The reason it's fast is that _everybody_ who is using BitTorrent to download a file is also contributing their upload resources to the community of people who want to download that file, and if people leave BitTorrent running after they've finished downlading they're also contributing. This means that you're not limited by the bandwidth of a single server like you are with FTP/HTTP, and you're not limited by the bandwidth of a single peer like you are with Napster, and because you're downloading from multiple peers, you can take advantage of the fastest peers as opposed to just the average or slowest peers.
There are times that single-stream downloads can be faster - if you're downloading from a fast lightly loaded server that has more bandwidth per user than your Internet connection, that's as fast as it gets, while if you're running BitTorrent on an asymmetric connection like ADSL or cable modem and the community of people who want the same file is still mostly downloading, then your download speed ends up limited by your upstream bandwidth until lots more people have the file. But it's pretty common for BitTorrent users to leave their clients uploading after they've finished downloading, particularly for big downloads that run overnight (because if it finishes before you get up in the morning, it keeps running.) There are exceptions - not just greedy leeching downloaders, but also people who download when the community has a lot of spare capacity and low demand, but that's when it's really not a problem.
Because BitTorrent doesn't need a fast server to support a lot of downloaders, there's also more material that can be published. If you're running your own tracker for the material you're publishing, that does take some bandwidth, but it's a lot less than actually downloading the file to lots of people.
I agree that a polished Hollywood flick would be the wrong way to do that movie, but the basic problem is that nobody takes the quantities of acid that were obviously involved in making the first one...
I haven't seen the trailer, and have no plans to see the movie, but if you _wanted_ to remake Willy Wonka for some bizarre reason, Depp is almost the only person I could think of to star in it. You could take it in somewhat different directions (e.g. Jim Carey or a younger Tim Curry) but Johnny Depp is really the right guy.
Sure, there are people who don't like the look of wind farms, because you can see the things, just like there are people who don't like seeing cell phone antennas. I've driven by the Altamont Pass wind farms fairly often, and once looked into renting a house that was located out there, and ok, it was a bit spacey, and if you've got epilepsy it might not be where you want to live, but the wind turbines are nowhere near as ugly as a smokestack or a coal strip-mine.
For many spammers, the "complain to their ISP or their upstream" method can work - for small spammers, it's a whack-a-mole game, but it did help take down Scott Richter's OptInRealBig network for a while, since their ISP shell had to buy bandwidth from actual larger ISPs.
The Artists Against 419 bandwidth sucker has taken out a couple hundred spammer websites from the Nigerian 419 crowd. I'm not running it today - the new NetBSD release came out so I'm wasting my bandwidth running Bittorrent instead (and there seem to be lots of high-bandwidth people seeding the torrent, so I've been downloading at 1.5 Mbps all morning.)
Active cracker DDOSing is mean and nasty and you shouldn't do it. But there are better-behaved ways to use group efforts to stop spammers.
Blocklists are of course a critical tool - identify the spammers or the relays/proxies/zombies they exploit, publish their addresses so that people can reject mail from them.
Sugarplums and other spam poisoners generate web pages full of bogus trap addresses for spammer address harvesters, so that they can DDOS themselves. Infinite-loop web pages, bogus email addresses, email addresses of other spammers, email addresses of teergrubes, spambait addresses on your machines that tell you to block anything from that IP address. Imagine if everybody set your 404-not-found page to include a few bogus addresses for spammers to email to...
Teergruben are modified tarpit mail servers that answer SMTP v...errrrryyyyyyyy... sssssssllllloooooooowwwwwwwlllllllly, and can keep SMTP senders that talk to them tied up for minutes or hours. If you're running real SMTP on the same machine, you can configure the tarpit function to only happen for recognized spammer IP addresses, or else you can run a dedicated server (e.g. if you're not running your own SMTP on your DSL or cable modem.) One of these doesn't make much difference. Lots of teergrubes can tie up lots of spammers.
Bandwidth Suckers like Artists Against 419 repeatedly download images from spammer websites to tie up their bandwidth. Because many web sites and ISPs charge for bandwidth on a 95th percentile basis, two days of heavy downloads can totally jack their bandwidth bill for a month, and small sites (e.g. free web pages) that have quotas can be taken out for the month by aggressive downloads (1GB is about 6 hours at 384kbps, so you can blow out a small quota overnight.)
I first met Dave Farber in ~1992. I was taking the train from DC back to New Jersey, and there was this old guy sitting across the table from me with the smallest laptop I'd ever seen, a cellphone smaller than a brick (Moto flip-phones were still amazing back then), and an alphanumeric Skypager. We started talking, and I recognized his name from various sources - his IP list was much smaller back then, and he'd been one of the advisors to the recently-formed EFF, and I was doing cypherpunk stuff with some of the same people. The laptop was an IBM Japan model, with Japanese keyboard, about six pounds, and it was quite a while before anything that light was available here. (And yes, it was a black rectangle design like most of the newer Thinkpads, but black was still cool back then.) He'd been toy-shopping at Akihabara.
Vince Cate, the author of this article, lives in Anguilla. He's a well-known ex-USAn cypherpunk who moved down there a number of years ago, and runs offshore hosting services and such. Anguilla's not a big place, maybe 10000 people, and while it's not rich, it's also not dirt-poor, and some things are expensive because of ridiculous import taxes (like many third-world countries), or because of inherent costs (electricity there is expensive and doesn't sound highly reliable.)
These machines aren't low price because some company is subsidizing them as a charity for poor people - they're low price because they've been designed for a lower cost, leaving behind some of the excess parts of the legacy PC design for something that's not fully PC-compatible and running a lower-priced OS. They're also running a less bloated OS than WinXP, so they can get by with a slower CPU and less RAM and disk, though they could of course have done that with Linux or ucLinux.
They're intended for markets where people can't pay as much for them, and there's probably less profit per box in them than in higher-end machines - but as other posters have pointed out, if rich people buy them, AMD can just crank out more of these, because they're Making A Profit on each one (or at least, they're making a Marginal Profit on each one; total profit depends on whether they sell enough of them to cover their initial development costs.) If rich people turn these things into X-terminals for their kids or whatever, that's just more sales, which is a Good Thing.
This isn't a charity project - this is a system designed to cost less because it's got less legacy clutter on it, plus it can use less memory / disk / CPU because it's running a less bloated operating system. They're still paying for WinCE, though the PDA market price points means that they probably pay less for WinCE than for WinXP, and Linux would still be a bit cheaper.
SPF will help with that, to the extent that you're getting joejob mail from sites that use SPF. To the people who get forged, yes, your C/R mail feels about as spammy as the anti-virus notices they're also getting.
But the real problem isn't just whether you mind sending challenges to unknown people who send you legitimate email - it's whether _they_ mind, or whether they ignore your TMDA and their mail to you gets lost. If they're spammers, of course, you don't care, but if they're people you'd be interested in hearing from (e.g. somebody asking a question about your web site), it would be nice not to annoy them.
SPF won't help you cut down much on the 200 messages you sent to real addresses, except for mail that was forged by viruses and maybe some Joe Jobs that forge addresses of people you hadn't already whitelisted. It'll make a big impact on the 3800 TMDA messages you sent to bogus addresses, assuming that many of them are for domains that adopt SPF (e.g. forged Yahoo addresses), because it's designed to do that.
But it isn't designed to block spammers, just forgers, and spammers are already starting to adopt SPF - so you can tell that mail you received really did orginate at bigspammer.com. This means that you can't use "SPF says it's plausible" to mean "it's not spam", so you'll still need to send them your TMDA if you want to prove they're a human.
First of all, nobody needs a second-level domain for these techniques to work - subdomains work quite well, and once you've written a couple of good perl scripts to feed your DNS server, they're free and effortless. Several ISPs I use automatically create subdomains for users, so username@example.com is also anything@username.example.com, which makes it easy to give out tagged addresses.
Secondly, SMTP supports tagged addresses of the form username+tag@example.com, and your email client can filter on the tags. Unfortunately, that's not foolproof - many web forms choke on the "+", because it has syntactic value to CGI and they're not always bright enough to escape the character. (But almost everything can handle subdomain-format tags.) Also unfortunately, Pobox's forwarding service and.forward files and other mail forwarders generally don't know how to preserve the tags while forwarding mail, though sometimes they can at least forward the mail.
The biggest problem with tagged emails is that to use them effectively, your email client needs to keep track of them, so if you get mail addressed to tag-for-alice@username.example.com, your reply to it will come From: tag-for-alice@username.example.com and not From: somedefaultvalue@username.example.com, and if you're sending mail to alice@alice.com, your mail client will know to send mail from tag-for-alice@username.example.com or whatever the last tag was that you used to send it to that address. Also, your email client should keep track of all the addresses you've sent out, because you might want to handle mail from unknown tags differently.
Using MIME attachments for the confirmation is annoying - it increases the chances that the original sender won't be able to read them, when the software could have perfectly well just sent a URL.
Setting "Precedence: Bulk" would seem to discourage reading, but at least it seems to be a common convention that vacation-mailers don't respond to it.
The real problem you've got is that of the 4000 unknown addresses that you received email from, as many as 1000 might have been from real people rather than spammers, but most of them didn't bother replying. It's possible that only 10 of them were real people, so maybe only a couple of the real people who'd sent you email you might have cared about didn't bother replying to your TMDA, but it's also possible that 992 of them didn't, i.e. 99% of the real senders. You can't easily tell, except of course for the undeliverable addresses which were probably forged (or else are on email systems that don't let strangers verify addresses any more because spammers abuse them.)
It's an obvious idea, but unfortunately an inadequate one. Too many spammers, especially phishers, include legitimate URLs to evade Bayesian filters or trick users into thinking they're really getting email from Citibank. Most of the legitimate URLs are big enough it's not a problem (e.g. Citibank probably wouldn't notice the hit), but some are smaller and would suffer. So you've got to check the URLs, probably manually, and only hit the ones that aren't legit. Also, lots of spammer URLs really go to free web sites where they redirect to their real ones using Meta-Refresh or whatever, so the URL you're pounding isn't the real target.
If their screensaver just downloads web pages from the spammer site, they could use something trivial like running wget a lot, and if they just want to fill out spammers' viagra order forms with lots of bogus data, they could get away with netcatting a standard request to them (which is easy if the screensaver downloads the target list from Lycos anyway - just download the target request page.)
The Artists Against 419 page uses a browser with some Javascript or equivalent to repeatedly download web pages from Nigerian 419 scammers. It's much less efficient, because the browser renders all the pages each time instead of just downloading them to/dev/null, and it's not a screensaver, but it's a no-brainer to use.
Sure, it's a whack-a-mole game, and if you only take out two of them, there'll be two more to take their place. It's obviously more effective if you can take out two big ROKSO-known spammers as opposed to two little ones, but it's at least a start.
UF Cartoon Pitr gets email saying "This is not Unsolicited Bulk Email", and decides to fixink their leetle red wagon. "What happened to our server?" "It's flooded. And there's an email here that says 'this is not a denial of service attack'".
There are different kinds of high-volume attacks against spammers. Some, like the Artists Against 419 web page just download lots of images from the spammer, burning their bandwidth quotas and their 95%ile billing systems. Some submit requests to the spammer's web forms filling them up with junk or complaints. Some send lots of complaint emails to the ISP. All of those seem perfectly fair, particularly if they're directed at the spammers' web pages which are usually cheap services. And yes, some of them try to take down the machine through various mechanisms, which can be rude.
The Artists Against 419 group's Lad Vampire page repeatedly downloads images from 419 scammers' (aka "The Lads"') web pages. It's using your browser to download and render them, so it uses more of your CPU resources than just running wget, but it's pretty easy. Also, on sites that pay by the popular "95th Percentile" bandwidth usage, you only need to pound on them for the first couple days a month to keep their bills high.
On the other hand, the articles on Lycos didn't explain exactly how their attacks worked, but if they're submitting lots of database queries to the spammer web pages to fill them up with garbage, it doesn't need as much bandwidth as a bandwidth-sucking attack.
The Artists Against 419 site repeatedly loads images from about 20 spammer websites, trying to burn their download bandwidth (both on an immediate basis and on a monthly-quota basis, which can shut down lots of cheaper web hosting accounts quickly.) It sounds like the Lycos screensaver only targets a couple of spammers at a time, and if the BBC article is correct (not that you ever expect newspapers to get technology descriptions 100% correct, so insert several grains of salt here) it sounds like Lycos is sending data _to_ the spammers, while AA419 is downloading images from the spammers (If you've got a cable modem or Asymmetric DSL, it's much easier for you to download.)
If Lycos is doing this efficiently, which you can't tell from the article, they could be filling in blanks on the spammer's CGI scripts, which could use up their CPU and database capabilities as opposed to just burning bandwidth, or they could be just sending email, which is less interesting.
AA419's Lad Vampire web page really beats up my 2.4GHz machine, because it keeps trying to render the images after downloading them, using a lot of Mozilla CPU. If you wanted to be more efficient, you could write something similar that just did lots of wgets to/dev/null (obviously trivial on Linux/BSD/GNU/Unix, but probably not too hard on Windows, at least with a bit of Cygwin help.) But loading the page is a no brainer, so I just do it when I'm not going to be using the machine for a while.
Sure, it's a Microsoft protocol, and an old funky one they mostly inherited. That means that if you implement it, you need to not only be careful in all the ways you're always supposed to be careful, like checking array bounds and checking for malicious input, but you need to think about assumptions that the protocol makes about problems that the MS file system or other OS parts would fix that you're going to have to handle for yourself (or that you can trust Unix to take care of for you), and threats that wouldn't have bothered Windows that might bother Linux, and features that Windows wants that you really shouldn't fix simply by running as root, etc.
I took Computer Science 101 back in 1974, and the first couple of lessons we learned were
Never trust your input data - always check it!
Always check for array-bounds and function return codes.
Document everything with comments and design documents.
If you forget a semicolon or close-parentheses, the compiler will try to fix it, but it'll probably do it wrong.
Always number your punch-cards so that you can resort them if you drop them.
One and a half of those things are no longer true, and most of the security holes seem to come from people ignoring the first two of these principles. I really really like the C Language, but people shouldn't use it for anything sensitive unless they're willing to be really careful, but there's too much badly written code out there.
I'm quite skeptical of the games "re-enactment shows it's possible" contention. Sure, you can make a computer simulation where it's easy for the FPS player to shoot the President, but it's unlikely that they're doing the physics well enough to let anybody prove or disprove the Magic Bullet theory, and the news article and web site don't sound like they're giving you the choice of shooting from the Grassy Knoll either - just that you get to watch the bullet flying in Neo-like slow motion. It ain't science. (Also, of course, even if Oswald was the only gunman, that doesn't prove he was a lone assassin vs. getting support and funding from the Mafia or the Cubans or the Pentagon or whomever your favorite conspiracy theory likes to pin it on.)
That's NOT why any ISP wants prioritization
on
SBC's VoIP End Run
·
· Score: 1
In fact, giving positive priorities to VOIP and interactive video-conferencing are the main reasons carriers are interested in QoS solutions. Some of that's for services they deploy themselves, but mostly it's to sell differentiated-quality services to businesses that are willing to pay more to get better real-time treatment.
For the business market, you could really accomplish 90% of the QoS needs by simple egress queuing that gives a lower priority to HTTP, SMTP, and maybe FTP, or a bit more crudely, by just giving a higher priority to UDP and a lower priority to TCP (though there are some database applications running on TCP that can get a bit grumpy about that.) For the home and university market, it's a bit different, because you need to give BitTorrent and other easily-recognized P2P protocols the lowest priority, and it may help to have three queues so you can give VOIP high priority and HTTP medium priority, but that's getting more complex.
Slashdot Mirror
There are times that single-stream downloads can be faster - if you're downloading from a fast lightly loaded server that has more bandwidth per user than your Internet connection, that's as fast as it gets, while if you're running BitTorrent on an asymmetric connection like ADSL or cable modem and the community of people who want the same file is still mostly downloading, then your download speed ends up limited by your upstream bandwidth until lots more people have the file. But it's pretty common for BitTorrent users to leave their clients uploading after they've finished downloading, particularly for big downloads that run overnight (because if it finishes before you get up in the morning, it keeps running.) There are exceptions - not just greedy leeching downloaders, but also people who download when the community has a lot of spare capacity and low demand, but that's when it's really not a problem.
Because BitTorrent doesn't need a fast server to support a lot of downloaders, there's also more material that can be published. If you're running your own tracker for the material you're publishing, that does take some bandwidth, but it's a lot less than actually downloading the file to lots of people.
I agree that a polished Hollywood flick would be the wrong way to do that movie, but the basic problem is that nobody takes the quantities of acid that were obviously involved in making the first one...
I haven't seen the trailer, and have no plans to see the movie, but if you _wanted_ to remake Willy Wonka for some bizarre reason, Depp is almost the only person I could think of to star in it. You could take it in somewhat different directions (e.g. Jim Carey or a younger Tim Curry) but Johnny Depp is really the right guy.
Sure, there are people who don't like the look of wind farms, because you can see the things, just like there are people who don't like seeing cell phone antennas. I've driven by the Altamont Pass wind farms fairly often, and once looked into renting a house that was located out there, and ok, it was a bit spacey, and if you've got epilepsy it might not be where you want to live, but the wind turbines are nowhere near as ugly as a smokestack or a coal strip-mine.
The Artists Against 419 bandwidth sucker has taken out a couple hundred spammer websites from the Nigerian 419 crowd. I'm not running it today - the new NetBSD release came out so I'm wasting my bandwidth running Bittorrent instead (and there seem to be lots of high-bandwidth people seeding the torrent, so I've been downloading at 1.5 Mbps all morning.)
I first met Dave Farber in ~1992. I was taking the train from DC back to New Jersey, and there was this old guy sitting across the table from me with the smallest laptop I'd ever seen, a cellphone smaller than a brick (Moto flip-phones were still amazing back then), and an alphanumeric Skypager. We started talking, and I recognized his name from various sources - his IP list was much smaller back then, and he'd been one of the advisors to the recently-formed EFF, and I was doing cypherpunk stuff with some of the same people. The laptop was an IBM Japan model, with Japanese keyboard, about six pounds, and it was quite a while before anything that light was available here. (And yes, it was a black rectangle design like most of the newer Thinkpads, but black was still cool back then.) He'd been toy-shopping at Akihabara.
Vince Cate, the author of this article, lives in Anguilla. He's a well-known ex-USAn cypherpunk who moved down there a number of years ago, and runs offshore hosting services and such. Anguilla's not a big place, maybe 10000 people, and while it's not rich, it's also not dirt-poor, and some things are expensive because of ridiculous import taxes (like many third-world countries), or because of inherent costs (electricity there is expensive and doesn't sound highly reliable.)
They're intended for markets where people can't pay as much for them, and there's probably less profit per box in them than in higher-end machines - but as other posters have pointed out, if rich people buy them, AMD can just crank out more of these, because they're Making A Profit on each one (or at least, they're making a Marginal Profit on each one; total profit depends on whether they sell enough of them to cover their initial development costs.) If rich people turn these things into X-terminals for their kids or whatever, that's just more sales, which is a Good Thing.
This isn't a charity project - this is a system designed to cost less because it's got less legacy clutter on it, plus it can use less memory / disk / CPU because it's running a less bloated operating system. They're still paying for WinCE, though the PDA market price points means that they probably pay less for WinCE than for WinXP, and Linux would still be a bit cheaper.
But the real problem isn't just whether you mind sending challenges to unknown people who send you legitimate email - it's whether _they_ mind, or whether they ignore your TMDA and their mail to you gets lost. If they're spammers, of course, you don't care, but if they're people you'd be interested in hearing from (e.g. somebody asking a question about your web site), it would be nice not to annoy them.
But it isn't designed to block spammers, just forgers, and spammers are already starting to adopt SPF - so you can tell that mail you received really did orginate at bigspammer.com. This means that you can't use "SPF says it's plausible" to mean "it's not spam", so you'll still need to send them your TMDA if you want to prove they're a human.
Secondly, SMTP supports tagged addresses of the form username+tag@example.com, and your email client can filter on the tags. Unfortunately, that's not foolproof - many web forms choke on the "+", because it has syntactic value to CGI and they're not always bright enough to escape the character. (But almost everything can handle subdomain-format tags.) Also unfortunately, Pobox's forwarding service and .forward files and other mail forwarders generally don't know how to preserve the tags while forwarding mail, though sometimes they can at least forward the mail.
The biggest problem with tagged emails is that to use them effectively, your email client needs to keep track of them, so if you get mail addressed to tag-for-alice@username.example.com, your reply to it will come From: tag-for-alice@username.example.com and not From: somedefaultvalue@username.example.com, and if you're sending mail to alice@alice.com, your mail client will know to send mail from tag-for-alice@username.example.com or whatever the last tag was that you used to send it to that address. Also, your email client should keep track of all the addresses you've sent out, because you might want to handle mail from unknown tags differently.
Setting "Precedence: Bulk" would seem to discourage reading, but at least it seems to be a common convention that vacation-mailers don't respond to it.
The real problem you've got is that of the 4000 unknown addresses that you received email from, as many as 1000 might have been from real people rather than spammers, but most of them didn't bother replying. It's possible that only 10 of them were real people, so maybe only a couple of the real people who'd sent you email you might have cared about didn't bother replying to your TMDA, but it's also possible that 992 of them didn't, i.e. 99% of the real senders. You can't easily tell, except of course for the undeliverable addresses which were probably forged (or else are on email systems that don't let strangers verify addresses any more because spammers abuse them.)
It's an obvious idea, but unfortunately an inadequate one. Too many spammers, especially phishers, include legitimate URLs to evade Bayesian filters or trick users into thinking they're really getting email from Citibank. Most of the legitimate URLs are big enough it's not a problem (e.g. Citibank probably wouldn't notice the hit), but some are smaller and would suffer. So you've got to check the URLs, probably manually, and only hit the ones that aren't legit. Also, lots of spammer URLs really go to free web sites where they redirect to their real ones using Meta-Refresh or whatever, so the URL you're pounding isn't the real target.
The Artists Against 419 page uses a browser with some Javascript or equivalent to repeatedly download web pages from Nigerian 419 scammers. It's much less efficient, because the browser renders all the pages each time instead of just downloading them to /dev/null, and it's not a screensaver, but it's a no-brainer to use.
Sure, it's a whack-a-mole game, and if you only take out two of them, there'll be two more to take their place. It's obviously more effective if you can take out two big ROKSO-known spammers as opposed to two little ones, but it's at least a start.
There are different kinds of high-volume attacks against spammers. Some, like the Artists Against 419 web page just download lots of images from the spammer, burning their bandwidth quotas and their 95%ile billing systems. Some submit requests to the spammer's web forms filling them up with junk or complaints. Some send lots of complaint emails to the ISP. All of those seem perfectly fair, particularly if they're directed at the spammers' web pages which are usually cheap services. And yes, some of them try to take down the machine through various mechanisms, which can be rude.
On the other hand, the articles on Lycos didn't explain exactly how their attacks worked, but if they're submitting lots of database queries to the spammer web pages to fill them up with garbage, it doesn't need as much bandwidth as a bandwidth-sucking attack.
If Lycos is doing this efficiently, which you can't tell from the article, they could be filling in blanks on the spammer's CGI scripts, which could use up their CPU and database capabilities as opposed to just burning bandwidth, or they could be just sending email, which is less interesting.
AA419's Lad Vampire web page really beats up my 2.4GHz machine, because it keeps trying to render the images after downloading them, using a lot of Mozilla CPU. If you wanted to be more efficient, you could write something similar that just did lots of wgets to /dev/null (obviously trivial on Linux/BSD/GNU/Unix, but probably not too hard on Windows, at least with a bit of Cygwin help.) But loading the page is a no brainer, so I just do it when I'm not going to be using the machine for a while.
At least, that's what Slashdot thought at 6:29am today. Maybe the collective wisdom has changed its mind? :-)
Sure, it's a Microsoft protocol, and an old funky one they mostly inherited. That means that if you implement it, you need to not only be careful in all the ways you're always supposed to be careful, like checking array bounds and checking for malicious input, but you need to think about assumptions that the protocol makes about problems that the MS file system or other OS parts would fix that you're going to have to handle for yourself (or that you can trust Unix to take care of for you), and threats that wouldn't have bothered Windows that might bother Linux, and features that Windows wants that you really shouldn't fix simply by running as root, etc.
- Never trust your input data - always check it!
- Always check for array-bounds and function return codes.
- Document everything with comments and design documents.
- If you forget a semicolon or close-parentheses, the compiler will try to fix it, but it'll probably do it wrong.
- Always number your punch-cards so that you can resort them if you drop them.
One and a half of those things are no longer true, and most of the security holes seem to come from people ignoring the first two of these principles. I really really like the C Language, but people shouldn't use it for anything sensitive unless they're willing to be really careful, but there's too much badly written code out there.I'm quite skeptical of the games "re-enactment shows it's possible" contention. Sure, you can make a computer simulation where it's easy for the FPS player to shoot the President, but it's unlikely that they're doing the physics well enough to let anybody prove or disprove the Magic Bullet theory, and the news article and web site don't sound like they're giving you the choice of shooting from the Grassy Knoll either - just that you get to watch the bullet flying in Neo-like slow motion. It ain't science. (Also, of course, even if Oswald was the only gunman, that doesn't prove he was a lone assassin vs. getting support and funding from the Mafia or the Cubans or the Pentagon or whomever your favorite conspiracy theory likes to pin it on.)
For the business market, you could really accomplish 90% of the QoS needs by simple egress queuing that gives a lower priority to HTTP, SMTP, and maybe FTP, or a bit more crudely, by just giving a higher priority to UDP and a lower priority to TCP (though there are some database applications running on TCP that can get a bit grumpy about that.) For the home and university market, it's a bit different, because you need to give BitTorrent and other easily-recognized P2P protocols the lowest priority, and it may help to have three queues so you can give VOIP high priority and HTTP medium priority, but that's getting more complex.