But honestly, why would anyone carry two devices (mobile phone, organizer) when the Sidekick (or another similar device) offers the functionality of both?
I can tell you why I carry two devices rather than one: I'm terrified of losing my contact and calendar information. I live and breathe, during the working week, with my palm nearby. I've had my PDA break on me in the past, and it's a near disaster until I can get it replaced.
Cellphones and multifunction devices are generally pretty fragile and given the amount of use they get, don't have great life spands. For me, if I had a combo device and got bumped on the subway or what have you, leading to a broken device, I'd not just be without my phone as without my appointment book, address book, essential files etc.
I carry two devices in order to make sure I only lose either the phone or the contact list, but not both.
Check out Red Flag Deals for a nice selection of rebates, coupons and specials from Canadian dealers. The site covers all sorts of merchandise, but seems to be predominantly geared towards geek-analia.
Interesting, they apparently are going after a Canadian this round:
whois 132.206.101.25
Country: CANADA
NOTE: More information appears to be available at MU-ORG-ARIN.
OrgName: McGill University OrgID: GILL Address: Department of Electrical Engineering Address: 3480 University Street City: Montreal StateProv: PostalCode: Quebec Country: CA
Oh, don't misunderstand me. I wasn't saying he shouldn't throttle back on his network usage. I'm pretty parsimonious at home myself, and think that if he's 100 times above the national median then he's probably hurting his neighbours and abusing the system.
My point was that a letter stating "100 times (!!!) the National (!!!) median (whisper)" sounds shocking yet it is more likely that he's only 10 times above the average. That's still too high, but it's not like he's running slashdot from home.
But don't expect them to beleive this is nothing illegal, and don't expect them to allow you to pay $49.95 or whatever for 100 times the average. I'm sure if you were 2 or 3 probably even 10 times the average, you'd be ok...
Careful there with the confusion of median and average. Given typical 'net usage patterns, the highly skewed distribution of bandwidth usage will lead to different values for the median and average. Essentially, I interpret the letter to mean that most of the people in the nation use practially no bandwidth while he is actually daring to use the fat pipe he's paid for to its potential.
Here's a little thought experiment: Say there are 101 total cable subscribers. 55 of those people only use 2 GB of bandwidth per month, 20 use 5 GB per month, 15 use 20 GB per month and the remaining 10 use 50 GB per month. Our poor lead poster is using up 100 GB per month. The median value for bandwidth consumption would be 2 GB while the average would be ~11 GB per month. As such, our user is 50 times above the median and only 9 times above the average.
If I were in his shoes, I'd ask what the average bandwidth usage is, and compare on that basis (it makes things more favourable for him). Or, ask for the quintiles or some such. IMHO the letter was well crafted to scare the bejesus out of him by playing with the stats.
We should focus instead on content-based spam filtering, and share that knowledge to improve efficiency. Accuracy skyrockets and collateral damage virtually disappears!
This is a typical response given by many people who don't understand that blacklists are needed since the spam problem scales. For the individual, content-based filtering is often the best way to go; it's easier and the technology has improved to the point of being very, very accurate.
However, this is little consolation to those mail admins responsible for hundreds, if not thousands, of users. When each of them is getting hundreds of spam messages daily the bandwidth costs become significant. When the users start to demand that you do the content-filtering on the server side (which is a reasonable request), you're talking about serious computing power, which entrains even more cost. And finally, when terra.es mailbombs your users with bestiality and pedophile porn adverts, you lose clients and risk being sued (these are litigious times). This all costs you money.
Now, if you're the mail admin you have the choice of either dealing with these problems daily and spending money to deal with the spew of scumbag lowlives, or you can block the nastiest parts of the net at the router level. If the anecdotal evidence from most mail admins is correct, when they blackhole terra.es, they DO NOT lose legitimate email. When.kr is blocked by a North American business, in most cases NO legitimate email is lost.
And finally, here's the thing that really gets my goat. If you are trying to contact someone for either personal or business reasons and your email gets bounced back because you are in a spam-infested part of the net, all is not lost. Your life has not come to an end. Chickens are not falling from the sky. Waht you should do is the following two things, immediately:
1. CALL YOUR CONTACT ON THE PHONE!!! Get yourself whitelisted. E-mail is notoriously unreliable, so for anything mission critical you should always be confirming receipt and get a response; if that doesn't happen, get on the freaking phone.
2. Move out of your slum; it's hard to be taken seriously as a business when you are connected and dependent on UUNet, for example, a corrupt criminal company which has regularly polluted the entire net with the crap from their customers.
Actually, I think it's worse than that. Pattern recognition, cluster analysis etc. are good ways to assign an individual to a group with a certain degree of confidence. Your degree of confidence in that assignment is related to the alpha and beta error rates you're willing to accept. So, the decision to place someone in the 'terrorist' or 'security threat' cluster is partially dependent on the false-positive and false-negative error rate you're willing to accept.
Now, consider the situation at hand. You (the company) are tracking and classifying people based on a number of variables (and ethnic origin couldn't possibly be one of those, could it?). False positives include placing a non-terrorist in the terrorist group. False negatives include not identifying a security risk. In 2003, which error rate do you think is going to be minimized? Knowing that there is an inverse relationship between the two, the false positive rate is going to be very high. This means that perfectly normal individuals (who happen to be of the 'wrong' ethnic group) are going to be subject to political persecution.
This statistical approach is useful in painting with a broad brush, but shouldn't be used for something as sensitive as the rights of individuals. The risk for errors/abuse is simply too great.
Actually, you can use this approach to another end as well. If you want to buy something from indigo books, for example, create and use the email address indigobooks@yourdomain.com for the transaction. If you later find that you're getting spam to that address, you have a good idea of who sold/leaked your email. Great way to build up and work a bitch-list of slimy companies.
I would have to second this suggestion. If the original poster isn't going to be doing complicated statistical analysis or curve-fitting (I mean, is he just ramming a straight line through his data cloud, or is he into splines and whatnot), then it probably isn't worth his time to learn the syntax. If, on the other hand, he's going to be doing some real stats, then he should definitely investigate R as a solution.
R is multi-platform, capable of producing graphs in a number of formats (eps, ps, pdf, png etc), can be run in batch mode (great for simulations) and can produce very pretty graphics. There is no gui, so all fine tuning of the figures has to be done through the command line. However, it can handle immense data sets, supports mathematical text (meaning you can throw a complicated equation into your graph and have it come out with publication-quailty finish) and can do some colour, three-dimensional plotting. A wonderful piece of software that has been indispensible during my Ph. D. research.
My old master's supervisor stated your point far more succinctly during a conference we were attending. Listening to a particular scientist's presentation, he leaned over and whispered to me:
"That's what we call a blinding glimpse of the perfectly obvious."
It's from Edmund Burke, an english politician/philosopher. However, the exact text of the quote is in some doubt, and there appear to be many variants of this phrase on the 'net.
Slashdot Mirror
But honestly, why would anyone carry two devices (mobile phone, organizer) when the Sidekick (or another similar device) offers the functionality of both?
I can tell you why I carry two devices rather than one: I'm terrified of losing my contact and calendar information. I live and breathe, during the working week, with my palm nearby. I've had my PDA break on me in the past, and it's a near disaster until I can get it replaced.
Cellphones and multifunction devices are generally pretty fragile and given the amount of use they get, don't have great life spands. For me, if I had a combo device and got bumped on the subway or what have you, leading to a broken device, I'd not just be without my phone as without my appointment book, address book, essential files etc.
I carry two devices in order to make sure I only lose either the phone or the contact list, but not both.
Check out Red Flag Deals for a nice selection of rebates, coupons and specials from Canadian dealers. The site covers all sorts of merchandise, but seems to be predominantly geared towards geek-analia.
whois 132.206.101.25
Oh, don't misunderstand me. I wasn't saying he shouldn't throttle back on his network usage. I'm pretty parsimonious at home myself, and think that if he's 100 times above the national median then he's probably hurting his neighbours and abusing the system.
My point was that a letter stating "100 times (!!!) the National (!!!) median (whisper)" sounds shocking yet it is more likely that he's only 10 times above the average. That's still too high, but it's not like he's running slashdot from home.
Careful there with the confusion of median and average. Given typical 'net usage patterns, the highly skewed distribution of bandwidth usage will lead to different values for the median and average. Essentially, I interpret the letter to mean that most of the people in the nation use practially no bandwidth while he is actually daring to use the fat pipe he's paid for to its potential.
Here's a little thought experiment: Say there are 101 total cable subscribers. 55 of those people only use 2 GB of bandwidth per month, 20 use 5 GB per month, 15 use 20 GB per month and the remaining 10 use 50 GB per month. Our poor lead poster is using up 100 GB per month. The median value for bandwidth consumption would be 2 GB while the average would be ~11 GB per month. As such, our user is 50 times above the median and only 9 times above the average.
If I were in his shoes, I'd ask what the average bandwidth usage is, and compare on that basis (it makes things more favourable for him). Or, ask for the quintiles or some such. IMHO the letter was well crafted to scare the bejesus out of him by playing with the stats.
This is a typical response given by many people who don't understand that blacklists are needed since the spam problem scales. For the individual, content-based filtering is often the best way to go; it's easier and the technology has improved to the point of being very, very accurate.
However, this is little consolation to those mail admins responsible for hundreds, if not thousands, of users. When each of them is getting hundreds of spam messages daily the bandwidth costs become significant. When the users start to demand that you do the content-filtering on the server side (which is a reasonable request), you're talking about serious computing power, which entrains even more cost. And finally, when terra.es mailbombs your users with bestiality and pedophile porn adverts, you lose clients and risk being sued (these are litigious times). This all costs you money.
Now, if you're the mail admin you have the choice of either dealing with these problems daily and spending money to deal with the spew of scumbag lowlives, or you can block the nastiest parts of the net at the router level. If the anecdotal evidence from most mail admins is correct, when they blackhole terra.es, they DO NOT lose legitimate email. When .kr is blocked by a North American business, in most cases NO legitimate email is lost.
And finally, here's the thing that really gets my goat. If you are trying to contact someone for either personal or business reasons and your email gets bounced back because you are in a spam-infested part of the net, all is not lost. Your life has not come to an end. Chickens are not falling from the sky. Waht you should do is the following two things, immediately:
1. CALL YOUR CONTACT ON THE PHONE!!! Get yourself whitelisted. E-mail is notoriously unreliable, so for anything mission critical you should always be confirming receipt and get a response; if that doesn't happen, get on the freaking phone.
2. Move out of your slum; it's hard to be taken seriously as a business when you are connected and dependent on UUNet, for example, a corrupt criminal company which has regularly polluted the entire net with the crap from their customers.
Actually, I think it's worse than that. Pattern recognition, cluster analysis etc. are good ways to assign an individual to a group with a certain degree of confidence. Your degree of confidence in that assignment is related to the alpha and beta error rates you're willing to accept. So, the decision to place someone in the 'terrorist' or 'security threat' cluster is partially dependent on the false-positive and false-negative error rate you're willing to accept.
Now, consider the situation at hand. You (the company) are tracking and classifying people based on a number of variables (and ethnic origin couldn't possibly be one of those, could it?). False positives include placing a non-terrorist in the terrorist group. False negatives include not identifying a security risk. In 2003, which error rate do you think is going to be minimized? Knowing that there is an inverse relationship between the two, the false positive rate is going to be very high. This means that perfectly normal individuals (who happen to be of the 'wrong' ethnic group) are going to be subject to political persecution.
This statistical approach is useful in painting with a broad brush, but shouldn't be used for something as sensitive as the rights of individuals. The risk for errors/abuse is simply too great.
Actually, you can use this approach to another end as well. If you want to buy something from indigo books, for example, create and use the email address indigobooks@yourdomain.com for the transaction. If you later find that you're getting spam to that address, you have a good idea of who sold/leaked your email. Great way to build up and work a bitch-list of slimy companies.
I would have to second this suggestion. If the original poster isn't going to be doing complicated statistical analysis or curve-fitting (I mean, is he just ramming a straight line through his data cloud, or is he into splines and whatnot), then it probably isn't worth his time to learn the syntax. If, on the other hand, he's going to be doing some real stats, then he should definitely investigate R as a solution.
R is multi-platform, capable of producing graphs in a number of formats (eps, ps, pdf, png etc), can be run in batch mode (great for simulations) and can produce very pretty graphics. There is no gui, so all fine tuning of the figures has to be done through the command line. However, it can handle immense data sets, supports mathematical text (meaning you can throw a complicated equation into your graph and have it come out with publication-quailty finish) and can do some colour, three-dimensional plotting. A wonderful piece of software that has been indispensible during my Ph. D. research.
Worked fine under Konqueror (3.0.2) using the standard flash plugin.
My old master's supervisor stated your point far more succinctly during a conference we were attending. Listening to a particular scientist's presentation, he leaned over and whispered to me:
"That's what we call a blinding glimpse of the perfectly obvious."
It's from Edmund Burke, an english politician/philosopher. However, the exact text of the quote is in some doubt, and there appear to be many variants of this phrase on the 'net.