I remain unconvinced. Consider Australian aborigines http://www.infoplease.com/spot/aboriginal1.html: isolated for 30,000 years, pop. 400,000 currently. In that population there are none that don't have European ancestry? From what I could glean from TFA they don't address this, nor isolated (LONG TERM ISOLATED) populations in New Guinea and the Amazon.
But "Chicago", which was to be named "Windows 4.0", was so late and had slipped so many times, that it was renamed "Windows 95" to force a 'drop-dead' ship date and encourage the troops.
As Samuel Johnson said: "Depend upon it, sir, when a man knows he is to be hanged in a fortnight, it concentrates his mind wonderfully."
Except, that is not how the lookups are done in practice.
Google for "Tony Li Variable Length addresses" in the group info.big-internet. You want the article of Jun 16 1994, 8:02 pm by Tony Li, Message-ID: 199406160722.AAA16646@lager.cisco.com
On to today, and modern silicon does deep packet inspection: copes with variable length and all sorts of other interesting things, and does it at wire rate.
ATM cells are 53 bytes long, of which 48 bytes are payload. See http://en.wikipedia.org/wiki/Asynchronous_Transfer _Mode. But, an ATM switch doesn't just look at the 5 byte header and send the cell on. It has to decide if it needs to delay, or drop the cell, depending on the quality of service on the associated service. Then there is the adaptation layer. While voice will be carried in AAL1, data is carried in AAL5, where cells are assembled into packets. Anyway, the latency is (53 bytes * 8 * circuit speed) + (decision time). The (decision time) for ATM switches tends to be much greater than modern IP switches. Also, if each switch in the chain enforces the circuit speed, then each switch imposes a serialisation delay, whereas in an IP network, the rate shaping (and associated latency burden) happens only at the edges. All intermediate switches pass packets at the trunk rate which has a latency burden of a few microseconds.
ATM has superior jitter characteristics owing to the very small size of its cells---not the fixed size of its cells. That is easily managed by increasing speed (for things like VoIP it only matters below 768kbps or thereabouts).
Latency arises from propagation delay, serialisation delay, queuing delay and switching delay. If I do a 64B ping it doesn't matter whether it is ATM or IP, the latency will be the same except for switching delay---where my IP routers have negligible delay and my ATM switches have significant delay. OK, that's implementation, but the point is that my real-life IP switches have superior latency performance to my real-life ATM switches.
During the IPv6 debate, as to what the addressing scheme should be, one proposal was for variable length addresses. Look up RFC 1347. Various arguments were raised about how horribly inefficient this would be vs fixed length addressing. At which point Tony Li posted pseudocode to route a fixed length address vs a variable length address and showed that they consumed identical router resources; took exactly the same time---the same number of processor cycles. Proof by counterexample. Fixed is not inherently better than variable. ATM (fixed cells) has not beaten packet (variable length) on speed or cost.
However, TUBA was contaminated with GOSIP, so regardless of its merits it was shelved.
SIP, which is the IETF way of doing VoIP (and is the basis for IMS, and lots of other Good Stuff) is fundamentally NAT unfriendly. Inside the SIP negotiation are commands that say "connect address x, port y to address p, port q". For this to work with NAT you need an Application Level Gateway (ALG) that looks inside the SIP packets and rewrites the addresses/ports to the NATed ones. Ugly, ugly, ugly. And fragile.
Enter SIPS: secure SIP, where the SIP conversation is encrypted for security. Bye-bye ALG. It can't see inside the encrypted stream, much less modify it.
For their cable customers, that would probably require deploying DOCSIS 3.0---which doesn't even (quite) exist yet. DOCSIS 1.0/1.1 and 2.0 only support IPv4. That is, the filters etc in the cable modem are IPv4 specific. You can filter on protocol type, which means you can permit/deny IPv6, but no more.
If Comcast want to provide any sort of filtering, which may be required for authentication/accounting as well as protection/control, they'll need DOCSIS 3.0 (which is supposed to support IPv6). Which requires new headend equipment, new modems, and new provisioning systems. Of course, you get other good things with DOCSIS 3.0, like >= 100Mbps symmetric bandwidth. But since that requires extra RF tuners in the modems, it's not something that can be done with a firmware upgrade.
It may be possible to firmware upgrade existing DOCSIS 1.1/2.0 modems to just support the enhanced IPv6 features of DOCSIS 3.0---I really have no idea at this stage.
They could always support 6in4 tunnels, which would be a good interim step, but that is hardly something that will work seamlessly for the average naive user.
Then there is the IPv6 support that will need to be added to all the cable/DSL routers. An opportunity to sell more kit I suppose...
Well, this guy has been around even longer, and seems to be doing OK. While he was with a major label (Capitol), it seems that the almost guaranteed ability to sell 100,000 copies of an album were not enough to keep him signed. See this article and wikipedia for some history.
He does record in his garage, tours, sells on-line (including via iTunes) and physical CDs through a small independent record label.
Did I come across as sympathetic? I'm not. At all.
Copyright is essentially a contract: We, the people, grant you, the artist, a limited monopoly over your creation in order to encourage you to create works for the benefit of us (the people).
We create a right, where none existed before (unlike your inalienable rights to life, liberty etc.). We do this in order to benefit us. Arguably, you might create art and literature anyway. You might sing, and dance, and draw, and paint, and write because it gives you pleasure to do so. Nevertheless, we have determined that our lives would be more greatly enhanced if we allowed you some encouragement. So, we allow you copyright, for a limited time. That's the deal: we give you copyright; you give us art and literature.
We, the people, are more than a little irritated that your soulless heirs and assigns, to wit, the corporations, are attempting to renegotiate the contract as it nears its expiry. Especially since they are offering nothing in return. We are not amused.
Disney has trademarked Mickey, which gives them rights over anyone trying to create any new Mickey Mouse stories (or merchandise), but would not prevent people reprinting the original stories when they were out of copyright.
It's a big site. He has a LOT of reviews. (2118 full album reviews, all written by himself). If you want to take the time to explore further you will find that he does think the Beatles are good. In fact, great. To quote from his introduction page,
Class A (formerly - 5 stars). An ideal band; sure enough, some of its output might be flawed, but it's the highest standard by which I judge everything else. It must meet such an awful lot of selective criteria that only three of the bands in existence (the Beatles, the Rolling Stones and the Who) have received this rating; Bob Dylan also passed the plank, although it was a tough choice for me - he's put out quite a solid load of stinkers. Still, the very inhumane strength of his 1965-66 records alone managed to get him through.
Never To Be Topped
[15 albums]
Go and buy this now. These are superalbums by the few really great and timeless masters of rock: the cream of the cream of the cream. These records set the highest standard for all those that follow them and they probably won't be superated by anybody, not in the nearest couple thousand years. And yes, maybe I'm a fool for giving the highest mark to five Beatles' albums in a row, but fifteen years of Beatle-listening haven't cured me of the attitude.
I guess you could say he likes the Beatles, and thinks they are great. And this is from someone who stumbled over their work without peer pressure that 'hey, these guys are great.'
Re:Would the Beatles have made it today?
on
How Songs Get Popular
·
· Score: 2, Informative
What, all of them? I would have thought that most would be using the default browser, which is Safari. Hence my surprise that Safari does not show up at all, if 3% of the hits are identified as from Macs.
Excellent idea. Alas, no such technology has thus far been forthcoming, nor is there any sign that one is imminent. This despite the very real financial (and other) incentives to develop it.
So, while waiting for a Deus ex machina, I'm all for pursuing other avenues, as well.
Just because the current system has been abused beyond all recognition does not mean that the concept of software patents is a bad thing.
Maybe not, but it is a data point supporting that argument.
If we want to go back to first principles, start by asking what the purpose of a 'patent' is. Then see if software patents achieve, or are ever likely to achieve that purpose.
Slashdot Mirror
I remain unconvinced. Consider Australian aborigines http://www.infoplease.com/spot/aboriginal1.html: isolated for 30,000 years, pop. 400,000 currently. In that population there are none that don't have European ancestry? From what I could glean from TFA they don't address this, nor isolated (LONG TERM ISOLATED) populations in New Guinea and the Amazon.
As Samuel Johnson said: "Depend upon it, sir, when a man knows he is to be hanged in a fortnight, it concentrates his mind wonderfully."
Google for "Tony Li Variable Length addresses" in the group info.big-internet. You want the article of Jun 16 1994, 8:02 pm by Tony Li, Message-ID: 199406160722.AAA16646@lager.cisco.com
On to today, and modern silicon does deep packet inspection: copes with variable length and all sorts of other interesting things, and does it at wire rate.
ATM cells are 53 bytes long, of which 48 bytes are payload. See http://en.wikipedia.org/wiki/Asynchronous_Transfer _Mode. But, an ATM switch doesn't just look at the 5 byte header and send the cell on. It has to decide if it needs to delay, or drop the cell, depending on the quality of service on the associated service. Then there is the adaptation layer. While voice will be carried in AAL1, data is carried in AAL5, where cells are assembled into packets. Anyway, the latency is (53 bytes * 8 * circuit speed) + (decision time). The (decision time) for ATM switches tends to be much greater than modern IP switches. Also, if each switch in the chain enforces the circuit speed, then each switch imposes a serialisation delay, whereas in an IP network, the rate shaping (and associated latency burden) happens only at the edges. All intermediate switches pass packets at the trunk rate which has a latency burden of a few microseconds.
ATM has superior jitter characteristics owing to the very small size of its cells---not the fixed size of its cells. That is easily managed by increasing speed (for things like VoIP it only matters below 768kbps or thereabouts).
Latency arises from propagation delay, serialisation delay, queuing delay and switching delay. If I do a 64B ping it doesn't matter whether it is ATM or IP, the latency will be the same except for switching delay---where my IP routers have negligible delay and my ATM switches have significant delay. OK, that's implementation, but the point is that my real-life IP switches have superior latency performance to my real-life ATM switches.
During the IPv6 debate, as to what the addressing scheme should be, one proposal was for variable length addresses. Look up RFC 1347. Various arguments were raised about how horribly inefficient this would be vs fixed length addressing. At which point Tony Li posted pseudocode to route a fixed length address vs a variable length address and showed that they consumed identical router resources; took exactly the same time---the same number of processor cycles. Proof by counterexample. Fixed is not inherently better than variable. ATM (fixed cells) has not beaten packet (variable length) on speed or cost.
However, TUBA was contaminated with GOSIP, so regardless of its merits it was shelved.
Fair point, but wrong example. localhost in IPv6 is ::1
Enter SIPS: secure SIP, where the SIP conversation is encrypted for security. Bye-bye ALG. It can't see inside the encrypted stream, much less modify it.
If Comcast want to provide any sort of filtering, which may be required for authentication/accounting as well as protection/control, they'll need DOCSIS 3.0 (which is supposed to support IPv6). Which requires new headend equipment, new modems, and new provisioning systems. Of course, you get other good things with DOCSIS 3.0, like >= 100Mbps symmetric bandwidth. But since that requires extra RF tuners in the modems, it's not something that can be done with a firmware upgrade.
It may be possible to firmware upgrade existing DOCSIS 1.1/2.0 modems to just support the enhanced IPv6 features of DOCSIS 3.0---I really have no idea at this stage.
They could always support 6in4 tunnels, which would be a good interim step, but that is hardly something that will work seamlessly for the average naive user.
Then there is the IPv6 support that will need to be added to all the cable/DSL routers. An opportunity to sell more kit I suppose...
He does record in his garage, tours, sells on-line (including via iTunes) and physical CDs through a small independent record label.
Death is too good for them.
Copyright is essentially a contract: We, the people, grant you, the artist, a limited monopoly over your creation in order to encourage you to create works for the benefit of us (the people).
We create a right, where none existed before (unlike your inalienable rights to life, liberty etc.). We do this in order to benefit us. Arguably, you might create art and literature anyway. You might sing, and dance, and draw, and paint, and write because it gives you pleasure to do so. Nevertheless, we have determined that our lives would be more greatly enhanced if we allowed you some encouragement. So, we allow you copyright, for a limited time. That's the deal: we give you copyright; you give us art and literature.
We, the people, are more than a little irritated that your soulless heirs and assigns, to wit, the corporations, are attempting to renegotiate the contract as it nears its expiry. Especially since they are offering nothing in return. We are not amused.
Disney has trademarked Mickey, which gives them rights over anyone trying to create any new Mickey Mouse stories (or merchandise), but would not prevent people reprinting the original stories when they were out of copyright.
And from the rating page
I guess you could say he likes the Beatles, and thinks they are great. And this is from someone who stumbled over their work without peer pressure that 'hey, these guys are great.'
George Starostin describes his introduction to the Beatles somewhat differently. As in someone really not exposed to them who now has very definite views.
http://www.deathpenaltyinfo.org/article.php?scid=8 &did=245
As they say: beware of programmers bearing screwdrivers.
What, all of them? I would have thought that most would be using the default browser, which is Safari. Hence my surprise that Safari does not show up at all, if 3% of the hits are identified as from Macs.
Odd statistics. Mac shows up as 3%, yet Safari does not rate at all in the browsers.
So, while waiting for a Deus ex machina, I'm all for pursuing other avenues, as well.
Mods---the "nuke from orbit" is a joking reference to Aliens. See http://www.moviequotes.com/archive/titles/22.html Modded Troll? Sheesh!
I hope your fallout shelter is really deep.
OTOH, maybe we have (briefly) a wonderful new energy source...
"It's yours"
"Here's your $10"
"That's $11"
"WTF?"
"I have to charge G.S.T."
"Then you should have said so"
That's all. If you have to charge G.S.T., then you have to include it in the stated price. Most casual sales don't have to charge G.S.T..
If we want to go back to first principles, start by asking what the purpose of a 'patent' is. Then see if software patents achieve, or are ever likely to achieve that purpose.
There you go then, your wish has been catered to.