Because a real Tier 1 provider has peering agreements with all other Tier 1's. Thus, all of its traffic can be routed to its destination through a peer for 'free' (since peering agreements are bilateral).
If we have provider X and Y, both Tier 1s, and provider Z that purchases transit from Y, Zs network will be advertised to other Tier 1s as part of Ys, and thus Ys peers can transfer to Z over the peer link.
Because they peer with everyone, they don't have to pay any more than infrastructure costs.
The SMTP thing is 99.9999% of the time a perfectly valid one. It prevents dumbass trojan spamming bots from sending SMTP data directly to the destination (ie. victim). The ISPs SMTP server probably also has a sane (say, 5 mails per minute) limit on the number of messages you can send. It also prevents exploiting open relays and other such nonsense.
As long as they provide a stable, working SMTP relay with sane usage limits, I don't see the beef with this. It cust down on crap coming from their network and improves the internet in general, whilst causing their customers to understand what they're doing to make it work.
The summary is incorrect. Companies that lease from Telus only get ATM transport from the DSLAM into their data centres. To the best of my knowledge, Telus doesn't provide end-to-end DSL resale agreements like they do with dialup (VPOPs). They provide transit and the actual DSL circuit, but not the router inbetween or any other infrastructure. It's raw, as you would expect of such an agreement.
The filtering is undoubtedly being done by the transparent HTTP proxying Telus imposed on all of its customers a couple years back, probably for precisely this purpose (well, ok, they sell a 'protect your kids' service that uses it as well).
Anyway, I'm on an ISP that leases the DSL circuit from Telus (which I'm an ex-employee of), and I can access the site just fine through my Uniserve-owned IP and over Navigata transit. My packets don't go near Telus' IP network.
The OpenSSL speed suite is about 2.87% (averaged over all of the throughput tests) faster with gcc 4 (Debian prerelease) than gcc 3.3.5 on my Duron machine. There were no cases where gcc 4 was more than 0.5% slower than 3.3.5.
You're using the exact same argument he is: spurious data and impossible to confirm assumptions. You're just putting the people on different sides of the equations. On the one hand (your side), people that otherwise would have bought the material don't. On the other hand, people enjoy the material so they purchase it.
Personally, I'm more inclined to believe that the lazy, fairly rich majority are more likely to be in the second category. Getting decent quality full CDs is still fairly difficult on P2P (though I doubt this will be true forever), and this is even more true of films. The poor quality definitely works for the movie studios too. If you get halfway through a film you're enjoying and it dies on you, most people wouldn't think twice before jumping in the car and driving down to Blockbuster to pony up the cash to rent the movie.
Both types exist; which prevail is anyone's guess, and your guessing is no more credible than anyone else's. I'll take record profits in recent years by the RIAA as an indicator, however. The economy has only been going downhill, while the RIAA sues everyone, calling a majority of middle-class America thieves...yet they still manage to break profit records.
Post 9/11, it seems most stations have two personnel stationed at all times. I'm sure this alleviates concerns about security, as well as improving turnaround times when problems arise. I choke everytime I think about how much it costs, though.
In any case, nowadays there are usually at least two SkyTrain employees at every station; if Control notices a train isn't leaving the station on schedule, they'll alert the security people to check it out.
I take the SkyTrain every day to school and back, and I don't think I've ever had a problem with the doors being held open longer than 5-10 seconds (though it's conceivable..the doors are tough, but they're designed not to hurt you). More common are hold-ups due to switching problems or e.g. someone on a train having a heart attack.
It runs remarkably smoothly considering it's all computer controlled and designed in 1985.
I'm not sure about the battery life issue. My two-year old Panasonic MP3 discman came with a set of Li-Ions (it's an ultra-thin model, and the batteries must be flat, and narrower on that axis than AAAs to fit). Battery life is well above 24 hours in practical use, well above that of an HD-based player, and approaching that of a flash player.
Sure, it's not the coolest thing ever, but it's a good medium between flash and HD players. Much higher practical capacity than a flash player, lower price, similar battery life.
Skipping is an absolute non-issue with any of these players. My particular player has an insanely long buffer, and when playing MP3s, the disc actually stops spinning for 30-45s at a time to reduce power consumption, and obviously the buffer is longer than that.
Wow, I totally remember checking those magazines out from the school library JUST for the BASIC games. Entered 'em into the parent's XT-compatible and hoped for the best. They usually didn't work, so I'd read through, double checking everything (it usually wasn't a typing error though). Eventually I got pretty decent at sorting out the problem with the code and getting it to run.
Yeesh, I guess I must've been 9 or 10, thanks for the reminder:D.
Once a release is licensed under a specific license, that license cannot be revoked. Program 1.0 is still GPL, and can be forked and extended, despite 1.1 being released under a non-GPL license.
There do exist some publically funded and run libraries around here. I saw one on one of the Saltspring Islands this summer. They probably don't get any tax dollars and are funded by nominal membership fees directly from the users.
Seems fairly reasonable to me. It's better than the alternative: no library. And you don't have to pay if you don't want to use it.
The firewall *rules* and *logic* would still be controlled fully by an external configuration program. But the filtering itself would be done directly in the core TCP stack, by the same code no matter the configuration utility in use. If done properly, it shouldn't be any less flexible than what we already have. It should also perform better, be more secure, and have less bugs.
Yes, I realize IE has a ton of CSS bugs and flaws in it's implementation. But it's possible to make a perfectly workable CSS compliant layout that renders properly in it. Because that's possible, it's viable to start migrating now; there is no reason to wait for full compliance (if it should ever appear in IE).
Why is this so difficult to understand? There isn't any valid argument I've heard for table-based layouts and other such nonsense. Full compatibility with version 4 browsers is a ton of work for little to no gain (and probably a substantial loss in the design elements availiable to the designer). Using CSS (even IEs broken subset and it's requisite hacks) is easier, more flexible, and more forward (and backward) looking. It works better for everyone involved, from Grandma on her Pentium 100 through RMS browsing with emacs to the developer writing the markup and stylesheets.
EVERYONE knows that IE is broken. That doesn't change the fact that CSS is superiour in every way, and that there's not really any reasonable excuse to use anything else.
I understand that the car will move with only the starter powering it..but in the two cars I've done it in (both small, light commuter cars), it jumps forward to, apparently, a certain point in the next engine cycle and stops dead. I'm assuming it's compression (with the added burden of pulling the car) that it doesn't have enough power for, but I'm not a mechanic. I doubt that quick jerk is enough to get the engine going with the car in gear.
If it was in neutral, you would be able to start the car normally, but then you'd need to get it into first, and the only way to do it cleanly without a clutch is to have both the output portion of the transmission already spinning (which requires the car to have been moving), and the engine at the right RPM for the target gear and current speed of the spinning dog. Impossible without rolling down a hill or pushing the car.
Either way much more of a pain, and more conspicuous than depressing a switch on the bottom of the steering column and careful driving somewhere safe using the parking brake and/or downshifting (while a pain, it still works to some extent with automatics).
Well I doubt you'll be able to get the starter to start the car with the transmission engaged and the wheels touching the ground. And good luck bomb starting it without a clutch...it might be doable but I imagine it's prohibitively difficult even for 2 or 3 people. Leaving it in neutral, obviously, avoids that issue entirely.
And yeah, it is possible to shift without a clutch, but I *think* (and I'm no mechanic) that you'd have quite a difficult time launching from a dead stop (shifting from neutral to 1st) without one. Down a hill or something, maybe. Still a lot of trouble and would definitely attract attention.
Still better than what you can do easily with an auto and the same device.
Your suggestion of locking the clutch in the disengaged position is obviously better, but I was just trying to give a better way to use the same device on a manual transmission car.
Except that the major browser is standards compliant enough to migrate today. Sure it takes some hacks and stuff to get it to work fine, but it complies with the standards and is far superiour to the methods of the past.
My point was that the 'major browser' consists of IE6 and IE5.5, both of which support CSS well enough to already be useful to both users and web developers. I was referring to people using version 4 browsers, and other obscure stuff like text readers or text-only browsers. They'll get a better experience, the web developer spends a lot less time maintaining pages, and the up to date users get a spiffier page. Everyone wins.
I think putting the lock on the clutch would work fine. Modern cars won't turn over without the clutch depressed, and vintage models will start in neutral, but it would be impossible to put them into gear without a clutch. This approach actual seems superiour to the brake pedal approach. An enterprising theif could use the interlock override to engage the transmission, and use the handbrake to try to steal the car (or at least get it to a safer location to remove the brake lock). With a standard it would be more difficult to move the car.
Of course, the device is probably designed for the wider automatic-style brake pedals and may not work well with standards.
Isn't the entire point (well, one of the many) of coding to standards that the page will degrade cleanly on lesser browsers?
If all is done properly, users will see a page that's perfectly usable, just not quite as 'pretty' as it could be. See large, successful sites that've already made the move to CSS (Wired was one of the first, DeviantArt being another notable example).
I don't see many negatives to migrating immediately. People using browsers that don't handle CSS properly are probably used to seeing things render poorly, and currently popular techniques don't degrade nearly as nicely. They should be glad for a more usable web on their ancient software. Everyone else shouldn't notice a difference (ie. your average user, though many will see improved usability), and it makes web dev's jobs much easier.
Slashdot Mirror
How about AdBusters?
;). Pretty pricey though.
More fun too
Because a real Tier 1 provider has peering agreements with all other Tier 1's. Thus, all of its traffic can be routed to its destination through a peer for 'free' (since peering agreements are bilateral).
If we have provider X and Y, both Tier 1s, and provider Z that purchases transit from Y, Zs network will be advertised to other Tier 1s as part of Ys, and thus Ys peers can transfer to Z over the peer link.
Because they peer with everyone, they don't have to pay any more than infrastructure costs.
Wow, and I thought that the English language was confusing enough already...
The SMTP thing is 99.9999% of the time a perfectly valid one. It prevents dumbass trojan spamming bots from sending SMTP data directly to the destination (ie. victim). The ISPs SMTP server probably also has a sane (say, 5 mails per minute) limit on the number of messages you can send. It also prevents exploiting open relays and other such nonsense.
As long as they provide a stable, working SMTP relay with sane usage limits, I don't see the beef with this. It cust down on crap coming from their network and improves the internet in general, whilst causing their customers to understand what they're doing to make it work.
The summary is incorrect. Companies that lease from Telus only get ATM transport from the DSLAM into their data centres. To the best of my knowledge, Telus doesn't provide end-to-end DSL resale agreements like they do with dialup (VPOPs). They provide transit and the actual DSL circuit, but not the router inbetween or any other infrastructure. It's raw, as you would expect of such an agreement.
The filtering is undoubtedly being done by the transparent HTTP proxying Telus imposed on all of its customers a couple years back, probably for precisely this purpose (well, ok, they sell a 'protect your kids' service that uses it as well).
Anyway, I'm on an ISP that leases the DSL circuit from Telus (which I'm an ex-employee of), and I can access the site just fine through my Uniserve-owned IP and over Navigata transit. My packets don't go near Telus' IP network.
A network is a communal resource. Is it really that far-out?
The OpenSSL speed suite is about 2.87% (averaged over all of the throughput tests) faster with gcc 4 (Debian prerelease) than gcc 3.3.5 on my Duron machine. There were no cases where gcc 4 was more than 0.5% slower than 3.3.5.
You're using the exact same argument he is: spurious data and impossible to confirm assumptions. You're just putting the people on different sides of the equations. On the one hand (your side), people that otherwise would have bought the material don't. On the other hand, people enjoy the material so they purchase it.
Personally, I'm more inclined to believe that the lazy, fairly rich majority are more likely to be in the second category. Getting decent quality full CDs is still fairly difficult on P2P (though I doubt this will be true forever), and this is even more true of films. The poor quality definitely works for the movie studios too. If you get halfway through a film you're enjoying and it dies on you, most people wouldn't think twice before jumping in the car and driving down to Blockbuster to pony up the cash to rent the movie.
Both types exist; which prevail is anyone's guess, and your guessing is no more credible than anyone else's. I'll take record profits in recent years by the RIAA as an indicator, however. The economy has only been going downhill, while the RIAA sues everyone, calling a majority of middle-class America thieves...yet they still manage to break profit records.
Post 9/11, it seems most stations have two personnel stationed at all times. I'm sure this alleviates concerns about security, as well as improving turnaround times when problems arise. I choke everytime I think about how much it costs, though.
In any case, nowadays there are usually at least two SkyTrain employees at every station; if Control notices a train isn't leaving the station on schedule, they'll alert the security people to check it out.
I take the SkyTrain every day to school and back, and I don't think I've ever had a problem with the doors being held open longer than 5-10 seconds (though it's conceivable..the doors are tough, but they're designed not to hurt you). More common are hold-ups due to switching problems or e.g. someone on a train having a heart attack.
It runs remarkably smoothly considering it's all computer controlled and designed in 1985.
Wow, someone I could recognize IRL got modded +5 on Slashdot. Amazing.
I'm not sure about the battery life issue. My two-year old Panasonic MP3 discman came with a set of Li-Ions (it's an ultra-thin model, and the batteries must be flat, and narrower on that axis than AAAs to fit). Battery life is well above 24 hours in practical use, well above that of an HD-based player, and approaching that of a flash player.
Sure, it's not the coolest thing ever, but it's a good medium between flash and HD players. Much higher practical capacity than a flash player, lower price, similar battery life.
Skipping is an absolute non-issue with any of these players. My particular player has an insanely long buffer, and when playing MP3s, the disc actually stops spinning for 30-45s at a time to reduce power consumption, and obviously the buffer is longer than that.
Wow, I totally remember checking those magazines out from the school library JUST for the BASIC games. Entered 'em into the parent's XT-compatible and hoped for the best. They usually didn't work, so I'd read through, double checking everything (it usually wasn't a typing error though). Eventually I got pretty decent at sorting out the problem with the code and getting it to run.
:D.
Yeesh, I guess I must've been 9 or 10, thanks for the reminder
What does that make FTP? MI-6?
But damn cool stunts, and some neat physics effects.
Are you trying to say that lockpicks exist to kill people?
Slackware is a Linux distro for BSD people. Linux people generally prefer Debian or Gentoo. Slackware is stuck in the stone ages.
It has it's uses, though.
You're missing the point.
Once a release is licensed under a specific license, that license cannot be revoked. Program 1.0 is still GPL, and can be forked and extended, despite 1.1 being released under a non-GPL license.
There do exist some publically funded and run libraries around here. I saw one on one of the Saltspring Islands this summer. They probably don't get any tax dollars and are funded by nominal membership fees directly from the users.
Seems fairly reasonable to me. It's better than the alternative: no library. And you don't have to pay if you don't want to use it.
Uh? Did you not understand what he was saying?
The firewall *rules* and *logic* would still be controlled fully by an external configuration program. But the filtering itself would be done directly in the core TCP stack, by the same code no matter the configuration utility in use. If done properly, it shouldn't be any less flexible than what we already have. It should also perform better, be more secure, and have less bugs.
Yes, I realize IE has a ton of CSS bugs and flaws in it's implementation. But it's possible to make a perfectly workable CSS compliant layout that renders properly in it. Because that's possible, it's viable to start migrating now; there is no reason to wait for full compliance (if it should ever appear in IE).
Why is this so difficult to understand? There isn't any valid argument I've heard for table-based layouts and other such nonsense. Full compatibility with version 4 browsers is a ton of work for little to no gain (and probably a substantial loss in the design elements availiable to the designer). Using CSS (even IEs broken subset and it's requisite hacks) is easier, more flexible, and more forward (and backward) looking. It works better for everyone involved, from Grandma on her Pentium 100 through RMS browsing with emacs to the developer writing the markup and stylesheets.
EVERYONE knows that IE is broken. That doesn't change the fact that CSS is superiour in every way, and that there's not really any reasonable excuse to use anything else.
I understand that the car will move with only the starter powering it..but in the two cars I've done it in (both small, light commuter cars), it jumps forward to, apparently, a certain point in the next engine cycle and stops dead. I'm assuming it's compression (with the added burden of pulling the car) that it doesn't have enough power for, but I'm not a mechanic. I doubt that quick jerk is enough to get the engine going with the car in gear.
If it was in neutral, you would be able to start the car normally, but then you'd need to get it into first, and the only way to do it cleanly without a clutch is to have both the output portion of the transmission already spinning (which requires the car to have been moving), and the engine at the right RPM for the target gear and current speed of the spinning dog. Impossible without rolling down a hill or pushing the car.
Either way much more of a pain, and more conspicuous than depressing a switch on the bottom of the steering column and careful driving somewhere safe using the parking brake and/or downshifting (while a pain, it still works to some extent with automatics).
Well I doubt you'll be able to get the starter to start the car with the transmission engaged and the wheels touching the ground. And good luck bomb starting it without a clutch...it might be doable but I imagine it's prohibitively difficult even for 2 or 3 people. Leaving it in neutral, obviously, avoids that issue entirely.
;).
And yeah, it is possible to shift without a clutch, but I *think* (and I'm no mechanic) that you'd have quite a difficult time launching from a dead stop (shifting from neutral to 1st) without one. Down a hill or something, maybe. Still a lot of trouble and would definitely attract attention.
Still better than what you can do easily with an auto and the same device.
Your suggestion of locking the clutch in the disengaged position is obviously better, but I was just trying to give a better way to use the same device on a manual transmission car.
But I assume neither of us has even seen it
Except that the major browser is standards compliant enough to migrate today. Sure it takes some hacks and stuff to get it to work fine, but it complies with the standards and is far superiour to the methods of the past.
My point was that the 'major browser' consists of IE6 and IE5.5, both of which support CSS well enough to already be useful to both users and web developers. I was referring to people using version 4 browsers, and other obscure stuff like text readers or text-only browsers. They'll get a better experience, the web developer spends a lot less time maintaining pages, and the up to date users get a spiffier page. Everyone wins.
I think putting the lock on the clutch would work fine. Modern cars won't turn over without the clutch depressed, and vintage models will start in neutral, but it would be impossible to put them into gear without a clutch. This approach actual seems superiour to the brake pedal approach. An enterprising theif could use the interlock override to engage the transmission, and use the handbrake to try to steal the car (or at least get it to a safer location to remove the brake lock). With a standard it would be more difficult to move the car.
Of course, the device is probably designed for the wider automatic-style brake pedals and may not work well with standards.
Isn't the entire point (well, one of the many) of coding to standards that the page will degrade cleanly on lesser browsers?
If all is done properly, users will see a page that's perfectly usable, just not quite as 'pretty' as it could be. See large, successful sites that've already made the move to CSS (Wired was one of the first, DeviantArt being another notable example).
I don't see many negatives to migrating immediately. People using browsers that don't handle CSS properly are probably used to seeing things render poorly, and currently popular techniques don't degrade nearly as nicely. They should be glad for a more usable web on their ancient software. Everyone else shouldn't notice a difference (ie. your average user, though many will see improved usability), and it makes web dev's jobs much easier.