What of it? He's talking about a Nuclear Winter, we'd be back in the stone age very quickly. I hope someone like you doesn't have his finger on the trigger.
Time to move past your 80's-style nuclear paranoia. Even if the U.S and Soviet Union had a full-out nuclear exchange, the idea that the entire planet would be so cloudy as to block all sun, freeze us, and kill all humans is just bogus. Again, 80's-style paranoia.
And, given current realities, we're not going to have a full-out nuclear war. China could lob everything it has against us and we could completely nuke China out of existence and the vast majority of the world (including most of the US) would survive just fine. China would be the only one left standing when the music stopped.
Pre-planned? Perhaps. The prank-puller(s) could have hiked there before and remembered to bring something funny for next time. Could have been their first trip but they had already checked out the lie of the land on the net first, and found the camera in their googling. Or could have been completely spontaneous.
Uh, it's pretty clear it was pre-planned and not spontaneous. The alternative is that someone was just hiking around this volcanic island with a blow-up Dino with him when he saw the webcam and decided to leave his blow-up buddy on the island.
Which raises the question what the guy was doing on this island with a blow-up Dino. But that may explain the goofy smile on Dino's face.:)
And they certainly have achieved fantastic press with this slashdot exposure: suddenly a large group of people know the name, what it does, how it works and how much it costs...
... And an SMTP server(s) that we can add to our "denied hosts" file to filter all incoming crap from them. If someone cares enough to add a webbug to their emails to violate my privacy, I care enough to filter them.
Would you now give your e-mail address away in the way you gave it away in 1994? Time was I was quite happy to post my address on the web as a mailto href and the link text of the href. I don't do that with my current address, which I've had since 1999, and I get about one spam a month, without any filters.
Congratulations. Spam isn't a problem for you. Obivously, based on all the noise on the subject, laws being passed, etc. you are not representative of the population as a whole. Spam is a problem for many of us and many of us do not feel that being chased from one email address to the next is an acceptable option. But I'm glad to hear it's worked for you.
I get both ham and spam which both contain just a zip file, just an image, or just a link to a web site that I have never seen before. The zip files can not be filtered because we get customer data that way. I also only get to see the spam not the ham. How are the Bayesian filters supposed to deal with this? The "best" way around bayes filtering is to keep your message short.
YOur points are valid. However:
1. The fact that you get emails with nothing but a zip file, a link, or am image, is NOT typical. So normally there is more to go on. You could be the exception rather than the rule on this point.
2. The message headers are a huge part of what Bayesian considers. Even in the absence of any real content, the headers themselves can make or break the case for spam in many, many cases.
3. Over time you will receive ham with the above characteristics from people you "know" so their particular headers (not to mention the "From" line) will become known to your Bayesian fiter.
4. As with any filter, you can improve performance by whitelisting those you often communicate with.
Bayesian filtering is a great solution at stopping you from seeing spam, but it does nothing to actually make it go away.
I'm not going to wait for spam to go away. Bayesian is something we can use now and it lets me get work done (or surf on Slashdot) rather than clicking delete every 2 minutes as my mail icon indicates a new mail has arrived.
Besides, in the end, bandwidth is cheap when compared to my time. If, eventually, someone can figure out how to make spam go away then that'll be a great help to reduce bandwidth usage and reduce server resources needed to filter my mail. But, in the grand scheme of things, that's a relatively small part of the cost of spam. The largest cost of spam is in humans having to take time to deal with it. And Bayesian solves that aspect of the problem.
Some of us don't like changing our email addresses every few weeks. My email address hasn't changed since 1994. I'm getting about 650 spams per day.
Luckily, with Bayesian filtering we aren't forced to abandon our email addresses to avoid spam. We can keep using the same address we've had for a decade and just lose the spam. Problem solved and no running involved, much less notifying dozens or hundreds of contacts you've established over the years what your new email address is.
In this day and age, anyone who isn't using Bayesian filtering for spam protection is just begging for spam. In this day and age, anyone that claims that Bayesian filters can't solve the current spam problem has either tried the wrong Bayesian-based product or doesn't know what he/she is talking about. And those that say Bayesian filtering won't work "forever" because spam will evolve to get around it are preaching gloom and doom about a future that is far from certain.
I've been using Bayesian filtering for 1-1/2 years and my success rate just keeps inching up. In May of last year I was at 99.5% success. My Bayesian corpus has grown in the last year and so far this month I'm averaging 99.98% success--only two have gotten through, and one of those was in a foreign language. I can't even remember the last time a pornographic spam got past my Bayesian filter.
And in the last year we've seen silly attempts to get around Bayesian filters, such as packing the message with lots of random words, or excerpts from books or the Constitution or what have you. Time and time again those messages actually get a higher Bayesian score than they would have if they had just left the random words out.
There is still no known effective way to get around Bayesian filtering. I personally don't think there will be a way around them, but regardless: Bayesian has been the answer for nearly two years and shows no sign of weakening in the near future. USE IT!.
Everything on my Linux system is free, except Win4Lin. The Gimp doesn't measure up to even the other open source programs I use. I actually prefer OpenOffice to Office, I prefer XMMS to WinAmp or CoolPlayer, I prefer Opera to Mozilla/Eudora on Windows.
There are plenty of quality apps with a GUI for Linux. That The Gimp is free is no excuse to have a crappy interface that is completely unintuitive.
I agree. I have always used Paint Shop Pro for years under Windows. Now that I've switched to Linux I've all but abandoned Microsoft Office in favor of Open Office. I use Opera instead of Eudora. I used Kate (awesome!) instead of notepad. I use Xine for movies and DVDs and XMMS works great for music.
But when it comes to working with images I still have to run Win4Lin to open a Win98 session and run Paint Shop Pro. The interface on The Gimp is just unusable to me. And maybe it has all the same features as Paint Shop Pro, but at least with PSP I can find them.
FWIW, I have no problem paying for software, even Linux software. I do have a problem with what appeared to be an open-source project being closed up, buried under patents, and the whole thing being sold to the public for a fee with the excuse that there was no choice.
I don't know and I don't know, but I've wondered the same thing. Unfortunately, I didn't save a copy of the version I downloaded early last year since I (stupidly, it seems) thought I'd always be able to find the same thing online. Now it turns out if I want the same thing I have to pay.
So I just use my modem as 14k. I use it so infrequently that that's fine. Certainly a pain when I do need it, but it's so infrequent that I'd rather not buy a license on principle.
All I know is that about a year ago I installed Redhat 7.3 on my laptop and loaded the Linuxant drivers (or whatever they were called then). It was free and I got either 33k or 56k and it worked great.
Late last year I upgraded to Redhat 9.0 and went back to reload the drivers. Now they insist that if I want anything better than 14k that I have to pay money and use the Conexant IP as an excuse to charge for it. Screw that!
Someone was providing relatively high-speed software for my modem last year for free. Now they force me to 14k if I don't pay money? Forget that. They can take their modem drivers and shove it up their bit bucket. I scarcely use my modem to start with. I almost always plug into a network or find some wireless access point. I would like to have my modem work just so that all the hardware I paid for on my laptop works, but it seems to me like sometime last summer the whole Conexant project went from a free open-source program to a rather closed-source, non-free program. That's not what I was looking for when I switched to Linux.
Spam evolves with time. As you wrote your spam filter, you can probably see why this is a problem for a Baysian system.
Why is evolving spam a problem for a Bayesian system? To the contrary, a Bayesian system is one of the best filters available for an evolving set of spam. Evolution doesn't usually happen in a single event which means the new tactic is implemented while old indicators of spam (headers, if nothing else) are still present in the incoming spam. So the new spam technique may not be immediately recognized as spam but other areas of the message will be--then, over a short amount of time, the new technique itself will be recognized as spam. At that point it doesn't matter if the old spam indicators disappear--the new tactic will be recognized by Bayesian.
My Bayesian filter just continues to increase in accuracy and has been moving up in accuracy since I started my corpus about a years ago. I'm at 99.97% so far this month with 96,000 spams and 8000 good mails in my corpus. Compare that with 99.35% from last June when I had about 4000 spam and 800 good emails in my corpus.
I see no evidence that an increasingly large corpus reduces the effectiveness of Bayesian. Nor have I seen any spammer countermeasure in the last year that has succeeded at circumventing my filter on any ongoing basis.
See another post somewhere above yours. It is a design failure. Yes, given the inefficient design of FAT it might not have made sense to have long-filename. This is simply evidence that the design itself is inefficient. A variable filename format could have been implemented that would use only the space needed for the filename length in question. As the other post said, tell the user: "You can use 255 character filenames, but you're using up disk space to do it." That simple.
I live in Mexico. Everything, including M&Ms, is more expensive in Mexico (except the ticket price itself which is currently hovering around US$5 for an evening show).
Me:...they can't see the difference between a real threat to security in the U.S. and sour-grapes retaliation against Americans, I'll spend my money at home.
You: I think tourists around the world are concluding the same about travelling to the USA which only hurts the US economy and puts your countrymen out of work.
So other countries are retaliating by implementing the same measures against the U.S. which hurts their economy and puts their countrymen out of work? At least the U.S. is doing it based on a demonstrated past threat. Countries that do it as a result of sour grapes and based on no credible security risk coming from the U.S. are hurting themselves for no good reason.
If you think that in the future, you as an American will somehow exempt from beign treated like a criminal everytime you cross a border you have another thing coming.
I have no problem with security and if security requires that I be treated as a criminal, fine. But, as of yet, I've seen no evidence of security risks to Europe from Americans. That is to say, if they're going to implement the same draconian security across the board, fine. If they're going to single out Americans for special security treatment despite a total lack of past security threats from the U.S. then that's just stupid, sour grapes, a waste of security resources, and hurting their own economy for no substantial increase in their security (since security threats don't historically come from the U.S.).
But, again, go for it. I'll just stay home. I have plenty to enjoy in the U.S., Canada, and Mexico. No need to sit on a plane for 10 hours to look at old buildings in Europe.:)
No kidding. The first time I snuck food into a theater was because I wanted a double cheeseburger from McDonalds in the same mall but there wasn't time to eat it and make it to the movie. So we stuffed it in my wife's purse. That's sort of when I realized how easy, better tasting, and of course cheaper "external" food is. We now regularly smuggle double cheeseburgers in, I almost always bring my own M&Ms, and my wife has gone so far as to smuggle in Sushi! I do still buy my soft drink at the the theater, though.
I know the theater's need to make a buck, but charging $2 for M&Ms that cost about 70 cents outside is just gouging.
Of course, the obligatory Steve Wright reference: "I got arrested for bringing my own food to the movie theater. My argument was that the concession stand prices were outrageous. Besides, I hadn't had a good barbeque in a long time!"
The lesson is clear: stay out of movie theaters and you won't get arrested.
The lession is clear: Don't take a videocamera to the movie theaters and you won't get arrested.
I hate the RIAA and the MPAA as much as the next guy, but come on, does anyone really think someone should be allowed to bring a videocamera in to the move theater, tape the movie, and not get arrested?
I hate to state the obvious, but measures should be based on potential usefulness and potential threat, not on retaliation.
Do I think it's cool that foreigners have to go through this added security? No! My wife is Mexican and on her last trip to the U.S. she had to be fingerprinted and photographed (to their credit she said the INS person was very friendly and the fingerprinting/photograph process was extremely quick). But the reality is that after 9/11 the U.S. has some serious justification for wanting to know who is entering the country. A bunch of anonymous faces is no longer acceptable. Why? Because the terrorists came from overseas, including from Europe (Germany). How many terrorists in Europe have come from the U.S.?
Further, getting the personal information ahead of time allows the government to research those that enter before they arrive. I suppose an alternative would be for arrivals to stand in line in immigration for 6-8 hours while their information is reviewed on arrival after
Finally, the U.S. isn't asking for anything that is out of line with what other countries often ask for. I don't know about Europe, but for me to live in Mexico (I'm an American) I had to fill out an application with a bunch of personal information, including my fingerprints. To bring my car into the country I had to provide a credit card. I do not consider this inappropriate: I am a guest in their country and they have a right to ask for this information. If I don't want to provide it, I have the right to not live here.
I don't support collecting information on domestic travelers (which is what this thread is really about), but I do support the government having all the information possible on who is coming into the country. This is justified based on where past threats have come from. Countries such as Brazil which decide to institute retaliatory security procedures against Americans despite the fact that no security threat has been demonstrated to be of American origin are just practicing sour grapes and, quite frankly, will not be getting any of my tourism dollars. I understand their frustration but if they can't see the difference between a real threat to security in the U.S. and sour-grapes retaliation against Americans, I'll spend my money at home.
Is there some reason why some people have this inexplicable need to have something other than a "ring" as a ringtone? And that this need is so great that they'll actually pay for it?
When my phone rings, it rings. My wife once wanted some song as the ring on her phone. So I used the phone's ring-editor and made it. I can't understand why anyone pays for such useless junk.
Spending money is always better then a tax cut which is pure flushing money down the drain and doesn't help anyone.
Oh man, you've been at the Democratic Institute for Tax & Spending Economic Theory too long. It's my money, not the government's money. The government has a responsibiity to do as much as it can while stealing as little of my money as possible. How anyone can suggest that the government stealing and spending more of my money is always better than letting me keep my money is just incredible.
The whole belief that government somehow knows how to spend money better than the taxpayers or somehow does it more efficiently betrays the underlying B.S. in much liberal-think.
Hmm, I wrote the Bayesian filter that I am using. The idea of "overtraining" doesn't make sense to me and after nearly 100,000 combined good/bad messages my filter is still creeping up in accuracy. I've seen no evidence that such an overtraining has or will happen, nor does it make sense to me logically.
I wonder if it was due to a bad implementation used by the Bayesian filter you are using?
Slashdot Mirror
Time to move past your 80's-style nuclear paranoia. Even if the U.S and Soviet Union had a full-out nuclear exchange, the idea that the entire planet would be so cloudy as to block all sun, freeze us, and kill all humans is just bogus. Again, 80's-style paranoia.
And, given current realities, we're not going to have a full-out nuclear war. China could lob everything it has against us and we could completely nuke China out of existence and the vast majority of the world (including most of the US) would survive just fine. China would be the only one left standing when the music stopped.
Uh, it's pretty clear it was pre-planned and not spontaneous. The alternative is that someone was just hiking around this volcanic island with a blow-up Dino with him when he saw the webcam and decided to leave his blow-up buddy on the island.
Which raises the question what the guy was doing on this island with a blow-up Dino. But that may explain the goofy smile on Dino's face. :)
Look Lane, I've been going to this high school for seven and a half years... I'm no dummy.
Congratulations. Spam isn't a problem for you. Obivously, based on all the noise on the subject, laws being passed, etc. you are not representative of the population as a whole. Spam is a problem for many of us and many of us do not feel that being chased from one email address to the next is an acceptable option. But I'm glad to hear it's worked for you.
YOur points are valid. However:
1. The fact that you get emails with nothing but a zip file, a link, or am image, is NOT typical. So normally there is more to go on. You could be the exception rather than the rule on this point.
2. The message headers are a huge part of what Bayesian considers. Even in the absence of any real content, the headers themselves can make or break the case for spam in many, many cases.
3. Over time you will receive ham with the above characteristics from people you "know" so their particular headers (not to mention the "From" line) will become known to your Bayesian fiter.
4. As with any filter, you can improve performance by whitelisting those you often communicate with.
I'm not going to wait for spam to go away. Bayesian is something we can use now and it lets me get work done (or surf on Slashdot) rather than clicking delete every 2 minutes as my mail icon indicates a new mail has arrived.
Besides, in the end, bandwidth is cheap when compared to my time. If, eventually, someone can figure out how to make spam go away then that'll be a great help to reduce bandwidth usage and reduce server resources needed to filter my mail. But, in the grand scheme of things, that's a relatively small part of the cost of spam. The largest cost of spam is in humans having to take time to deal with it. And Bayesian solves that aspect of the problem.
Luckily, with Bayesian filtering we aren't forced to abandon our email addresses to avoid spam. We can keep using the same address we've had for a decade and just lose the spam. Problem solved and no running involved, much less notifying dozens or hundreds of contacts you've established over the years what your new email address is.
I've been using Bayesian filtering for 1-1/2 years and my success rate just keeps inching up. In May of last year I was at 99.5% success. My Bayesian corpus has grown in the last year and so far this month I'm averaging 99.98% success--only two have gotten through, and one of those was in a foreign language. I can't even remember the last time a pornographic spam got past my Bayesian filter.
And in the last year we've seen silly attempts to get around Bayesian filters, such as packing the message with lots of random words, or excerpts from books or the Constitution or what have you. Time and time again those messages actually get a higher Bayesian score than they would have if they had just left the random words out.
There is still no known effective way to get around Bayesian filtering. I personally don't think there will be a way around them, but regardless: Bayesian has been the answer for nearly two years and shows no sign of weakening in the near future. USE IT!.
There are plenty of quality apps with a GUI for Linux. That The Gimp is free is no excuse to have a crappy interface that is completely unintuitive.
But when it comes to working with images I still have to run Win4Lin to open a Win98 session and run Paint Shop Pro. The interface on The Gimp is just unusable to me. And maybe it has all the same features as Paint Shop Pro, but at least with PSP I can find them.
So I just use my modem as 14k. I use it so infrequently that that's fine. Certainly a pain when I do need it, but it's so infrequent that I'd rather not buy a license on principle.
Late last year I upgraded to Redhat 9.0 and went back to reload the drivers. Now they insist that if I want anything better than 14k that I have to pay money and use the Conexant IP as an excuse to charge for it. Screw that!
Someone was providing relatively high-speed software for my modem last year for free. Now they force me to 14k if I don't pay money? Forget that. They can take their modem drivers and shove it up their bit bucket. I scarcely use my modem to start with. I almost always plug into a network or find some wireless access point. I would like to have my modem work just so that all the hardware I paid for on my laptop works, but it seems to me like sometime last summer the whole Conexant project went from a free open-source program to a rather closed-source, non-free program. That's not what I was looking for when I switched to Linux.
Why is evolving spam a problem for a Bayesian system? To the contrary, a Bayesian system is one of the best filters available for an evolving set of spam. Evolution doesn't usually happen in a single event which means the new tactic is implemented while old indicators of spam (headers, if nothing else) are still present in the incoming spam. So the new spam technique may not be immediately recognized as spam but other areas of the message will be--then, over a short amount of time, the new technique itself will be recognized as spam. At that point it doesn't matter if the old spam indicators disappear--the new tactic will be recognized by Bayesian.
My Bayesian filter just continues to increase in accuracy and has been moving up in accuracy since I started my corpus about a years ago. I'm at 99.97% so far this month with 96,000 spams and 8000 good mails in my corpus. Compare that with 99.35% from last June when I had about 4000 spam and 800 good emails in my corpus.
I see no evidence that an increasingly large corpus reduces the effectiveness of Bayesian. Nor have I seen any spammer countermeasure in the last year that has succeeded at circumventing my filter on any ongoing basis.
You: I think tourists around the world are concluding the same about travelling to the USA which only hurts the US economy and puts your countrymen out of work.
So other countries are retaliating by implementing the same measures against the U.S. which hurts their economy and puts their countrymen out of work? At least the U.S. is doing it based on a demonstrated past threat. Countries that do it as a result of sour grapes and based on no credible security risk coming from the U.S. are hurting themselves for no good reason.
If you think that in the future, you as an American will somehow exempt from beign treated like a criminal everytime you cross a border you have another thing coming.
I have no problem with security and if security requires that I be treated as a criminal, fine. But, as of yet, I've seen no evidence of security risks to Europe from Americans. That is to say, if they're going to implement the same draconian security across the board, fine. If they're going to single out Americans for special security treatment despite a total lack of past security threats from the U.S. then that's just stupid, sour grapes, a waste of security resources, and hurting their own economy for no substantial increase in their security (since security threats don't historically come from the U.S.).
But, again, go for it. I'll just stay home. I have plenty to enjoy in the U.S., Canada, and Mexico. No need to sit on a plane for 10 hours to look at old buildings in Europe. :)
I know the theater's need to make a buck, but charging $2 for M&Ms that cost about 70 cents outside is just gouging.
Of course, the obligatory Steve Wright reference: "I got arrested for bringing my own food to the movie theater. My argument was that the concession stand prices were outrageous. Besides, I hadn't had a good barbeque in a long time!"
The lession is clear: Don't take a videocamera to the movie theaters and you won't get arrested.
I hate the RIAA and the MPAA as much as the next guy, but come on, does anyone really think someone should be allowed to bring a videocamera in to the move theater, tape the movie, and not get arrested?
Do I think it's cool that foreigners have to go through this added security? No! My wife is Mexican and on her last trip to the U.S. she had to be fingerprinted and photographed (to their credit she said the INS person was very friendly and the fingerprinting/photograph process was extremely quick). But the reality is that after 9/11 the U.S. has some serious justification for wanting to know who is entering the country. A bunch of anonymous faces is no longer acceptable. Why? Because the terrorists came from overseas, including from Europe (Germany). How many terrorists in Europe have come from the U.S.?
Further, getting the personal information ahead of time allows the government to research those that enter before they arrive. I suppose an alternative would be for arrivals to stand in line in immigration for 6-8 hours while their information is reviewed on arrival after Finally, the U.S. isn't asking for anything that is out of line with what other countries often ask for. I don't know about Europe, but for me to live in Mexico (I'm an American) I had to fill out an application with a bunch of personal information, including my fingerprints. To bring my car into the country I had to provide a credit card. I do not consider this inappropriate: I am a guest in their country and they have a right to ask for this information. If I don't want to provide it, I have the right to not live here.
I don't support collecting information on domestic travelers (which is what this thread is really about), but I do support the government having all the information possible on who is coming into the country. This is justified based on where past threats have come from. Countries such as Brazil which decide to institute retaliatory security procedures against Americans despite the fact that no security threat has been demonstrated to be of American origin are just practicing sour grapes and, quite frankly, will not be getting any of my tourism dollars. I understand their frustration but if they can't see the difference between a real threat to security in the U.S. and sour-grapes retaliation against Americans, I'll spend my money at home.
When my phone rings, it rings. My wife once wanted some song as the ring on her phone. So I used the phone's ring-editor and made it. I can't understand why anyone pays for such useless junk.
Oh man, you've been at the Democratic Institute for Tax & Spending Economic Theory too long. It's my money, not the government's money. The government has a responsibiity to do as much as it can while stealing as little of my money as possible. How anyone can suggest that the government stealing and spending more of my money is always better than letting me keep my money is just incredible.
The whole belief that government somehow knows how to spend money better than the taxpayers or somehow does it more efficiently betrays the underlying B.S. in much liberal-think.
I wonder if it was due to a bad implementation used by the Bayesian filter you are using?