Technically, IANA owns all IPv4 addresses (and maybe IPv6 too). Everyone else is leasing them from someone up the tree to IANA. As a general rule, address space not leased from a RIR is not portable -- meaning it's not yours to be subdivided all over the internet.
An antenna aimed at the metal skin of an airplane will have zero effect inside the tin can. (that's basic physics) The best (or worst) one could do is damage the radar and radio equipment that has exterior antenna.
I was wondering why the electronics inside aircraft would be such insanely sensitive devices seeing how almost nothing else on earth even notices the RF from a pager/cellphone. But after considering the RF cage all that equipment lives in and the age of the entire system design... I'm really not surprised it's so easy to screw it up.
(fwiw, there's a Lucent notice hanging on our 5ESS phone switch(es) warning about RF problems from cellphones within 5ft of the equipment. I think I'm the only one who's bothered to read it in years.)
It's not the accelerational stress, it's the amount and magnitude of vibration. In any event, it should hold together for the required 10s +/-. (insert duct tape comment here)
This thing doesn't have to make it to mars. And even if it bursts into a cloud of flames at 5000ft, it'll still be worth the price of admission:-)
Has anyone considered the dopler effect @ mach 3? I don't know how wide a tolerance any 802.11b radios are. Those things weren't designed for shuttle telemetry, after all.
In NC, they are officially designated "utility poles". While the power companies plant and maintain them, the NCPUC licenses and controls them. In effect, as soon as Duke Power finishes erecting it's poles, Bellsouth and Time Warner (among others) have access to the right-of-way.
As for the cost of cabling (buried or otherwise)... a surprising majority of the rural US was cabled decades ago. Explosive growth in urban and suburban areas are where most of the expense can be found today. (If you listen to the telco's, your great grandchildren will still be paying for that cable. Bellsouth collected in the neighborhood of a billion dollars from the "touch-tone service" fee they were allowed to charge for two decades. (they stopped that shit a few years ago.))
Case in point... my phone line (urban Raleigh, NC) was buried in 1997 -- all 4.7 miles of it. I watched their progress. My parent's house was built in 1966. Asuming the cable hadn't already been run -- very likely as there were no other houses on that side of the creek -- their phone line is 36 years old. It hasn't been replaced in my 31 years. The line to my grandparent's on the other side of the creek was put there circa 1940. And it has never been replaced as was confirmed by a bell tech who had to move the junction box in the early 90's after lightning and a severed transmission line fused about 20ft of cable in the front yard (they just dug the box up and moved it to the other side of the yard.)
Most telephone wiring is underground, not on the "telephone" poles. Back home, maybe 100ft of the miles of cable feeding my parents phone is above ground -- basically everywhere it crosses water. Besides, those poles are property of the power company, not the telco.
When the last ice storm blew through, everyone had to wait for the power company to fix the poles before they could fix the phones, cable, etc. that were on those poles. (of course Time Warner took a few weeks to finally fix everything.)
Exactly. Nobody's going to actually check their bill and ask where a few pennies are going. Repeat for a few dozen line items... Multiply by the number of subscribers, and you get a very large amount of pure profit.
(check your local PUC tariffs) DSL is run over an "Unbundled Network Element" (UNE) which is telco speak for an unloaded ("dry") copper loop. There's no services on the line. (For IDSL, it's basically the same thing... an ISDN loop not attached to a switch. It's transported through the PSTN wiring just like any other ISDN loop. It can be a dry loop, but usually isn't...)
Had I ordered by ISDN line as "data only", I wouldn't be (or wouldn't have been -- they've changed the rules several times over the years) charged for universal service (or the 911 access fee.)
[I've worked with these things for several years. Actually, now that I think about, since the very beginning of DSL.]
Exactly to whom does speakeasy provide "universal service"? Where exactly is that money going? In the traditional telco world, that money is collected and kept by the telco. The FCC allows a maximum line item amount for the telco to recover some/all of their costs for providing universal service. 99% of the time, that line item is many thousands of times their cost and thus one hell of a profit padding. I'll add, the FCC does not require telco's to pass the charge on to the customer.
Yes, but the rural infrastructure is far cheaper (as in quality) in it's construction. In a city, there's a fiber loop (or two...) with muxes feeding much shorter copper runs. Out in the sticks, there maybe fiber run to the CO for the entire county, but everything from the CO out is copper -- many, many miles of it. And most of it is many decades old today.
Don't let the FCC and RBOC's fool you. Most of the UFC charges are, by definition, a scam. See that line item on your phone bill? Think it goes to the FCC? WRONG. It goes right in Bell's pocket "to offset the costs of providing universal services."
To be accurate, TiVo has not (and never will) released the source to the non-GPL'd stuff that makes the Tivo what it is... the kernel modules and "tivoapp". The source to all the GPL components and the entire linux kernel running on the tivo are available.
His claims aren't entirely trollish... a 6000x increase in throughput would indicate a Really F**king Broken Network (tm). It doesn't take many dropped packets to kill throughput. Some simple tweaks to the backoff timer(s) goes a long way to fixing that -- I did just that many years ago to address broken VJ header compression on a netopia router.
The "holy grail" is in finding the maximum packet flow rate for a given connection. This depends on a number of constantly moving targets. Packets are dropped due to errors (rare) and queue overflows (very common.) In the case of errors, immediate retransmittion is best. But for overflows, immediately retransmitting the packet is very likely to get it dropped again; of course, there's an ICMP message code for this ("source quench"), but nobody uses it.
Therein is the problem... the people making the legal decissions do not have sufficent technical qualifications to sit in judgement. It's not easy to say what's common and what's inovative, unique code. At any rate, Linux has a lengthy history. Almost every line of code can be traced to when it was introduced and in many cases who put it there.
I recall a program once used at the NCSU computer science department to determine "cheating" on programming assignments. It was wrong ~100% of the time. It was actually rather comical.
InterNIC didn't start to enforce it until it was far too late. Back in the day, people asked for what they wanted to use (or thought they eventually would) instead of the modern day hoarding of domain names... register everything that exists and hold it ransom.
In fact, when I registered troz.com, Network Solutions offered up all three in one package.
News of a serial spam killer would be a ray of sunshine in that dark world. As the saying goes, nobody would shed a tear.
Costs are hard to quantify...
How does one put a dollar sign on latency? Yes, spam consumes bandwidth requiring ISPs to have substaintially larger connections than otherwise necessary -- bandwidth is cheap, but it isn't free. However, the connections are wide enough that spam isn't taking available bits away from other users -- assuming a fair queue, the more users of the bandwidth, the more spread out a users packets become (instead of 3 packets between each of mine, there are 9 or 50.)
How much is one's time worth? And just how much time is spent downloading, qualifying as spam, and deleting email? Even using a 28.8 dialup connection, it takes just a few (annoying) minutes to fetch 30 or 40 messages. It takes a few more minutes to sift through them to weed out all the spam. I've heard of people spending hours per day in this process, but I just don't buy it; unless they are studying every message and bitching to everyone within 100miles, there's no way it can take that long from eyeball to trash.
The only things that are easy to calculate are the administrative costs: software, support for said software, server(s) to run said software, store and service email, and monetary compensation for administrators of said software and systems. And of course, there's the cost(s) of repairs when the spammer's spew disrupts services. Admin overhead amounts to very little of $10billion -- even if we blur the line between "email service" and "spam fighting."
.net,.com,.org... there's been almost zero distinction since around '97 or so. Companies (and even individuals) routinely snarf up foo.everything-they-can-aford. InterNIC only breifly enforced the distinction (or tried to.)
And being in the UK, I don't think Spamhaus really cares about some idiot lawyer-spammer(s) in the US. (Ok, maybe if they're bored.)
Your network connection cost money. The time it takes to download their crap is time you cannot be downloading pr0n and warez.
Bandwidth "comsumed" in the same manner from your ISP (and every network between you and the spammer)... plus the server processing and storage space consumed holding that crap until you delete it.
And the cost to the ISP(s) in admin time maintain mail servers.
--
Case in point, without a dedicated spam filtering system (running mail marshal), no one would be able to access their email because the exchange server would be on fire. There were many hours lost to spam overloading everything and people wading through deleting it (and a great many complaining to helpdesk staff about it.)
The volume of traffic is more than most companies could support. Being the corp. office of a Telco/ISP, that's not really a problem. And in fact is more than most of our customers can support.
Because a great deal of the internet isn't configured the way it should. Most mail handlers will do some basic verification of who the sender claims to be, but that usually stops at making sure the domain exists ("HELO foo.com".) Very few check forward and inverse DNS.
And almost zero process the headers of the incoming spew (that's left to the user/admin to do after the message is received.) At any rate, once you've received the entire message, what's the point? You've already "paid" for it. Spammers go to unbeleivable lengths to randomize their spew to thwart pattern recognition systems.
I've been very tempted to build something like LaBrea to trap spammers. "You can send me that spam... at one character per hour, and I'm not listening."
Billion dollar industry... blah, blah, freakin' blah. Prove the damned numbers. Unlike RIAA and MPAA, no one is going to let spammers make up their own balance sheets. There are numerous reports world wide giving hard proof of the costs brought about by all the stupid spammers. The only people who stand to be finacially injured and unemployed (and unemployable after a background check) are the asses sending all the spam.
I'll see their billions and raise by trillions -- the costs of software development and administrator headaches addressing the problem of spam, software development and administrative overhead to block loopholes in internet protocols, ever increasing server and bandwidth needs to move, process, and store all this crap... SPAM is a very expensive problem with the burden everywhere but the spammer.
Laws are useless unless swiftly and strictly enforced. Speeding is illegal, but that hasn't made much of a dent.
Well, having delt with one spammer in particular (yes, listed by Spamhaus), I'll just say they are very open liars. This individual said -- and I wished I'd been recording the call "for quality assurance purposes":-) -- Spamhaus was a company run by one of his competitors. We had to mute the phone for a few minutes. They insist they are not sending "spam" -- even tho' I have spam reports from every batch of crap they sent.
(650m + 987k + 231k) / 175m =~ 3.72lbs per computer with almost of that being lead. No computer in the world (that's still in use anyway) contains that much lead. CRT monitor certainly have a high lead content... radiation sheilding inside the walls of the picture tube. Computers contain a few ounces of lead in the solder. That's about the only place lead exists in there.
Cadmium? How fsckin' old was that study? No laptop in nearly a decade has used NiCad battery packs. Heck, even my pager has a radio shack NiMh "AA" in it.
Slashdot Mirror
Technically, IANA owns all IPv4 addresses (and maybe IPv6 too). Everyone else is leasing them from someone up the tree to IANA. As a general rule, address space not leased from a RIR is not portable -- meaning it's not yours to be subdivided all over the internet.
An antenna aimed at the metal skin of an airplane will have zero effect inside the tin can. (that's basic physics) The best (or worst) one could do is damage the radar and radio equipment that has exterior antenna.
I was wondering why the electronics inside aircraft would be such insanely sensitive devices seeing how almost nothing else on earth even notices the RF from a pager/cellphone. But after considering the RF cage all that equipment lives in and the age of the entire system design... I'm really not surprised it's so easy to screw it up.
(fwiw, there's a Lucent notice hanging on our 5ESS phone switch(es) warning about RF problems from cellphones within 5ft of the equipment. I think I'm the only one who's bothered to read it in years.)
It's not the accelerational stress, it's the amount and magnitude of vibration. In any event, it should hold together for the required 10s +/-. (insert duct tape comment here)
:-)
This thing doesn't have to make it to mars. And even if it bursts into a cloud of flames at 5000ft, it'll still be worth the price of admission
Has anyone considered the dopler effect @ mach 3? I don't know how wide a tolerance any 802.11b radios are. Those things weren't designed for shuttle telemetry, after all.
In NC, they are officially designated "utility poles". While the power companies plant and maintain them, the NCPUC licenses and controls them. In effect, as soon as Duke Power finishes erecting it's poles, Bellsouth and Time Warner (among others) have access to the right-of-way.
As for the cost of cabling (buried or otherwise)... a surprising majority of the rural US was cabled decades ago. Explosive growth in urban and suburban areas are where most of the expense can be found today. (If you listen to the telco's, your great grandchildren will still be paying for that cable. Bellsouth collected in the neighborhood of a billion dollars from the "touch-tone service" fee they were allowed to charge for two decades. (they stopped that shit a few years ago.))
Case in point... my phone line (urban Raleigh, NC) was buried in 1997 -- all 4.7 miles of it. I watched their progress. My parent's house was built in 1966. Asuming the cable hadn't already been run -- very likely as there were no other houses on that side of the creek -- their phone line is 36 years old. It hasn't been replaced in my 31 years. The line to my grandparent's on the other side of the creek was put there circa 1940. And it has never been replaced as was confirmed by a bell tech who had to move the junction box in the early 90's after lightning and a severed transmission line fused about 20ft of cable in the front yard (they just dug the box up and moved it to the other side of the yard.)
Most telephone wiring is underground, not on the "telephone" poles. Back home, maybe 100ft of the miles of cable feeding my parents phone is above ground -- basically everywhere it crosses water. Besides, those poles are property of the power company, not the telco.
When the last ice storm blew through, everyone had to wait for the power company to fix the poles before they could fix the phones, cable, etc. that were on those poles. (of course Time Warner took a few weeks to finally fix everything.)
The consumer versions have SMP and BIGMEM support. Maybe not at the same time, 'tho.
Exactly. Nobody's going to actually check their bill and ask where a few pennies are going. Repeat for a few dozen line items... Multiply by the number of subscribers, and you get a very large amount of pure profit.
(check your local PUC tariffs) DSL is run over an "Unbundled Network Element" (UNE) which is telco speak for an unloaded ("dry") copper loop. There's no services on the line. (For IDSL, it's basically the same thing... an ISDN loop not attached to a switch. It's transported through the PSTN wiring just like any other ISDN loop. It can be a dry loop, but usually isn't...)
Had I ordered by ISDN line as "data only", I wouldn't be (or wouldn't have been -- they've changed the rules several times over the years) charged for universal service (or the 911 access fee.)
[I've worked with these things for several years. Actually, now that I think about, since the very beginning of DSL.]
Exactly to whom does speakeasy provide "universal service"? Where exactly is that money going? In the traditional telco world, that money is collected and kept by the telco. The FCC allows a maximum line item amount for the telco to recover some/all of their costs for providing universal service. 99% of the time, that line item is many thousands of times their cost and thus one hell of a profit padding. I'll add, the FCC does not require telco's to pass the charge on to the customer.
Yes, but the rural infrastructure is far cheaper (as in quality) in it's construction. In a city, there's a fiber loop (or two...) with muxes feeding much shorter copper runs. Out in the sticks, there maybe fiber run to the CO for the entire county, but everything from the CO out is copper -- many, many miles of it. And most of it is many decades old today.
Don't let the FCC and RBOC's fool you. Most of the UFC charges are, by definition, a scam. See that line item on your phone bill? Think it goes to the FCC? WRONG. It goes right in Bell's pocket "to offset the costs of providing universal services."
To be accurate, TiVo has not (and never will) released the source to the non-GPL'd stuff that makes the Tivo what it is... the kernel modules and "tivoapp". The source to all the GPL components and the entire linux kernel running on the tivo are available.
To be fair, it's the pre-processor that eats the comments, not the compiler. (or have the monkeys changed that in the years I haven't been watching?)
His claims aren't entirely trollish... a 6000x increase in throughput would indicate a Really F**king Broken Network (tm). It doesn't take many dropped packets to kill throughput. Some simple tweaks to the backoff timer(s) goes a long way to fixing that -- I did just that many years ago to address broken VJ header compression on a netopia router.
The "holy grail" is in finding the maximum packet flow rate for a given connection. This depends on a number of constantly moving targets. Packets are dropped due to errors (rare) and queue overflows (very common.) In the case of errors, immediate retransmittion is best. But for overflows, immediately retransmitting the packet is very likely to get it dropped again; of course, there's an ICMP message code for this ("source quench"), but nobody uses it.
Therein is the problem... the people making the legal decissions do not have sufficent technical qualifications to sit in judgement. It's not easy to say what's common and what's inovative, unique code. At any rate, Linux has a lengthy history. Almost every line of code can be traced to when it was introduced and in many cases who put it there.
I recall a program once used at the NCSU computer science department to determine "cheating" on programming assignments. It was wrong ~100% of the time. It was actually rather comical.
InterNIC didn't start to enforce it until it was far too late. Back in the day, people asked for what they wanted to use (or thought they eventually would) instead of the modern day hoarding of domain names... register everything that exists and hold it ransom.
In fact, when I registered troz.com, Network Solutions offered up all three in one package.
Ah, but they do increase. My dialup account has increased twice in the last three years -- a total of 5$/month. (multiply by thousands of customers.)
News of a serial spam killer would be a ray of sunshine in that dark world. As the saying goes, nobody would shed a tear.
Costs are hard to quantify...
How does one put a dollar sign on latency? Yes, spam consumes bandwidth requiring ISPs to have substaintially larger connections than otherwise necessary -- bandwidth is cheap, but it isn't free. However, the connections are wide enough that spam isn't taking available bits away from other users -- assuming a fair queue, the more users of the bandwidth, the more spread out a users packets become (instead of 3 packets between each of mine, there are 9 or 50.)
How much is one's time worth? And just how much time is spent downloading, qualifying as spam, and deleting email? Even using a 28.8 dialup connection, it takes just a few (annoying) minutes to fetch 30 or 40 messages. It takes a few more minutes to sift through them to weed out all the spam. I've heard of people spending hours per day in this process, but I just don't buy it; unless they are studying every message and bitching to everyone within 100miles, there's no way it can take that long from eyeball to trash.
The only things that are easy to calculate are the administrative costs: software, support for said software, server(s) to run said software, store and service email, and monetary compensation for administrators of said software and systems. And of course, there's the cost(s) of repairs when the spammer's spew disrupts services. Admin overhead amounts to very little of $10billion -- even if we blur the line between "email service" and "spam fighting."
.net, .com, .org... there's been almost zero distinction since around '97 or so. Companies (and even individuals) routinely snarf up foo.everything-they-can-aford. InterNIC only breifly enforced the distinction (or tried to.)
And being in the UK, I don't think Spamhaus really cares about some idiot lawyer-spammer(s) in the US. (Ok, maybe if they're bored.)
Well, technically, this would not be managed by sendmail. And I can do this for almost nothing on the router...
Go find a version of LaBrea and learn how it does it. (arp spoofing to deal with nimbda.) It's actually pretty damned evil and 105% effective.
Your network connection cost money. The time it takes to download their crap is time you cannot be downloading pr0n and warez.
Bandwidth "comsumed" in the same manner from your ISP (and every network between you and the spammer)... plus the server processing and storage space consumed holding that crap until you delete it.
And the cost to the ISP(s) in admin time maintain mail servers.
--
Case in point, without a dedicated spam filtering system (running mail marshal), no one would be able to access their email because the exchange server would be on fire. There were many hours lost to spam overloading everything and people wading through deleting it (and a great many complaining to helpdesk staff about it.)
The volume of traffic is more than most companies could support. Being the corp. office of a Telco/ISP, that's not really a problem. And in fact is more than most of our customers can support.
Because a great deal of the internet isn't configured the way it should. Most mail handlers will do some basic verification of who the sender claims to be, but that usually stops at making sure the domain exists ("HELO foo.com".) Very few check forward and inverse DNS.
And almost zero process the headers of the incoming spew (that's left to the user/admin to do after the message is received.) At any rate, once you've received the entire message, what's the point? You've already "paid" for it. Spammers go to unbeleivable lengths to randomize their spew to thwart pattern recognition systems.
I've been very tempted to build something like LaBrea to trap spammers. "You can send me that spam... at one character per hour, and I'm not listening."
These people piss me off...
Billion dollar industry... blah, blah, freakin' blah. Prove the damned numbers. Unlike RIAA and MPAA, no one is going to let spammers make up their own balance sheets. There are numerous reports world wide giving hard proof of the costs brought about by all the stupid spammers. The only people who stand to be finacially injured and unemployed (and unemployable after a background check) are the asses sending all the spam.
I'll see their billions and raise by trillions -- the costs of software development and administrator headaches addressing the problem of spam, software development and administrative overhead to block loopholes in internet protocols, ever increasing server and bandwidth needs to move, process, and store all this crap... SPAM is a very expensive problem with the burden everywhere but the spammer.
Laws are useless unless swiftly and strictly enforced. Speeding is illegal, but that hasn't made much of a dent.
Well, having delt with one spammer in particular (yes, listed by Spamhaus), I'll just say they are very open liars. This individual said -- and I wished I'd been recording the call "for quality assurance purposes" :-) -- Spamhaus was a company run by one of his competitors. We had to mute the phone for a few minutes. They insist they are not sending "spam" -- even tho' I have spam reports from every batch of crap they sent.
(650m + 987k + 231k) / 175m =~ 3.72lbs per computer with almost of that being lead. No computer in the world (that's still in use anyway) contains that much lead. CRT monitor certainly have a high lead content... radiation sheilding inside the walls of the picture tube. Computers contain a few ounces of lead in the solder. That's about the only place lead exists in there.
Cadmium? How fsckin' old was that study? No laptop in nearly a decade has used NiCad battery packs. Heck, even my pager has a radio shack NiMh "AA" in it.