Hypothesis 1: the same number of cables get cut in the past two weeks as in the entire 50 years previous to that.
The problem is that cable "cuts" (actual term = "disruptions of service") happen ALL THE TIME. There's nearly always a half-dozen or more ongoing major incidents. There's enough spare capacity on parallel and diverse cables so that most folks don't notice a change in service. It's when there are simultaneous cuts on several parallel cables that people really notice. As an example, the cable breaks after the tsunami in Asia.
So, let me restate your hypothesis #1:
Hypothesis 1: The same number of cables get disrupted in the past two weeks as in the prior month. This time, they're all in the same vicinity.
CDBaby has been great for me - my band has found some listeners we otherwise wouldn't have, and I've found a couple of great other groups. I'm a big fan of the "random walk" approach to trying music, and their interface lets you listen to 2 minutes of the songs for free before buying. Much awesomeness.
IPv6 has better security provisions within the protocol itself, making the usual run of D- through to F- on Federal security audits less likely.
No, it doesn't. The IPSec header field in IPv6 works in the exact same way that it does in IPv4. The possible benefit of including it in the spec is that it'd theoretically be easier to have interoperable implementations of IP6Sec. The reason.gov gets a D- or F doesn't have to do with the level of or quality of the encryption used, it has to do with things like password control, physical security, wireless implementations, network ports in conference rooms and public areas, and the like. I am not aware of a case of an agency getting cited for implementing IPsec incorrectly, and IP6sec isn't likely to be implemented anywhere where IPsec wasn't.
The protocol incorporates many of the features back-engineered into IPv4 as standard, producing a cleaner design with fewer compromises and fewer flaws
huh? oh you mean autoconfiguration, which conveniently forgot to include DNS server location, right? The "elegance" of the protocol does not matter to an end user or an agency. SNA/Token Ring is far more elegant than IP/Ethernet, but which one is more common? Also, it's hard not to call the/64 host address a compromise - that burns half of the available space, and assumes that every interface is an Ethernet.
Built-in support for protocol expansion means future updates should have less impact and be adoptable faster
Wasn't that what you were saying was the problem with IPv4? That people had written updates (for instance, repurposing the ToS byte)
Automatic configuration means fewer errors and less maintenance
Show me a case of IPv6 autoconfiguration working better than DHCPv4, and I'll be very surprised. You still have to run DHCPv6 if you want hosts to find DNS servers, and further, autoconfiguration means that if you change the NIC on your server, your autoconfigured address changes. Ugh!
Alignment of entries in the header means potentially greater throughput
You've got to be kidding: the size of the header grew tremendously - once there are nicely-spun asics for IPv6, the forwarding performance will be approximately equal to v4, but there's certainly not going to be a performance improvement.
Skript Kiddies will end up jumping off bridges as they won't know what to do Software contracting firms are located in regions in which elections are due, creating excellent opportunities on both sides of the table
this band isn't affiliated with the RIAA either. I've gotta agree with the parent: CDBaby and myspace are pretty good, although I've noticed that a lot of bands have stopped offering their MP3s for free on myspace.
Another site which I like a lot is 3hive, which does reviews of bands which offer MP3s for free.
Wouldn't the Commerce Clause be uniquely appropriate to Internet traffic? Peering sites could be presumed to carry more Interstate than Intrastate traffic...
(This is a theoretical question, not arguing whether the alleged interception is legal or not - it certainly sounds sketchy...)
This happened recently, where the anonymous blogger "orthomom" got sued by a School Board member for saying nasty things about her. The suit was thrown out of court. Some details are available here.
I still have some old Ethernet hubs with BNC Coax uplinks on them...
Sidenote: if you've got a pair of point to point coax, you could also use them as a DS3 - there are lots of cheap, old M13s and other stuff like that out there - you might not have to futz with DOCSIS, which I believe is asymmetric.
Fair point, and good information. Of course, until the ruling was issued, you had the highest legal officials in the country asserting its legality, and I'm not aware that the DoJ's position has changed on this - so telcos are a little stuck.
That's tremendously appreciated and I understand: I should have presented this as more of a Devil's advocate position (because I'm certainly not in favor of giving away customer information...).
I do think the administration seemed to be playing a bit fast and loose with their requests, and the telcos gave the appearance of eagerness in giving information. I don't know whether that impression is right, but I think anyone involved in the slightest lost a lot of goodwill...
I was not saying "because the government said / did X is was necessarily legal"
I was saying "because the government said / did X, the entity who was asked for assistance does have a reasonable presumption that X is legal" - i.e. they can be shown to be acting in good faith.
Has anyone who has standing to do so actually ruled that AT&T did anything illegal as opposed to unpleasant? Perhaps they do have legions of lawyers after all...
(also, there are PR implications in refusing a direct order when it's couched in national security too...)
Given that AT&T HAS had its ass handed to it by the courts before, I am going to go out on a limb and guess that they've got a lot of lawyers who work there, and that their policies about compliance with government directives is probably thoroughly vetted.
Imagine you are driving, and a cop is directing traffic. He directs you to drive the one way on a one way street. If you comply with his directive, that would be an affirmative defense against another cop giving you a ticket for driving the wrong way.
Use the same analogy for AT&T. By no means is AT&T "too big to fail" - dialtone and ISP services will be provided by *someone* and a lot of equipment which currently has AT&T labels on it would be used, but if the company messed up badly enough, the courts could dismember it again.
You occasionally get cases where it's not that the balance of traffic is equal, but that the value of the traffic is equal: consider the Level3 vs. Cogent battle a year or so ago - L3 eventually had to give up and re-peer Cogent, because L3's users complained louder than Cogent's.
ISPs only carry traffic either to or from someone who pays them.
AT&T carries traffic from its customers to other ISPs, and AT&T carries traffic from other ISPs to its customers.
A bunch of ISPs are themselves customers of AT&T. This is true of any of the big players - nobody carries anything for free: "peering" is what happens when two ISPs agree that metering their back and forth traffic isn't worth the cost of doing so.
Slashdot Mirror
My wife and I met on usenet in 1995.
Hypothesis 1: the same number of cables get cut in the past two weeks as in the entire 50 years previous to that.
The problem is that cable "cuts" (actual term = "disruptions of service") happen ALL THE TIME. There's nearly always a half-dozen or more ongoing major incidents. There's enough spare capacity on parallel and diverse cables so that most folks don't notice a change in service. It's when there are simultaneous cuts on several parallel cables that people really notice. As an example, the cable breaks after the tsunami in Asia.
So, let me restate your hypothesis #1:
Hypothesis 1: The same number of cables get disrupted in the past two weeks as in the prior month. This time, they're all in the same vicinity.
That does actually make a bit more sense, eh?
CDBaby has been great for me - my band has found some listeners we otherwise wouldn't have, and I've found a couple of great other groups. I'm a big fan of the "random walk" approach to trying music, and their interface lets you listen to 2 minutes of the songs for free before buying. Much awesomeness.
No, it doesn't. The IPSec header field in IPv6 works in the exact same way that it does in IPv4. The possible benefit of including it in the spec is that it'd theoretically be easier to have interoperable implementations of IP6Sec. The reason
huh? oh you mean autoconfiguration, which conveniently forgot to include DNS server location, right? The "elegance" of the protocol does not matter to an end user or an agency. SNA/Token Ring is far more elegant than IP/Ethernet, but which one is more common? Also, it's hard not to call the
Wasn't that what you were saying was the problem with IPv4? That people had written updates (for instance, repurposing the ToS byte)
Show me a case of IPv6 autoconfiguration working better than DHCPv4, and I'll be very surprised. You still have to run DHCPv6 if you want hosts to find DNS servers, and further, autoconfiguration means that if you change the NIC on your server, your autoconfigured address changes. Ugh!
You've got to be kidding: the size of the header grew tremendously - once there are nicely-spun asics for IPv6, the forwarding performance will be approximately equal to v4, but there's certainly not going to be a performance improvement.
huh?
You are the wind beneath my wings.
this band isn't affiliated with the RIAA either. I've gotta agree with the parent: CDBaby and myspace are pretty good, although I've noticed that a lot of bands have stopped offering their MP3s for free on myspace.
Another site which I like a lot is 3hive, which does reviews of bands which offer MP3s for free.
Only the old people in Korea spam slashdot...
Natalie Portman spams Slashdot with hot grits?
Wouldn't the Commerce Clause be uniquely appropriate to Internet traffic? Peering sites could be presumed to carry more Interstate than Intrastate traffic...
(This is a theoretical question, not arguing whether the alleged interception is legal or not - it certainly sounds sketchy...)
This happened recently, where the anonymous blogger "orthomom" got sued by a School Board member for saying nasty things about her. The suit was thrown out of court. Some details are available here.
I like to think that I wrote this song (Seven Layer Cake) as a teaching tool...
Everything old is new again...
I still have some old Ethernet hubs with BNC Coax uplinks on them...
Sidenote: if you've got a pair of point to point coax, you could also use them as a DS3 - there are lots of cheap, old M13s and other stuff like that out there - you might not have to futz with DOCSIS, which I believe is asymmetric.
Fair point, and good information. Of course, until the ruling was issued, you had the highest legal officials in the country asserting its legality, and I'm not aware that the DoJ's position has changed on this - so telcos are a little stuck.
I didn't say "moral" "right" or "a good idea" - I said "legal." No irony at all, other than that the law is not necessarily moral or right.
This area of law isn't nearly as clear as premeditated murder is.
It's more like getting advice from a tax accountant about whether a deduction is legal: whether or not you are correct, you acted in good faith.
Now, as a matter of "is this a good idea," I think the telcos could have done a bit more homework first...
That's tremendously appreciated and I understand: I should have presented this as more of a Devil's advocate position (because I'm certainly not in favor of giving away customer information...).
I do think the administration seemed to be playing a bit fast and loose with their requests, and the telcos gave the appearance of eagerness in giving information. I don't know whether that impression is right, but I think anyone involved in the slightest lost a lot of goodwill...
Agreed- but what if their lawyers said it was ok?
That's an interesting tidbit- where did you find the list of the oldest domains? I don't have a hosts.txt that's QUITE that old...
I should have been more clear.
I was not saying "because the government said / did X is was necessarily legal"
I was saying "because the government said / did X, the entity who was asked for assistance does have a reasonable presumption that X is legal" - i.e. they can be shown to be acting in good faith.
I assume you can grasp the difference between telling a government official who called whom when and summarily executing thousands of people.
seeya.
Fair point.
Has anyone who has standing to do so actually ruled that AT&T did anything illegal as opposed to unpleasant? Perhaps they do have legions of lawyers after all...
(also, there are PR implications in refusing a direct order when it's couched in national security too...)
Given that AT&T HAS had its ass handed to it by the courts before, I am going to go out on a limb and guess that they've got a lot of lawyers who work there, and that their policies about compliance with government directives is probably thoroughly vetted.
Imagine you are driving, and a cop is directing traffic. He directs you to drive the one way on a one way street. If you comply with his directive, that would be an affirmative defense against another cop giving you a ticket for driving the wrong way.
Use the same analogy for AT&T. By no means is AT&T "too big to fail" - dialtone and ISP services will be provided by *someone* and a lot of equipment which currently has AT&T labels on it would be used, but if the company messed up badly enough, the courts could dismember it again.
You occasionally get cases where it's not that the balance of traffic is equal, but that the value of the traffic is equal: consider the Level3 vs. Cogent battle a year or so ago - L3 eventually had to give up and re-peer Cogent, because L3's users complained louder than Cogent's.
No, of course not.
However, if a member of a law enforcement branch of the government says "this is legal" and it's plausible, I might answer differently.
ISPs only carry traffic either to or from someone who pays them.
AT&T carries traffic from its customers to other ISPs, and AT&T carries traffic from other ISPs to its customers.
A bunch of ISPs are themselves customers of AT&T. This is true of any of the big players - nobody carries anything for free: "peering" is what happens when two ISPs agree that metering their back and forth traffic isn't worth the cost of doing so.
Send the sales guy and the tech guy an email, saying "just to confirm and document our prior conversation ..."
Keep a copy.