As far as Mexicans earning more than Indians, that Computerworld I mentioned above says, India avg programmer salary $6400, Mexico $5150. Source, NeoIT Inc of San Ramon, CA. Maybe it's right, maybe it's wrong, I'm no expert. Just reporting what I read.
Ok, then you are forgiven because you had data you based your comment on.:) But I can tell you first-hand that NeoIT is wrong. Although I earn American rates and work primarily with Americn companies I've lived in Mexico for the last 8 years. The only software developers that earn $5150 a year would be those maybe right out of college or finishing college, or opting to work at a very low-paying firm. $10,000/year would be much more believable and $15,000/year would not be surprising.
Even when those multiple articles prove with undisputable, acknowledged football statistics that what Rush said was right. Even other sportswriters are confessing it to be true. But don't let reality or facts get in the way of your preconceived notions and stereotypes. Hold your banner up high and shine a light on it for all to see, that way we'll all know who to snicker at.:)
Except prices up a little faster than they otherwise would, causing inflation to increase just a little more, causing every company in the country to be pressured to raise their employees' salaries just a little bit more to compensate for that added cost of living. No, a minor adjustment to the minimum wage every few years isn't going to cause the sky to fall, but obviously it has an effect.
Simply do a graph. If we plot the current minimum wage and number of people employed on the graph, then set an arbitrary minimum wage of $100/hr and the number of people employed we have our two extremes. There is a curve on the graph that connects those and, yes, increasing the minimum wage costs jobs. Otherwise we would just pass a $1000/hr minimum wage law and we could all retire after 2 or 3 years.
You can't legislate prosperity. The only thing a minimum wage can and should do is protect the absolute low-end of the workforce from gross exploitation. It's to avoid sweatshops. It is NOT to make sure that an employee at McDonald's can afford a car, a 27" color TV, cable, and still have money left over for a vacation in Cancun. Those that strive for such goods will have to educate themselves, work hard, get ahead, and start leapfrog into better jobs. The fact that McDonald's doesn't provide a sufficient income to live as well as most people want should be MOTIVATION for these people to better themselves.
FYI, Indian companies already outsource to China, today. China, Eastern Europe/Russia, Vietnam, Mexico, etc.
I can't imagine India outsourcing to Mexico. I've seen two projects firsthand in MEXICO outsourced to India, and Mexicans DO earn more than Indians. So I'd be surprised to see India outsource to Mexico.
In many cases, they've just cancelled the offshore projects and re-hired the original consultants.
You're 100% right. I have seen overseas outsourcing in action twice. Both times it was a *MEXICAN* company outsourcing to India. Both times the project just dragged on, drained money, and honestly didn't produce results. In one case, what was supposed to be a $500k contract in 6 months dragged on for a year and a half and they doled out $2 million as the Indian company kept asking for more and more time, more and more money, etc.
Once you sink $500k into a project it's hard to write it off as a loss--so they just kept spending more and more money on it. Eventually, they just cancelled the Indian outsourcing, brought it back home to Mexico--and even had me (an American) on board as a consultant to help them!
Andy Grove says that all the telecomm infrastructure makes it as if a guy in India was in the next cubicle. That's the simplistic promise of outsourcing that tempts companies to do it. There may be some industries where it will work. Software development is NOT one of them.
I'm a software engineer and I'm NOT worried about outsourcing at all. Yes, we'll have some short-term pain as many companies experiment with it. But the two cases I've seen of MEXICAN companies trying to save a buck by outsourcing to India made it very clear that it just won't work long-term. I completely believe many jobs will be outsourced to India and China, but many (if not most) will bounce back to the U.S. after the 1-2 years it takes to realize it has failed.
But if you've read enough of his articles
you'd realize that he knows as much about this topic as Rush Limbaugh knows about american football.
And ifyouread enougharticles, you'd know that what he said last week was plenty accurate and his assessment was more accurate than many that supposedly "know football." But keep carrying that liberal banner way up high if it makes you feel better about yourself. Just don't be surprised when others give you about the same credibility as we give Rob Enderle.
Besides, how do you explain "statistical intrusion detection" to the average home user who just wants to read e-mail and surf the Web?
Probably about the same way you explain TCP/IP to the average home user who just wants to read e-mail and surf the web. You don't. That doesn't mean it can't be of use to the user even if he or she doesn't understand it--or probably even knows it exists.
The various reports of spammers "getting around the Bayesian filters" that have surfaced since the appearance of Graham's first article are completely inconsistent with Graham's own results, virtually proving that the people implementing a lot of this stuff are not RTFA.
Bingo! I love it when people say that spammers are successfully getting around Bayesian filters. Either those Bayesian filters were poorly implemented or the people saying it just have something against Bayesian filtering and don't really know what they're talking about.
The percentage of spam being caught by my Bayesian filter continues to climb (currently around 99.7%, just like Graham's) even in the face of random words and inserted "neutral text." You have to break Bayesian pretty bad for any of those tricks to actually get a piece of spam through the filter.
Filtering is merely automating "just hit delete." It still gets sent, it still travels the wires to your box, it still hits your spool.
Unless you earn only pennies per hour, "just hitting the delete key" is the most expensive part of spam. Like I said in my original message, I agree that it would be better if it weren't sent at all. But the largest cost to me--and to SOCIETY--is the time it takes to DEAL with spam, not the bandwidth--although I don't deny the bandwidth is a cost... just not as significant as my time.
and worse, now you're spinning extra cycles to scan the mail.
Better my CPU consume some cycles than spam consume my time. My CPU is cheaper than my time.
Just hit delete means you kill 1000 this month -- and 10000 a year later.
Again, re-read my above message. In the short-term, yes. But as more and more people use filters fewer people will receive spam and even fewer will respond to it. Over time spam becomes less lucrative so there is less reason to send it.
It seems to me like you are looking for an overnight fix. There isn't one. But effective and pervasive spam filtering is a solution that keeps me from having to deal with spam on a daily basis in the short-term, and will lead to a reduction of spam when it no longer becomes profitable to send spam since everyone filters it.
That's my point. There are a lot of implications to saying that the filter's decoder must do exactly what the GUI's decoders does.
Oh, I misunderstood. But even so, the filter decoder doesn't have to do EVERYTHING the GUI decoder does. All it has to do is decode and parse, it doesn't have to make a visual presentation.
For example, it is sufficient to be able to decode MIME messages (including mutli-part), handle quoted-printable, handle Base64, and be able to decode HTML and quoted-printable escape sequences. You then just throw out all HTML comments, only keep HTML commands you are interested in (FORMs, IMGs, HREFs, FRAMEs, FONTs), throw out multiple whitespace, and then preferably parse out all remaining content into either headers, text, HTML, or script. You can then filter base on each of those content areas as you like.
What I mean is you certainly don't have to be able to execute Javascript or anything complicated like that. And short of anything executable (which would be dangerous anyway) the rest isn't really that hard.
In fact, one thing Paul Graham didn't address in his latest pieces is the risk of following links that can have embedded identifying information in them. So, yes, you slam their server but you may have inadvertently just told them WHO you are by requesting a certain URL.
Many times the message decoder and the filter are two different systems. The filter might be middlewear and the decoder might be part of the GUI. Under those circumstances, the message decoder can present readable text to the user that the filter ignores (read: that's bad).
I've never seen a filter that works in that fashion. Even if the filter is middleware (which is common), it must have an integrated message decoder. If it doesn't it is useless. The message needs to be fully decoded during spam filtering and then, probably, again when it is displayed to the user.
Bayesian filtering is a solution whose time came long ago. It works outstandingly for me
Great! I'll tell you, it almost makes spam fun for me. I look at my statistics and just grin at all the garbage my filter kept out of my life. "Let's see, how many spams did my filter block THIS month?":)
However, that's why email programs and ISPs should have Bayesian spam filtering. In email programs, it should be enabled by default,
I agree 100%. I personally think they should be enabled by default both in email programs and by ISPs--of course with the option of disabling them. But the idiots won't bother to disable them so they'll be protected, and those that don't want the ISP meddling with filtering can disable it.
I agree, the spam volume won't go down until everyone--especially the idiots--start using Bayesian, whether they make the effort to do so (doubtful) or whether they simply don't make the effort to disable it (probable).
But still, I must say... as much as I'd like to see spam stop wasting bandwidth, the most expensive aspect of spam is the time it wastes for the receiver. And Bayesian fixes that for us whether the spam is still being sent or not.
I want to know why Baysian filters cant checksum (not MD5, a real honest to goodness, simple checksum) so that when you've got a file with checksum 48299014 and another with 48291633 a red flag goes up.
That wouldn't be a Bayesian filter, it'd be a simple checksum filter. And the reason no-one does it is because usually every spam sent out is slightly different. Spammers insert random garbage words, etc. which would make a simple checksum different for every spam.
So far analyzing the IMG of spams is unnecessary. The HTML content alone used to display even a single image is usually sufficient for Bayesian to conclude it's spam. There's no reason to make things more complicated than they need be by analyzing IMGs, etc. The evidence is right there in the spam they send you.
We've seen first hand how the early Bayesian filters were circumvented.
How early? I haven't seen that.
Remember the images instead of text, then the HTML Entities (like instead of the letter 'A')? The second and third generations of the Bayesian filters had to account for them.
ALL filters, Bayesian or otherwise, have to be able to properly interpret and translate an email. It is insane to think you can filter email without being able to properly decode it. This is a limitation of the message decoder, not Bayesian filtering.
I can just see how a DoS filter would be circumvented early: redirects and browser scripts.
As Paul Graham said in an earlier piece (although surpirisngly not in this one), if your goal is to detect spam and an email link takes you to a redirect or browser script that's a high probability of spam. If you want to pound the spammer as suggested in this latest article, follow the redirect. Yes, you will also pound the first site but perhaps they'll notice it and take down the redirect altogether. Either way the end result is that gullible people that would otherwise follow the link to purchase from the spammer won't be able to get to the target site. It will either be Slashdotted or the redirect will be removed completely. In either case the end goal is achieved.
During the past 100 years, there has ALWAYS been a way of copying works, and yet, people have been paying for *copies* of music for years.
But, up until the last 6 or 7 years, it has generally been more hassle and time-consuming to copy the works rather than just buy them. That is no longer the case and changes the situation significantly.
If, while I'm working, I can download a song I want in the background and have it in 5 or 10 minutes for free or stop working and go to the mall and pickup a FREE CD, I'd rather just download the song. That saves me the hassle of going to the store, picking it up (even if it's free), and ripping it. And if I can do it without identifying myself or sharing my credit card number, all the better!
especially given that people have *always* been happy to pay for the priviledge of being able to reproduce music through speakers when ever and where ever they are.
And kings and queens used to travel their countries in horse-drawn carriages. That doesn't mean you'll see kings and queens today thrilled to pay to use obsolete technology. Just because something has always been so doesn't mean it always will.
I'm afraid a generation of teenagers has grown up through their most music-buying years in the digital era. My sister-in-law started downloading music when she was 18. She's now 21. I think she has something like 4000 MP3s and she has a MODEM. I don't think she's purchased a single CD in those 3 years.
Times are changing. They always have and always will.
but I am confident that people in developed nations that have the money dont mind paying for anything when they think they're paying a fair price. Thats the very defition of fair.
I agree with you. But the question is, "What is fair?" I'll be blunt--most of the music out there isn't worth a penny. It's not even worth the drive to the store or the effort to whip out my credit card.
That said, there's a lot of music out there that I wouldn't be willing to buy but I *would* be willing to pay $20 to see in concert.
What "makes sense" to a person has nothing to do with what is most likely to be correct.
Yes and no. Can reality be something other than I can imagine? Sure. But can an educated person take existing knowledge and formulate a theory that is more likely right than wrong? Yes.
These guys are seeing soccer balls because they want to see soccer balls. There is no reason, at this time, for us to believe that if we travel in a certain direction on the universe that we'll end up where we started. There is no observational evidence to suggest that. The article itself indicates that data to that effect is completely lacking, and no-one has been able to detect the "repeat patterns" that you would see if we were seeing a small universe from lots of different angles.
Now, if we traveled in some direction and found ourselves back where we started then certainly we would want to analyze why that happened.
On what evidence is your belief that the universe is infinite based? Why does it trump these guys' evidence?
Read the article. The article itself suggests problems in their conclusion and indicates that an infinite universe could produce the same data they've acquired.
So... we have the possibility that the universe is pretty much as we expect it and the data collected does not necessarily contradict that. Or we can conclude the universe is a soccer ball and that we'll get back to where we started if we go off in a certain direction long enough.
By the same logic, we can conclude there is a wormhole over the Pacific at the International dateline. The world is flat but when you go past 180 degrees we travel through a warp that takes us back to the other side of the flat world.
Yeah, maybe that's possible, but it's more likely the world is just round.
They are only winning to those that don't do anything to help themselves.
The Verisign SiteFinder was a bad thing, obviously, but I laughed at the reaction "It's breaking my spam filter." What kind of archaeic, obsolete spam filter were these people using?
Likewise, that spammers are using trojaned systems is bad, of course. Any system compromise is bad. But this is just normal virus and hacking. It doesn't make it any harder to get rid of your spam.
I've said it once and I'll say it again, Bayesian filers is the solution. It works today and it depends on no-one but yourself to start using it. Since I started using it in May, I've received 20,596 spams--of those I've seen 89 of them. I.e., only 0.43%. It comes out to one spam every other day, though that's deceptive since probably half of those that got by were cases of a single spam sent 5 times in rapid-fire mode and they all happened to get through at once--the same spam 6 hours later would've been filtered. In reality, I'd guess I see one spam per week. In a perefect world I wouldn't see any, but that's good enough for me in this imperfect world.
Now, some will say "But that doesn't solve the bandwidth problem." In the short-term, no, it doesn't. But in the short-term it doesn't waste my time which is my single largest expense when it comes to spam. And, in the long-term, if more people started using Bayesian the response rate on spam would continue to plummet making it less and less useful to spam in the first place.
But those that are being bothered by spam on a daily basis simply aren't using the tools and technology that are available to them, and have been for over a year.
True. I was purposefully trying to be very liberal with how much time it would take and allowing ample time to, perhaps, export to some external database or system.
Anyway, the point is that they paid an absurd amout of money for a site that virtually anyone could have done in less than a month. And if the person had experience and was efficient I agree with you that we'd be talking about hours or a few days rather than weeks or months.
Now some may just say my mind is small, but to some extent I think Occam's Razor applies here. I'm all for scientific investigation, and the excercise of contemplating weird shapes for the universe may lead to interesting discoveries.
But which really makes more sense? That the universe is a giant void filled with matter in various areas? Or that some weird mechanics cause the universe to actually be small and traveling in one direction brings us back to where we started as if we were traveling on the surface of a globe.
Personally, I think the "universe" is infinite. That's not to say that there are galaxies and stars filling every space of the infinite area, but I think if you are at any place in the universe you can always continue further in any direction even if you don't find anything there but the absence of matter. I don't think you're going to hit a wall where you can't go further nor reach a point that "wraps you around" to the the other side of the universe.
Perhaps the definition of "universe" is the "outer" extent of where matter can be found. In this sense, the universe seems to be expanding--but it's not that the universe is really getting bigger, just that existing material is moving further and further out into the available void which is infinite. My opinion only, of course.
I won't say these guys are wrong, but it just seems that what they are suggesting is probably more complicated than it needs to be or probably is. But at least they got their 15 minutes of fame and were able to mention "soccer" and "science" in the same sentence.
illegal p2p file sharing will lose ground to legit music download services. most people -will- pay to get the right song at the right bit rate at the best download speeds the first time.
Maybe, but I don't think so. P2P will win because it's anonymous, easy, and free. You don't have to give up your name or credit card number, you don't have to logon to some website, you don't have to deal with some non-standard music file format. Just power up the app and download the song you want and you're golden.
The whole "pay for Internet music services" is based on the premise that CDs may be obsolete but the concept of paying for recorded music is still valid. I believe that that whole concept is obsolete. Music is free. Concerts, merchandise and endorsements is where the money will be for those musicians that want to go that route. But music itself is free.
Any product or service that fails to understand that will fail. That's why there's, what, 4 million users on Kazaa at any given time and yet iTunes is thrilled because they sold a million songs in a couple weeks? Kazaa probably does that in a few hours.
Legal or not, people have tasted free music and it tasted good--especially when the RIAA was charging $15-$20 for the same thing at the time. I don't think there's any way to sugar-coat the fact that you're asking them to now pay for something they've grown accustomed to getting for free, especially younger ones that just like downloading MP3s because they like to collect them (whether or not they like the music) and also since they don't have much money to start with.
The RIAA should've introduced a pay-for-download music service 6 months before Napster came out. Now, it's too late.
Yeah, well to me the amazing thing in the story is that the government apparently paid AT&T $3.5 million to build the website. Have you visited the website? I've built more complex websites in a matter of weeks. Even charging $200/hour that would be easily less than $32k.
I would hope that "building" the site for $3.5 million also includes running it, ongoing maintenance, etc. Because if the government really paid AT&T $3.5 million to BUILD it and still has to pay some ongoing fee, they got ripped by an order or two of magnitude.
Government waste isn't surprising, but it's sad when it is made so obvious. A good percentage of the folks here at Slashdot could have done just as good a job for a fraction of the cost and STILL recorded a very good year income-wise.
So I have to wonder why it is such a proposterous notion that I wish to live in the lower->middle class bracket without performing my music live if my work is in fact played in clubs, bedrooms, etc across the nation?
If people are willing to pay for your music without you performing and you are able to make a lower, middle, or even upper class living, great! That's the free market. What I've been saying for years is that the free market is slowly reaching the conclusion that it is NOT willing to pay just for music. And, unfortunately, that's not even your fault. It's the RIAA. They've robbed so many customers that now customers are just not willing to pay for music anymore, period.
So it's not a proposterous notion that you want to make money doing what you love. You are mistaken, however, if you think that the simple fact that people are listening to your music entitles you to that money. The free market will dictate what you earn, and no-one is entitled to anything more than what the free market is willing to pay them.
It is the loud minority that cause posters like you to suggest that musicians are just looking for a Get Rich Quick
I apologize if that's what my post caused you to believe. I know a number of musicians and I know they aren't trying to get rich quick. I was really referring to the RIAA-type musicians, and those that aspire to be RIAA-type musicians. My mistake.
Pair programming is uncomfortable on our reduced space. And it's noisy. Are the inconveniences worth it?
If I'm at a computer and NOT typing, I'm falling asleep. Honestly. I've had people try to explain their code to me when they left a company and I literally had to try to stay awake, and that might have only been for a couple of hours. If I had to watch someone else work for hours on end I would be bored and want to do some real work--if I was able to stay awake.
Theoretical programming styles don't interest me much, and probably each applies to certain projects better than others. But I definitely would not work for a company that wants me to sit around and watch someone else work, nor would I employ someone to sit around and watch someone else work.
Pre-recording days the musicians of the world did earn their keep by playing live music, very few got rich doing so but I'm sure the majority made an adequate living just like the rest of society.
I agree completely. Is there some reason why musicians think they should be rolling in the big bucks? When a hamburger flipper earns $15k/year, some entry level positions $30k, and software engineers with decent experience make $50k, is there something particularly special about a musician that entitles him to earn $200k, $500k, a few million per year? Yes, a musician has special talents that allow him to compose or perform music that others enjoy. So? Software engineers have special talents that allow him to develop software that hopefully make the lives easier and business more productive.
A very few musicians have made big bucks because if you wanted the music you had to buy the single, the 8-track, the record, or the CD. And since the RIAA moved these in the millions some artists got rich. Most don't, but a few did. So that has become the definition of success.
Those days are over and musicians need to look at making music as something they want to do for the love of music, not chasing the almighty buck. Sure, they might do well just like there's an occasional Bill Gates in the software industry that makes some exceptional money. But the idea that musicians are automatically entitled to some above-average income level just because they are musicians has got to change. Also, the idea that we should feel sorry for artists that make millions of dollars a year because they've lost 20% of their income to P2P is rediculous. I'll feel sorry for the guy serving up my Big Mac before I feel sorry for any losses at the RIAA, or their artists.
There is a building backlash against these players/owners coming, it will take a few years but what will happen is there will be much fewer fans of the games because the fans simply can't afford to watch them play anymore.
Hmm, that'll be interesting to see if it happens. I agree that sports players make an unreasonable amount of money, but I don't think fans are going to just say "Hey, that's too much money. I'm not going to watch the game anymore."
Rather, fans will stop buying tickets and merchandise when the price of both goes too high. And I doubt that would be a sudden rebellion, but as the price of tickets goes up, the number of sales go down. I'm sure every time they increase ticket/merchandise prices they monitor sales very accurately to make sure they are selling their products at the optimum price to maximize revenue. And that maximized revenue is what makes ungodly salaries possible for players to have fun playing a game.
I went to a Yankees game about a week and a half ago and probably half the seats were empty. I also was thinking of buying a Yankees shirt but it cost US$70. No thanks, I don't need a shirt that bad. Especially when I could get it in Chinatown for $10.:) But there must be enough people that pay US$70 for an official Yankee shirt to make it worthwhile to drive a large number of potential buyers to Chinatown.
Slashdot Mirror
Ok, then you are forgiven because you had data you based your comment on. :) But I can tell you first-hand that NeoIT is wrong. Although I earn American rates and work primarily with Americn companies I've lived in Mexico for the last 8 years. The only software developers that earn $5150 a year would be those maybe right out of college or finishing college, or opting to work at a very low-paying firm. $10,000/year would be much more believable and $15,000/year would not be surprising.
Even when those multiple articles prove with undisputable, acknowledged football statistics that what Rush said was right. Even other sportswriters are confessing it to be true. But don't let reality or facts get in the way of your preconceived notions and stereotypes. Hold your banner up high and shine a light on it for all to see, that way we'll all know who to snicker at. :)
Simply do a graph. If we plot the current minimum wage and number of people employed on the graph, then set an arbitrary minimum wage of $100/hr and the number of people employed we have our two extremes. There is a curve on the graph that connects those and, yes, increasing the minimum wage costs jobs. Otherwise we would just pass a $1000/hr minimum wage law and we could all retire after 2 or 3 years.
You can't legislate prosperity. The only thing a minimum wage can and should do is protect the absolute low-end of the workforce from gross exploitation. It's to avoid sweatshops. It is NOT to make sure that an employee at McDonald's can afford a car, a 27" color TV, cable, and still have money left over for a vacation in Cancun. Those that strive for such goods will have to educate themselves, work hard, get ahead, and start leapfrog into better jobs. The fact that McDonald's doesn't provide a sufficient income to live as well as most people want should be MOTIVATION for these people to better themselves.
I can't imagine India outsourcing to Mexico. I've seen two projects firsthand in MEXICO outsourced to India, and Mexicans DO earn more than Indians. So I'd be surprised to see India outsource to Mexico.
You're 100% right. I have seen overseas outsourcing in action twice. Both times it was a *MEXICAN* company outsourcing to India. Both times the project just dragged on, drained money, and honestly didn't produce results. In one case, what was supposed to be a $500k contract in 6 months dragged on for a year and a half and they doled out $2 million as the Indian company kept asking for more and more time, more and more money, etc.
Once you sink $500k into a project it's hard to write it off as a loss--so they just kept spending more and more money on it. Eventually, they just cancelled the Indian outsourcing, brought it back home to Mexico--and even had me (an American) on board as a consultant to help them!
Andy Grove says that all the telecomm infrastructure makes it as if a guy in India was in the next cubicle. That's the simplistic promise of outsourcing that tempts companies to do it. There may be some industries where it will work. Software development is NOT one of them.
I'm a software engineer and I'm NOT worried about outsourcing at all. Yes, we'll have some short-term pain as many companies experiment with it. But the two cases I've seen of MEXICAN companies trying to save a buck by outsourcing to India made it very clear that it just won't work long-term. I completely believe many jobs will be outsourced to India and China, but many (if not most) will bounce back to the U.S. after the 1-2 years it takes to realize it has failed.
And if you read enough articles, you'd know that what he said last week was plenty accurate and his assessment was more accurate than many that supposedly "know football." But keep carrying that liberal banner way up high if it makes you feel better about yourself. Just don't be surprised when others give you about the same credibility as we give Rob Enderle.
Probably about the same way you explain TCP/IP to the average home user who just wants to read e-mail and surf the web. You don't. That doesn't mean it can't be of use to the user even if he or she doesn't understand it--or probably even knows it exists.
Bingo! I love it when people say that spammers are successfully getting around Bayesian filters. Either those Bayesian filters were poorly implemented or the people saying it just have something against Bayesian filtering and don't really know what they're talking about.
The percentage of spam being caught by my Bayesian filter continues to climb (currently around 99.7%, just like Graham's) even in the face of random words and inserted "neutral text." You have to break Bayesian pretty bad for any of those tricks to actually get a piece of spam through the filter.
Unless you earn only pennies per hour, "just hitting the delete key" is the most expensive part of spam. Like I said in my original message, I agree that it would be better if it weren't sent at all. But the largest cost to me--and to SOCIETY--is the time it takes to DEAL with spam, not the bandwidth--although I don't deny the bandwidth is a cost... just not as significant as my time.
and worse, now you're spinning extra cycles to scan the mail.
Better my CPU consume some cycles than spam consume my time. My CPU is cheaper than my time.
Just hit delete means you kill 1000 this month -- and 10000 a year later.
Again, re-read my above message. In the short-term, yes. But as more and more people use filters fewer people will receive spam and even fewer will respond to it. Over time spam becomes less lucrative so there is less reason to send it.
It seems to me like you are looking for an overnight fix. There isn't one. But effective and pervasive spam filtering is a solution that keeps me from having to deal with spam on a daily basis in the short-term, and will lead to a reduction of spam when it no longer becomes profitable to send spam since everyone filters it.
Oh, I misunderstood. But even so, the filter decoder doesn't have to do EVERYTHING the GUI decoder does. All it has to do is decode and parse, it doesn't have to make a visual presentation.
For example, it is sufficient to be able to decode MIME messages (including mutli-part), handle quoted-printable, handle Base64, and be able to decode HTML and quoted-printable escape sequences. You then just throw out all HTML comments, only keep HTML commands you are interested in (FORMs, IMGs, HREFs, FRAMEs, FONTs), throw out multiple whitespace, and then preferably parse out all remaining content into either headers, text, HTML, or script. You can then filter base on each of those content areas as you like.
What I mean is you certainly don't have to be able to execute Javascript or anything complicated like that. And short of anything executable (which would be dangerous anyway) the rest isn't really that hard.
In fact, one thing Paul Graham didn't address in his latest pieces is the risk of following links that can have embedded identifying information in them. So, yes, you slam their server but you may have inadvertently just told them WHO you are by requesting a certain URL.
I've never seen a filter that works in that fashion. Even if the filter is middleware (which is common), it must have an integrated message decoder. If it doesn't it is useless. The message needs to be fully decoded during spam filtering and then, probably, again when it is displayed to the user.
Great! I'll tell you, it almost makes spam fun for me. I look at my statistics and just grin at all the garbage my filter kept out of my life. "Let's see, how many spams did my filter block THIS month?" :)
However, that's why email programs and ISPs should have Bayesian spam filtering. In email programs, it should be enabled by default,
I agree 100%. I personally think they should be enabled by default both in email programs and by ISPs--of course with the option of disabling them. But the idiots won't bother to disable them so they'll be protected, and those that don't want the ISP meddling with filtering can disable it.
I agree, the spam volume won't go down until everyone--especially the idiots--start using Bayesian, whether they make the effort to do so (doubtful) or whether they simply don't make the effort to disable it (probable).
But still, I must say... as much as I'd like to see spam stop wasting bandwidth, the most expensive aspect of spam is the time it wastes for the receiver. And Bayesian fixes that for us whether the spam is still being sent or not.
That wouldn't be a Bayesian filter, it'd be a simple checksum filter. And the reason no-one does it is because usually every spam sent out is slightly different. Spammers insert random garbage words, etc. which would make a simple checksum different for every spam.
So far analyzing the IMG of spams is unnecessary. The HTML content alone used to display even a single image is usually sufficient for Bayesian to conclude it's spam. There's no reason to make things more complicated than they need be by analyzing IMGs, etc. The evidence is right there in the spam they send you.
How early? I haven't seen that.
Remember the images instead of text, then the HTML Entities (like instead of the letter 'A')? The second and third generations of the Bayesian filters had to account for them.
ALL filters, Bayesian or otherwise, have to be able to properly interpret and translate an email. It is insane to think you can filter email without being able to properly decode it. This is a limitation of the message decoder, not Bayesian filtering.
I can just see how a DoS filter would be circumvented early: redirects and browser scripts.
As Paul Graham said in an earlier piece (although surpirisngly not in this one), if your goal is to detect spam and an email link takes you to a redirect or browser script that's a high probability of spam. If you want to pound the spammer as suggested in this latest article, follow the redirect. Yes, you will also pound the first site but perhaps they'll notice it and take down the redirect altogether. Either way the end result is that gullible people that would otherwise follow the link to purchase from the spammer won't be able to get to the target site. It will either be Slashdotted or the redirect will be removed completely. In either case the end goal is achieved.
But, up until the last 6 or 7 years, it has generally been more hassle and time-consuming to copy the works rather than just buy them. That is no longer the case and changes the situation significantly.
If, while I'm working, I can download a song I want in the background and have it in 5 or 10 minutes for free or stop working and go to the mall and pickup a FREE CD, I'd rather just download the song. That saves me the hassle of going to the store, picking it up (even if it's free), and ripping it. And if I can do it without identifying myself or sharing my credit card number, all the better!
especially given that people have *always* been happy to pay for the priviledge of being able to reproduce music through speakers when ever and where ever they are.
And kings and queens used to travel their countries in horse-drawn carriages. That doesn't mean you'll see kings and queens today thrilled to pay to use obsolete technology. Just because something has always been so doesn't mean it always will.
I'm afraid a generation of teenagers has grown up through their most music-buying years in the digital era. My sister-in-law started downloading music when she was 18. She's now 21. I think she has something like 4000 MP3s and she has a MODEM. I don't think she's purchased a single CD in those 3 years.
Times are changing. They always have and always will.
but I am confident that people in developed nations that have the money dont mind paying for anything when they think they're paying a fair price. Thats the very defition of fair.
I agree with you. But the question is, "What is fair?" I'll be blunt--most of the music out there isn't worth a penny. It's not even worth the drive to the store or the effort to whip out my credit card.
That said, there's a lot of music out there that I wouldn't be willing to buy but I *would* be willing to pay $20 to see in concert.
Yes and no. Can reality be something other than I can imagine? Sure. But can an educated person take existing knowledge and formulate a theory that is more likely right than wrong? Yes.
These guys are seeing soccer balls because they want to see soccer balls. There is no reason, at this time, for us to believe that if we travel in a certain direction on the universe that we'll end up where we started. There is no observational evidence to suggest that. The article itself indicates that data to that effect is completely lacking, and no-one has been able to detect the "repeat patterns" that you would see if we were seeing a small universe from lots of different angles.
Now, if we traveled in some direction and found ourselves back where we started then certainly we would want to analyze why that happened.
On what evidence is your belief that the universe is infinite based? Why does it trump these guys' evidence?
Read the article. The article itself suggests problems in their conclusion and indicates that an infinite universe could produce the same data they've acquired.
So... we have the possibility that the universe is pretty much as we expect it and the data collected does not necessarily contradict that. Or we can conclude the universe is a soccer ball and that we'll get back to where we started if we go off in a certain direction long enough.
By the same logic, we can conclude there is a wormhole over the Pacific at the International dateline. The world is flat but when you go past 180 degrees we travel through a warp that takes us back to the other side of the flat world.
Yeah, maybe that's possible, but it's more likely the world is just round.
They are only winning to those that don't do anything to help themselves.
The Verisign SiteFinder was a bad thing, obviously, but I laughed at the reaction "It's breaking my spam filter." What kind of archaeic, obsolete spam filter were these people using?
Likewise, that spammers are using trojaned systems is bad, of course. Any system compromise is bad. But this is just normal virus and hacking. It doesn't make it any harder to get rid of your spam.
I've said it once and I'll say it again, Bayesian filers is the solution. It works today and it depends on no-one but yourself to start using it. Since I started using it in May, I've received 20,596 spams--of those I've seen 89 of them. I.e., only 0.43%. It comes out to one spam every other day, though that's deceptive since probably half of those that got by were cases of a single spam sent 5 times in rapid-fire mode and they all happened to get through at once--the same spam 6 hours later would've been filtered. In reality, I'd guess I see one spam per week. In a perefect world I wouldn't see any, but that's good enough for me in this imperfect world.
Now, some will say "But that doesn't solve the bandwidth problem." In the short-term, no, it doesn't. But in the short-term it doesn't waste my time which is my single largest expense when it comes to spam. And, in the long-term, if more people started using Bayesian the response rate on spam would continue to plummet making it less and less useful to spam in the first place.
But those that are being bothered by spam on a daily basis simply aren't using the tools and technology that are available to them, and have been for over a year.
Anyway, the point is that they paid an absurd amout of money for a site that virtually anyone could have done in less than a month. And if the person had experience and was efficient I agree with you that we'd be talking about hours or a few days rather than weeks or months.
But which really makes more sense? That the universe is a giant void filled with matter in various areas? Or that some weird mechanics cause the universe to actually be small and traveling in one direction brings us back to where we started as if we were traveling on the surface of a globe.
Personally, I think the "universe" is infinite. That's not to say that there are galaxies and stars filling every space of the infinite area, but I think if you are at any place in the universe you can always continue further in any direction even if you don't find anything there but the absence of matter. I don't think you're going to hit a wall where you can't go further nor reach a point that "wraps you around" to the the other side of the universe.
Perhaps the definition of "universe" is the "outer" extent of where matter can be found. In this sense, the universe seems to be expanding--but it's not that the universe is really getting bigger, just that existing material is moving further and further out into the available void which is infinite. My opinion only, of course.
I won't say these guys are wrong, but it just seems that what they are suggesting is probably more complicated than it needs to be or probably is. But at least they got their 15 minutes of fame and were able to mention "soccer" and "science" in the same sentence.
Maybe, but I don't think so. P2P will win because it's anonymous, easy, and free. You don't have to give up your name or credit card number, you don't have to logon to some website, you don't have to deal with some non-standard music file format. Just power up the app and download the song you want and you're golden.
The whole "pay for Internet music services" is based on the premise that CDs may be obsolete but the concept of paying for recorded music is still valid. I believe that that whole concept is obsolete. Music is free. Concerts, merchandise and endorsements is where the money will be for those musicians that want to go that route. But music itself is free.
Any product or service that fails to understand that will fail. That's why there's, what, 4 million users on Kazaa at any given time and yet iTunes is thrilled because they sold a million songs in a couple weeks? Kazaa probably does that in a few hours.
Legal or not, people have tasted free music and it tasted good--especially when the RIAA was charging $15-$20 for the same thing at the time. I don't think there's any way to sugar-coat the fact that you're asking them to now pay for something they've grown accustomed to getting for free, especially younger ones that just like downloading MP3s because they like to collect them (whether or not they like the music) and also since they don't have much money to start with.
The RIAA should've introduced a pay-for-download music service 6 months before Napster came out. Now, it's too late.
I would hope that "building" the site for $3.5 million also includes running it, ongoing maintenance, etc. Because if the government really paid AT&T $3.5 million to BUILD it and still has to pay some ongoing fee, they got ripped by an order or two of magnitude.
Government waste isn't surprising, but it's sad when it is made so obvious. A good percentage of the folks here at Slashdot could have done just as good a job for a fraction of the cost and STILL recorded a very good year income-wise.
If people are willing to pay for your music without you performing and you are able to make a lower, middle, or even upper class living, great! That's the free market. What I've been saying for years is that the free market is slowly reaching the conclusion that it is NOT willing to pay just for music. And, unfortunately, that's not even your fault. It's the RIAA. They've robbed so many customers that now customers are just not willing to pay for music anymore, period.
So it's not a proposterous notion that you want to make money doing what you love. You are mistaken, however, if you think that the simple fact that people are listening to your music entitles you to that money. The free market will dictate what you earn, and no-one is entitled to anything more than what the free market is willing to pay them.
It is the loud minority that cause posters like you to suggest that musicians are just looking for a Get Rich Quick
I apologize if that's what my post caused you to believe. I know a number of musicians and I know they aren't trying to get rich quick. I was really referring to the RIAA-type musicians, and those that aspire to be RIAA-type musicians. My mistake.
If I'm at a computer and NOT typing, I'm falling asleep. Honestly. I've had people try to explain their code to me when they left a company and I literally had to try to stay awake, and that might have only been for a couple of hours. If I had to watch someone else work for hours on end I would be bored and want to do some real work--if I was able to stay awake.
Theoretical programming styles don't interest me much, and probably each applies to certain projects better than others. But I definitely would not work for a company that wants me to sit around and watch someone else work, nor would I employ someone to sit around and watch someone else work.
I agree completely. Is there some reason why musicians think they should be rolling in the big bucks? When a hamburger flipper earns $15k/year, some entry level positions $30k, and software engineers with decent experience make $50k, is there something particularly special about a musician that entitles him to earn $200k, $500k, a few million per year? Yes, a musician has special talents that allow him to compose or perform music that others enjoy. So? Software engineers have special talents that allow him to develop software that hopefully make the lives easier and business more productive.
A very few musicians have made big bucks because if you wanted the music you had to buy the single, the 8-track, the record, or the CD. And since the RIAA moved these in the millions some artists got rich. Most don't, but a few did. So that has become the definition of success.
Those days are over and musicians need to look at making music as something they want to do for the love of music, not chasing the almighty buck. Sure, they might do well just like there's an occasional Bill Gates in the software industry that makes some exceptional money. But the idea that musicians are automatically entitled to some above-average income level just because they are musicians has got to change. Also, the idea that we should feel sorry for artists that make millions of dollars a year because they've lost 20% of their income to P2P is rediculous. I'll feel sorry for the guy serving up my Big Mac before I feel sorry for any losses at the RIAA, or their artists.
There is a building backlash against these players/owners coming, it will take a few years but what will happen is there will be much fewer fans of the games because the fans simply can't afford to watch them play anymore.
Hmm, that'll be interesting to see if it happens. I agree that sports players make an unreasonable amount of money, but I don't think fans are going to just say "Hey, that's too much money. I'm not going to watch the game anymore."
Rather, fans will stop buying tickets and merchandise when the price of both goes too high. And I doubt that would be a sudden rebellion, but as the price of tickets goes up, the number of sales go down. I'm sure every time they increase ticket/merchandise prices they monitor sales very accurately to make sure they are selling their products at the optimum price to maximize revenue. And that maximized revenue is what makes ungodly salaries possible for players to have fun playing a game.
I went to a Yankees game about a week and a half ago and probably half the seats were empty. I also was thinking of buying a Yankees shirt but it cost US$70. No thanks, I don't need a shirt that bad. Especially when I could get it in Chinatown for $10. :) But there must be enough people that pay US$70 for an official Yankee shirt to make it worthwhile to drive a large number of potential buyers to Chinatown.