it has nothing whatsoever to do with ideology. It's to make driver maintainers verify their code when there is an API change. For a large driver API change, they have to go through everything and make absolutely sure everything will work. Otherwise if the interface is changed just a little bit, the driver maintainer might not go through every API call, and subtle bugs can creep in.
Finally, dachshund has gotten to the kernel of the case! Congress's responsiveness to the private citizen has basically gone away. They don't have much time for you unless you have a lot of money; it's basically a big payola scheme (the US, run by payola - that makes me shudder). If Congress starts making laws that completely pervert the intent of what the Constitutional framers intended, which is exactly what they're doing nowadays, then who can stop them? The President? He's a big rubber stamp; don't count on much intelligence there. The Supreme Court also has the power to limit Congress, and this is exactly correct way to go about it.
The only part that concerns me comes from the notes that whoever took at the proceedings. According to his report, one of the justices said that the court does not do "policy", but only "legality". This seems a bit more policy-like from where I sit.