It IS a relevant comparison. Most humans in the developed world eat oil (read fossil fuels). Indirectly maybe, but that's what they eat.
The PS2 eats fossil fuels (assuming coal power) so do they.
Work out how much fossil fuels it takes to get a steak and potato on your plate - from fertilizer, pesticides, transportation, building the tractors and farms etc.
Compare the caloric value of the steak and the caloric value of the oil used to grow and eventually put a steak on your plate.
I bet you eat more indirect calories in oil than you eat in steak.
In countries where people drink goat milk from goats that just wander around munching shrubs (or eat the goats themselves), the CO2 they exhale is part of the normal carbon cycle.
If you already have enough public IPv4 addresses and the NAT/proxy sort of stuff is acceptable, then the ISP could just as easily do IPv4 NAT. The IPv4 to IPv4 NAT stuff is likely to be more mature and reliable than the IPv6 to IPv4 stuff.
2) Your 6to4 gateway will have to do something very similar to NAT (keep track of connections, translating addresses) or proxying.
The IPv4 packets from the IPv4 only server are not going to reach your IPv6 only clients behind that 6to4 gateway.
They may reach the 6to4 gateway, but the gateway needs to do some NAT/Proxy style stuff.
Yes I know we are running out of IPv4 addresses. But the world doesn't appear to be in a hurry to move to IPv6.
But the Internet is currently IPv4, and doesn't look like it's moving to IPv6 in a hurry.
Currently it's hard to get an IPv6 only client to talk with an IPv4 only server. AFAIK you need an proxy/NAT with an IPv4 address in between.
And the last I checked there are lots of very popular servers that are IPv4 only. I'd be surprised if they all start supporting IPv6 before 2013 and more importantly are _reachable_ via IPv6 through major ISPs around the world.
Ironically what appears to be more likely is that there will be more use/abuse of NAT. Whether IPv4-IPv4 or IPv6-IPv4 NAT.
As it is IPv6 networks are as much part of the Internet as Novell IPX networks that are tunnelled through the Internet.
Dunno if you'd see this (haven't had a net connection for nearly a week).
Anyway reread my post - it does say NAT provides some security BUT with some caveats.
Try this yourself:
Get a "NAT router" with a ethernet "WAN" port and "LAN" ports. Let's assume that the LAN IP range is 192.168.1.0/24, and the router WAN IP is 4.1.1.1 and the LAN IP is 192.168.1.1.
Plug a machine to the WAN port to simulate the ISP. We call this "A"
Plug another machine to the LAN port to simulate your server, leave an open port listening say on 192.168.1.3:80. On the server ensure you have a default gateway to 192.168.1.1. This simulates your supposedly your protected Intranet server that has some internet access.
Now on machine A, add a route to 192.168.1.0/24 via WAN port IP (4.1.1.1).
If your NAT router is typical (and without a firewall), you'll find that on machine A you can connect to 192.168.1.3:80.
This would not be possible if it was a stateful firewall that just allowed outbound connections with no inbound possible.
Anyway, the rest of the IPv6 people posting replies to me seem to be assuming that we would be able to magically move away from IPv4 very quickly. Good luck with that.
I am not an astronaut, but is it really "stuff happens" or is it that the training/practice didn't stick enough?
Before I watched the video I was actually expecting her to have bumped it away by accident so I was a bit surprised to see that she actually pushed/pulled it away.
From the video it looks like she got distracted by the problem at hand (or was in a rush) and fell into the "Earth mode" of "rummaging through stuff".
I heard that astronauts in space get used to being able to put stuff in "midair" and having it stay there while they do something else - in "earth mode" you can't do that - since it will hit the ground soon after.
In "Earth mode", if you push a bag away to uncover the stuff you want to get at, it usually stops moving soon after.
In orbit, the bag tends to continue moving. So you'd have to: 0) Make sure you are secured in position - very important step:) 1) move stuff away and stop it at the new position you want it to be. 2) Let go and it will stay there (until you or someone bumps it).
To me she didn't lose control of the bag - she was in control of the bag. She momentarily lost awareness of the situation, and so miscontrolled the bag.
I bet there are a fair number who bought the game and still haven't played it, and an even larger number who bought the game and just played the single player story for a few minutes.
What did they play instead? Counter-Strike...
Must be thousands of kids in cybercafes who just click the "Play Counter-Strike" shortcut on the desktop and don't even realize that Half-Life has a single player game:).
1) You still need to use IPv4 if sites you need to use still don't support IPv6 or are unreachable from your network.
For example - say you have a machine without an IPv4 address at all. How would you access the following sites: mail.google.com www.windowsupdate.com security.ubuntu.com mail.yahoo.com
I can list more.
2) You still need NAT if you are using dynamic IPv4 addresses.
Why?
Imagine what happens if the ISP gives you public IP range 4.5.5.0/252
But you drop and reconnect and are given public IP range 4.6.6.0/252
How long will it take for your machine to realize that it's IP address, DNS server and default gateway settings are wrong?
3) You still need NAT even if you are using static IPv4 addresses
There is an IPv4 shortage, so you need NAT to share the address(es) you get from the ISP.
If you think we can ignore the IPv4 shortage by switching your machines to IPv6, see 1).
Lastly, saying that DNS problem still applies because of NAT is wrong. NAT devices could randomize port numbers, there is nothing about doing NAT that requires a NAT device to not randomize port numbers. It's just like BIND could have randomized port numbers like djbdns did, but it didn't, so whose fault was that?
In order for Joe Public's PC to talk to other computers on the Internet, it needs an address.
Joe's ISP's routers all have addresses of their own which are fairly fixed in practice. The ISP can't just change them and automatically expect the rest of the Internet to still be able to reach them.
Without NAT, Joe's PC needs addresses that belongs to Joe's ISP before it can talk to the rest of the Internet.
BUT before Joe's router is connected to the ISP, how does his router or PC know what address they should be using?
With the popular NAT stuff, Joe's PC can be given important stuff like DNS server, default gateway, IP address - all using RFC1918 addresses, way before Joe's router connects to the ISP.
And then stuff can work for Joe almost immediately after connection.
Joe would not have to wait for "dhcp renewal time" seconds, or "some other public IP update period" seconds, before his PC realizes that "Oh I'm supposed to be using this public IP address and this gateway".
In short, with the NAT system when Joe sees the "Internet" LED lit on his router, he knows that PCs connected to the router should be able to access the Internet - if they can't there is a problem somewhere.
With the "public IP" system, when Joe sees the "Internet" LED lit, if the PCs can't access the internet it doesn't mean there is a problem or there isn't a problem. He has to wait a few minutes first (timeouts, renewals etc). Go ask an ISP call center manager how much a few minutes of waiting costs.
Maybe to you that's "almost the same thing", but to me it's not.
"because you tell it which set of IP addresses is your local subnet? Firewalls aren't magic..."
Wrong answer.
With the current NAT router+ISP stuff, Joe Public at the most needs to provide the username+password. And in some ISP configs, Joe doesn't even need to provide that- they just plug it in and it works "like magic" - and the sort of magic that Joe Public barely notices.
So either you'll keep getting router advertisements on your network indefinitely, or your computers will have to keep requesting for it (instead of eventually giving up- which is what happens now).
Next question: What url does Joe Public enter on his browser to get to the router config page, so that he can enter the username and password in order to get access to the ISP's network?
Without a NAT, how does a "NoNAT router" know what public IP range to give via DHCP (or other means) to Joe User's WinXP/Mac box, BEFORE it manages to get that public IP range from the ISP?
A public IPv4/IPv6 range that needs to be preconfigured on the router, is one more thing for the ISP and router manufacturers to deal with and one more thing for Joe User to screw up or have trouble with.
Go think about that.
My guess is you can't be a fucking moron since you're a slashdotter.
The difference between a "NAT router" and a "stateful firewall with public IP stuff behind"
You need the NAT working in order to reach the stuff behind it.
You don't need the stateful stuff working in order to reach the stuff behind it.
So in event of bugs, the hacker is more likely to have to work harder to exploit the stuff behind a NAT.
Now the issue with "just NAT" is the ISP can usually access the stuff behind the NAT - just as long as they know what IP range you have behind- they just have to get IP packets with dest=your.private.ip to your NAT device and _typically_ it will pass it through (some NAT devices also have a stateful firewall so they may not pass it through).
This means a 3rd party could get past your NAT if they have control over your ISP's routers route tables. But if they achieve that control you're probably screwed anyway.
Anyway, it's good enough protection, the hackers and malware bunch hardly do direct network attacks anymore against Joe User, much easier to convince Joe User to run stuff:).
1) I'm not in university 2) Which universities are allowed access? I'll find it interesting and amazing if somehow all universities worldwide have access.
As it is, such results are useless and a big waste of time to me.
"There are lot of topics, such as disability rehabilitation and linguistics, that I can't search for without Google returning a bunch of results from sites that require a subscription "
To me that's a breach of Google's own guidelines.
Here are Google's guidelines:
'Make pages primarily for users, not for search engines. Don't deceive your users or present different content to search engines than you display to users, which is commonly referred to as "cloaking."'
I've actually reported some of those "subscriber only" sites to Google, but not surprised that nothing much happened - since I suspect Google gets $$$ from them, and the unwritten guidelines is don't deceive users unless you pay us $$$:).
As Google's user, I very rarely want to get search results for content that I can't access. If they want that feature, at least I should be allowed to opt in/out much like their "safesearch".
So much for Google's don't be evil eh?
You should try search.yahoo.com and search.live.com once in a while to see if they are better. So far they are about as good as Google. If Google becomes worse I have no qualms about switching.
A game becomes crap if the players get to buy the referee/umpire or the rule makers.
Why is regulation so often bad?
Go figure what happens if voters keep voting for politicians mainly because those politicians got the most money from companies... Hint: you are selecting for politicians who are more likely to be bought/buyable.
Q: How many free market economists does it take to change a lightbulb? A: Free market economists don't change lightbulbs - they continue writing their papers in the dark and wait for Adam Smith's Invisible Hand to do it.
I get about 0.08 milliseconds ping between two PCs via a gigabit switch on my home network. That's for 64 byte packets. For 1472 byte packets it's about 0.15 ms. Doubt my el-cheapo switch is a "cut through".
7200RPM = 8.3 milliseconds per rev. So average latency due to rotation = 4ms. 15000 RPM will give you about 2ms.
Latency limits the maximum transactions per second - since when you commit you have to write to disk.
Bursts could be faster with battery backed up ram, but there might not be enough RAM for sustained TPS.
If the SSDs write transaction speeds are/become much faster, many people would be very interested in them.
Slashdot Mirror
Stupid people outnumber the smart people.
In a democracy don't be surprised if the politicians pander to the stupid.
It IS a relevant comparison. Most humans in the developed world eat oil (read fossil fuels). Indirectly maybe, but that's what they eat.
The PS2 eats fossil fuels (assuming coal power) so do they.
Work out how much fossil fuels it takes to get a steak and potato on your plate - from fertilizer, pesticides, transportation, building the tractors and farms etc.
Compare the caloric value of the steak and the caloric value of the oil used to grow and eventually put a steak on your plate.
I bet you eat more indirect calories in oil than you eat in steak.
In countries where people drink goat milk from goats that just wander around munching shrubs (or eat the goats themselves), the CO2 they exhale is part of the normal carbon cycle.
But for "industrialized nations" that's not true.
1) You still need a public IPv4 address.
If you already have enough public IPv4 addresses and the NAT/proxy sort of stuff is acceptable, then the ISP could just as easily do IPv4 NAT. The IPv4 to IPv4 NAT stuff is likely to be more mature and reliable than the IPv6 to IPv4 stuff.
2) Your 6to4 gateway will have to do something very similar to NAT (keep track of connections, translating addresses) or proxying.
The IPv4 packets from the IPv4 only server are not going to reach your IPv6 only clients behind that 6to4 gateway.
They may reach the 6to4 gateway, but the gateway needs to do some NAT/Proxy style stuff.
Yes I know we are running out of IPv4 addresses. But the world doesn't appear to be in a hurry to move to IPv6.
Maybe so.
But the Internet is currently IPv4, and doesn't look like it's moving to IPv6 in a hurry.
Currently it's hard to get an IPv6 only client to talk with an IPv4 only server. AFAIK you need an proxy/NAT with an IPv4 address in between.
And the last I checked there are lots of very popular servers that are IPv4 only. I'd be surprised if they all start supporting IPv6 before 2013 and more importantly are _reachable_ via IPv6 through major ISPs around the world.
Ironically what appears to be more likely is that there will be more use/abuse of NAT. Whether IPv4-IPv4 or IPv6-IPv4 NAT.
As it is IPv6 networks are as much part of the Internet as Novell IPX networks that are tunnelled through the Internet.
Dunno if you'd see this (haven't had a net connection for nearly a week).
Anyway reread my post - it does say NAT provides some security BUT with some caveats.
Try this yourself:
Get a "NAT router" with a ethernet "WAN" port and "LAN" ports. Let's assume that the LAN IP range is 192.168.1.0/24, and the router WAN IP is 4.1.1.1 and the LAN IP is 192.168.1.1.
Plug a machine to the WAN port to simulate the ISP. We call this "A"
Plug another machine to the LAN port to simulate your server, leave an open port listening say on 192.168.1.3:80. On the server ensure you have a default gateway to 192.168.1.1. This simulates your supposedly your protected Intranet server that has some internet access.
Now on machine A, add a route to 192.168.1.0/24 via WAN port IP (4.1.1.1).
If your NAT router is typical (and without a firewall), you'll find that on machine A you can connect to 192.168.1.3:80.
This would not be possible if it was a stateful firewall that just allowed outbound connections with no inbound possible.
Anyway, the rest of the IPv6 people posting replies to me seem to be assuming that we would be able to magically move away from IPv4 very quickly. Good luck with that.
I am not an astronaut, but is it really "stuff happens" or is it that the training/practice didn't stick enough?
:)
Before I watched the video I was actually expecting her to have bumped it away by accident so I was a bit surprised to see that she actually pushed/pulled it away.
From the video it looks like she got distracted by the problem at hand (or was in a rush) and fell into the "Earth mode" of "rummaging through stuff".
I heard that astronauts in space get used to being able to put stuff in "midair" and having it stay there while they do something else - in "earth mode" you can't do that - since it will hit the ground soon after.
In "Earth mode", if you push a bag away to uncover the stuff you want to get at, it usually stops moving soon after.
In orbit, the bag tends to continue moving. So you'd have to:
0) Make sure you are secured in position - very important step
1) move stuff away and stop it at the new position you want it to be.
2) Let go and it will stay there (until you or someone bumps it).
To me she didn't lose control of the bag - she was in control of the bag. She momentarily lost awareness of the situation, and so miscontrolled the bag.
I bet there are a fair number who bought the game and still haven't played it, and an even larger number who bought the game and just played the single player story for a few minutes.
:).
What did they play instead? Counter-Strike...
Must be thousands of kids in cybercafes who just click the "Play Counter-Strike" shortcut on the desktop and don't even realize that Half-Life has a single player game
I'm curious - how long would it take for the clients to get new IPs?
.here for free private use just like RFC1918.
Secondly "router." isn't a reserved TLD. So what RFC compliant TLD should be used?
Many years ago I personally tried convincing ICANN etc to reserve
But they didn't listen - maybe it's because I didn't give them lots of $$$.
1) You still need to use IPv4 if sites you need to use still don't support IPv6 or are unreachable from your network.
For example - say you have a machine without an IPv4 address at all. How would you access the following sites:
mail.google.com
www.windowsupdate.com
security.ubuntu.com
mail.yahoo.com
I can list more.
2) You still need NAT if you are using dynamic IPv4 addresses.
Why?
Imagine what happens if the ISP gives you public IP range 4.5.5.0/252
But you drop and reconnect and are given public IP range 4.6.6.0/252
How long will it take for your machine to realize that it's IP address, DNS server and default gateway settings are wrong?
3) You still need NAT even if you are using static IPv4 addresses
There is an IPv4 shortage, so you need NAT to share the address(es) you get from the ISP.
If you think we can ignore the IPv4 shortage by switching your machines to IPv6, see 1).
Lastly, saying that DNS problem still applies because of NAT is wrong. NAT devices could randomize port numbers, there is nothing about doing NAT that requires a NAT device to not randomize port numbers. It's just like BIND could have randomized port numbers like djbdns did, but it didn't, so whose fault was that?
In order for Joe Public's PC to talk to other computers on the Internet, it needs an address.
Joe's ISP's routers all have addresses of their own which are fairly fixed in practice. The ISP can't just change them and automatically expect the rest of the Internet to still be able to reach them.
Without NAT, Joe's PC needs addresses that belongs to Joe's ISP before it can talk to the rest of the Internet.
BUT before Joe's router is connected to the ISP, how does his router or PC know what address they should be using?
With the popular NAT stuff, Joe's PC can be given important stuff like DNS server, default gateway, IP address - all using RFC1918 addresses, way before Joe's router connects to the ISP.
And then stuff can work for Joe almost immediately after connection.
Joe would not have to wait for "dhcp renewal time" seconds, or "some other public IP update period" seconds, before his PC realizes that "Oh I'm supposed to be using this public IP address and this gateway".
In short, with the NAT system when Joe sees the "Internet" LED lit on his router, he knows that PCs connected to the router should be able to access the Internet - if they can't there is a problem somewhere.
With the "public IP" system, when Joe sees the "Internet" LED lit, if the PCs can't access the internet it doesn't mean there is a problem or there isn't a problem. He has to wait a few minutes first (timeouts, renewals etc). Go ask an ISP call center manager how much a few minutes of waiting costs.
Maybe to you that's "almost the same thing", but to me it's not.
"because you tell it which set of IP addresses is your local subnet? Firewalls aren't magic..."
Wrong answer.
With the current NAT router+ISP stuff, Joe Public at the most needs to provide the username+password. And in some ISP configs, Joe doesn't even need to provide that- they just plug it in and it works "like magic" - and the sort of magic that Joe Public barely notices.
That's why it's far from "almost the same thing".
So either you'll keep getting router advertisements on your network indefinitely, or your computers will have to keep requesting for it (instead of eventually giving up- which is what happens now).
Next question: What url does Joe Public enter on his browser to get to the router config page, so that he can enter the username and password in order to get access to the ISP's network?
Without a NAT, how does a "NoNAT router" know what public IP range to give via DHCP (or other means) to Joe User's WinXP/Mac box, BEFORE it manages to get that public IP range from the ISP?
A public IPv4/IPv6 range that needs to be preconfigured on the router, is one more thing for the ISP and router manufacturers to deal with and one more thing for Joe User to screw up or have trouble with.
Go think about that.
My guess is you can't be a fucking moron since you're a slashdotter.
Actually it is.
:).
The difference between a "NAT router" and a "stateful firewall with public IP stuff behind"
You need the NAT working in order to reach the stuff behind it.
You don't need the stateful stuff working in order to reach the stuff behind it.
So in event of bugs, the hacker is more likely to have to work harder to exploit the stuff behind a NAT.
Now the issue with "just NAT" is the ISP can usually access the stuff behind the NAT - just as long as they know what IP range you have behind- they just have to get IP packets with dest=your.private.ip to your NAT device and _typically_ it will pass it through (some NAT devices also have a stateful firewall so they may not pass it through).
This means a 3rd party could get past your NAT if they have control over your ISP's routers route tables. But if they achieve that control you're probably screwed anyway.
Anyway, it's good enough protection, the hackers and malware bunch hardly do direct network attacks anymore against Joe User, much easier to convince Joe User to run stuff
Could be something like this:
http://blog.wired.com/defense/2008/05/army-removes-pa.html
http://www.rense.com/general37/skull.htm
http://www.fas.org/sgp/othergov/dod/vts.html
"One application of V2K is use as an electronic scarecrow to frighten birds in the vicinity of airports."
Scarecrow for birds? They really had to stretch to come up with something more innocuous eh? ;).
1) I'm not in university
2) Which universities are allowed access? I'll find it interesting and amazing if somehow all universities worldwide have access.
As it is, such results are useless and a big waste of time to me.
"There are lot of topics, such as disability rehabilitation and linguistics, that I can't search for without Google returning a bunch of results from sites that require a subscription "
To me that's a breach of Google's own guidelines.
Here are Google's guidelines:
'Make pages primarily for users, not for search engines. Don't deceive your users or present different content to search engines than you display to users, which is commonly referred to as "cloaking."'
In 2006 they blacklisted BMW for breaching them:
http://news.bbc.co.uk/1/hi/technology/4685750.stm
I've actually reported some of those "subscriber only" sites to Google, but not surprised that nothing much happened - since I suspect Google gets $$$ from them, and the unwritten guidelines is don't deceive users unless you pay us $$$ :).
As Google's user, I very rarely want to get search results for content that I can't access. If they want that feature, at least I should be allowed to opt in/out much like their "safesearch".
So much for Google's don't be evil eh?
You should try search.yahoo.com and search.live.com once in a while to see if they are better. So far they are about as good as Google. If Google becomes worse I have no qualms about switching.
"Except we're not in a free market"
But quite a number seem to claim that you're in a free enough market for companies to buy politicians and legislators ;).
The way I see it, there MUST be regulation.
The issue is not whether you get less or more regulation, the issue whether you get good or bad regulation.
You are more likely to get bad regulation if the companies to be regulated get to create or influence the creation of regulation.
See: http://www.rense.com/general33/fd.htm
A game becomes crap if the players get to buy the referee/umpire or the rule makers.
Why is regulation so often bad?
Go figure what happens if voters keep voting for politicians mainly because those politicians got the most money from companies... Hint: you are selecting for politicians who are more likely to be bought/buyable.
Q: How many free market economists does it take to change a lightbulb?
A: Free market economists don't change lightbulbs - they continue writing their papers in the dark and wait for Adam Smith's Invisible Hand to do it.
"Your call is important to us"
*music*
<voice style=advert> NASA: For the Benefit of All... </voice>
I believe they also export cheese and pork in various forms.
In fact, Danish cheese is the only Danish stuff I can recall buying in the past few years.
I get about 0.08 milliseconds ping between two PCs via a gigabit switch on my home network. That's for 64 byte packets. For 1472 byte packets it's about 0.15 ms. Doubt my el-cheapo switch is a "cut through".
7200RPM = 8.3 milliseconds per rev. So average latency due to rotation = 4ms.
15000 RPM will give you about 2ms.
Latency limits the maximum transactions per second - since when you commit you have to write to disk.
Bursts could be faster with battery backed up ram, but there might not be enough RAM for sustained TPS.
If the SSDs write transaction speeds are/become much faster, many people would be very interested in them.
Latency might still be better with 15krpm drives.
If latency is important it makes sense to have those drives, even if throughput is good enough with 7200rpm drives.
Nuclear reactors + sea water = fresh water.
Easier to fight the battle against the sea when the sea is mainly in one direction.
Lot more expensive when you are surrounded.
If it's so simple you can afford to live underwater too.
There are advantages to living underwater - the waves and hurricanes won't cause as much problems.
Seems like McD is moving quick:
http://inventorspot.com/articles/mcdonalds_japan_goes_nobrand_with_quarter_pounder_shops_19505