Just send them a bill for a reasonable amount, say $500 or so. They will know they did something wrong and will either pay or rectify. Stay courteous and don't threaten legal action of the bat. You're likely to make some money.
yet another article that's skeptical of how ready IPv6 is. The amount of brokenness that is there is not very big. Of all the people that have the full Internet (that is IPv4 *and* IPv6) most will simply connect to any IPv6 website without issues.
And apart from the fact that yahoo seems to be a US only thing, and even there is not so relevant anymore, I applaud them doing IPv6, when they get to it. (and after Google, Comcast, Akamai and many others)
I wish we'd get over this "brokenness" story and simply deploy and then fix it for the 1% that has issues. Would be nice if it gets rolled out to the point of 20% traffic in 2011, the year we'll run out of available IPv4 addresses.
Skype is popular because it can function even when both ends of the conversation are going through a NAT. NATs are inherently evil (no, they don't offer more protection than a simple ingress filter on any "real" router) and break the end-to-end principle. With IPv6, you will not get just 1 IP address like in IPv4, but a shitload (currently the thinking is a/48, which is over 65000 subnets, each subnet containing roughly 4 billion * 4 billion addresses, but thinking is changing towards/56's, only 256 subnets, still an impressive number of addresses).
What this will do it 2 things:
1- allow for more machines to do any particular service. (multiple VoIP devices, multiple webservers, no extra configuration to restore ICQ file transfers, easy webcamming)
2- reduce worm problems (because most addresses are not used, simply scanning address ranges will not be successful, limiting worm propagation by several orders of magnitude)
So, no, it doesn't remove the need for servers per se, but most applications that set up peer to peer connections these days need some other machine to bypass the NAT problem. With IPv6, NAT is no longer needed (although there will be idiots that think it makes things more secure and demand it for IPv6 as well. I'm praying it won't catch on)
That is the current situation. Verisign does not operate all.com servers. There are other companies that run slave servers for them, spread over the world. DNS clients (recursive servers, like the DNS server of your ISP that you use) round robin between the authoritative servers they know to get the fastest one and that's the one they'll stick with for a while.
I thought, the root servers themselves are authoritative for their entire toplevel domain
You are correct. They are authoritative for "." , the root zone. the.com zone is under that in the hierarchy. FQDN (fully qualified domain names) always end with a dot, to indicate the root. By convention, this is usually left out. So the FQDN for "slashdot.org" is "slashdot.org." , just try it in your browser, it will work.
So, I would automatically ask the server authoritative for bar.com about foo.com? Does not sound plausible -- either you are a guru or you are wrong:-)
w00t! So I'm a guru!:-) (you would ask the same server about foo.com as bar.com, that's the way that remains manageable. You don't have to re-ask every question from the root down to where you need to be. You can re-use the most specific answer you still have in your cache about authority)
The main difference would be that ICANN would maintain the master server - but that server would not be reachable by anything other than the replica DNS servers and the registrars. You could run that on a Athlon over a T1 line.
Much like it is today for the roots. DNS servers don't need massive cpu power anyway, they are more memory intensive.
The system you describe is active for the root. Verisign runs the master server for that too. And the other organisations get the zone from a hidden master. And even for.com... there were a few slave servers for.com that removed the sitefinder wildcard. A bad move, imo, because it works around the problem instead of solving it, like ICANN did. (then again, the Internet usually works around problems).
It simply comes down to wether or not you think it is easy to maintain the.com zone file and deal with all the registrars. You seem to think so. I think it is not.
Unlike DNSSEC, my plan only requires modifications to the root-servers' software, which, I'm sure, is already heavily modified and customized. And it does not have to happen in one instant either -- those servers can be updated gradually.
Not really, as the load on the root and tld servers is lessened by caching the information (for a ttl). So if you do a query for www.bar.com to a root server, you are getting a referral to the servers authoritative for.com. You would cache this information, so if you then want to go to www.foo.com, you will reuse this information and thus cannot go to the other.com server.
The only way in which it would work is if you set the TTL of the records extremely low, which would increase the level of traffic well over manageable levels.
Oh, and just so you know, most root servers run regular versions of Bind and NSD, not specific changed versions. My company runs a root server and it has no special software. That would increase the possibility of special bugs too much.
You mean the current setup? Where ICANN has a company (registry) manage the master server, which gets updated through registrars, who are multiple companies handling the registration of domain names under the particular TLD that the company/registry is chosen for. (in this case.com)
There is a reason that ICANN has outsourced the infrastructure management of the TLDs. It simply does not have the resources to do it itself, nor the legal background to handle ccTLDs.
I am not a DNS guru, but yes, you are totally wrong. DNS is embedded in so many machines and is already such a complex standard, that it is impossible to change something so fundamental and have it work. Even adding something like DNSSEC took more than 10 years (and some broken attempts) before we reached the standard (just happened).
Basically, it has to be 1 company. Since delegations are on label boundaries. There is no way to get around that, unless you want to develop a new DNS-alike protocol and get everyone to switch over to it. Oh, and if you, build security into it while you're at it!
Cue the comments about shortage of IP addresses in China and how the US has almost all of the addresses. This is simply not true.
Currently about a third of the IPv4 addresses are still available, with about a sixth of the addresses used up in the last 12 years, world wide.
Current estimates (based on historical usage) give us about 20 to 35 years before we run out of IPv4 addresses, although this can be much less if new technologies get widely adopted (china and india becoming as connected as the western world is, mobile phones all getting addresses, voip taking off)
Big companies that are not even making sure that new hardware with a 5 year or more lifespan is IPv6 capable (or firmware upgradable) are shortsighted. On the other hand, companies that feel they need IPv6 now, and don't use it to pump up their tech credibility (about the only thing you can do currently with IPv6 that can't be done otherwise) are throwing away money.
I know a lot of people say this is too big, to much this or that, but really, if you have over 20GB of storage, you are not really targeting the casual music listener (other than the gullible ones, who think bigger is always better), but people with an interest in having their music collections in a good quality with them.
Of course, the iPod doesn't support lossless compressed formats, but this is about a harddisk that could also be used by better audioplayer manufacturers.
Anyway, a record, ripped in good quality, or even lossless will run between 100MB and 300MB. Let's be conservative and say 150MB per album. That means that on this disc will have space for around 500 albums. (rounded down to be on the save side, if you have only mp3 playback this number might grow to be around 800-900 albums)
500 albums is a medium sized collection for music lovers. (and 800-900 is not excessive) Personally, I would really like to see players with 80GB that are small and have good battery life. I don't care for colour screens and video, image and other capabilities (apart maybe from recording or digital in/out) and I would really like to design a menu for a music player. (is it so hard to have different random modes: artist, album, year, genre? or the ability to schedule songs to play next without generating a playlist?)
Oh well, I guess I'm not a good target market, I want to control how I listen to the music I love...
For me, the Mac interface has always been an example of how not to do it. I want my menu's where my program window is, I don't want to drag cd's to the garbage to get them out, I don't want to have programs running after I click to close them.
I seem to be in the minority, but I don't like to work that way. So, unless they port KDE to the x86 OSX port, I won't be using it, no matter how streamlined it all works together.
Not all artists are prefab performers. Some are musicians, actually. There are plenty of artists I like that regularly completely overhaul the arrangements of songs, or play songs they haven't released yet, or play covers that you simply cannot hear studio versions of.
Besides, it also happens that the studio mix simply isn't very good but that live recordings sound astounding.
It all comes down to preference, live recordings can add excitement, overview of an artists growth and they can show you songs in a different light, as live artists are prone to let a song breath some more, giving it the room to blossom that simply isn't always possible in studio recordings!
Re:Also, there are more addresses!
on
The State of IPv6
·
· Score: 1
the myth about address shortage in Asia is long standing, but does not get truer the more people repeat it.
The RIR system makes sure everyone that needs addresses can get them (regardless of who has the legacy address space from when we used classes). So when Asia or Europe runs out, the US runs out as well. The only ones who have addresses left are the ones who have the old legacy classes and they are already slowly getting more and more pressure to give back unused networks.
IPv6 has more addresses, yes, but that is mostly on the end-user side of things. I can't wait to have more then 1 address that I can have visible from the outside world if I choose it to be so.
From an ISP perspective though, IPv6 is a bit of work (=money) with no current returns. Let's hope more people ask for it. Luckily, here in the Netherlands we already have 1 ADSL provider that can give a native IPv6 connection. With a/48, of course, so over 65000 subnets! Yay!
Personally, I prefer CoolMon as a small lightweight desktop display program. It is for W2K and XP mostly, with client/server options and a lot of plugins possible.
Actually, in trying to find something similar for Linux, I found Karamba, although I found it bit complicated to set up correctly.
It is strange that there are hardly any truly silent powersupplies, when there is almost no laptop with a powersupply that makes noise..
If powersupply manufacturers simply take out the heat generating part so that you can have it well ventilated and only let the dc wires go into the case, then the completely silent pc would be so much closer.
the current system of handing out addresses (which have no value as such and should not be charged for, although an administrative fee can be charged for setting it up, which quite a few ISP's do) is essentially global through RIPE NCC, APNIC, LACNIC and ARIN. So if China runs out, it means we have all run out.
Admittedly, the US has quite a bit of legacy space, but I'm sure that large chunks of it will be reclaimed for everyone, should the need arise.
In the US, the idea is still that the Internet is American, so the US will be ok. That is exactly why the ARIN region is (too) slow to pick up on IPv6.
Ofcourse, IPv6 may not happen in the end (there are still quite a few bugs to be ironed out by the IETF et al), but I hope it does, because NAT is getting old real fast. Port forwarding helps a little, but remains a hack at best. The pain of having several machines do the same things behind one IP address (ICQ, webserver, netmeeting) is simply not worth it when I can get over 65000 subnets (with billions of addresses in each one) assigned to me with IPv6.(Everyone that could subnet, should receive a/48 according to current policy, no extra charge)
We could then finally do all the things that we should have been able to now.
And currently, IPv6 is totally free. Everyone gives free transit to everyone, IPv6 is not taken into account with the fee that the RIR's charge their members (at least in the RIPE region, I think the other regions too).
This will change ofcourse, but IPv6 is already a major improvement over IPv4, the US will feel the pain of coming late everywhere if they don't prepare.
If Australian law can be applied to US citizens, it's likely (though I ain't anal) that US law would apply to those same citizens.
The US has no problem with applying US laws to citizens in other countries doing things that are legal in those countries (i.e. Sklyarov/Elcomsoft), so it would only make sense if other countries do the same.
In this case it would ofcourse relate to acts performed in Oz that are illegal under Oz law..
Slashdot Mirror
Just send them a bill for a reasonable amount, say $500 or so. They will know they did something wrong and will either pay or rectify. Stay courteous and don't threaten legal action of the bat. You're likely to make some money.
yet another article that's skeptical of how ready IPv6 is. The amount of brokenness that is there is not very big. Of all the people that have the full Internet (that is IPv4 *and* IPv6) most will simply connect to any IPv6 website without issues.
And apart from the fact that yahoo seems to be a US only thing, and even there is not so relevant anymore, I applaud them doing IPv6, when they get to it. (and after Google, Comcast, Akamai and many others)
I wish we'd get over this "brokenness" story and simply deploy and then fix it for the 1% that has issues. Would be nice if it gets rolled out to the point of 20% traffic in 2011, the year we'll run out of available IPv4 addresses.
Skype is popular because it can function even when both ends of the conversation are going through a NAT. NATs are inherently evil (no, they don't offer more protection than a simple ingress filter on any "real" router) and break the end-to-end principle. With IPv6, you will not get just 1 IP address like in IPv4, but a shitload (currently the thinking is a /48, which is over 65000 subnets, each subnet containing roughly 4 billion * 4 billion addresses, but thinking is changing towards /56's, only 256 subnets, still an impressive number of addresses).
What this will do it 2 things:
1- allow for more machines to do any particular service. (multiple VoIP devices, multiple webservers, no extra configuration to restore ICQ file transfers, easy webcamming)
2- reduce worm problems (because most addresses are not used, simply scanning address ranges will not be successful, limiting worm propagation by several orders of magnitude)
So, no, it doesn't remove the need for servers per se, but most applications that set up peer to peer connections these days need some other machine to bypass the NAT problem. With IPv6, NAT is no longer needed (although there will be idiots that think it makes things more secure and demand it for IPv6 as well. I'm praying it won't catch on)
CC
That is the current situation. Verisign does not operate all .com servers. There are other companies that run slave servers for them, spread over the world. DNS clients (recursive servers, like the DNS server of your ISP that you use) round robin between the authoritative servers they know to get the fastest one and that's the one they'll stick with for a while.
CC
You are correct. They are authoritative for "." , the root zone. the
w00t! So I'm a guru!
(you would ask the same server about foo.com as bar.com, that's the way that remains manageable. You don't have to re-ask every question from the root down to where you need to be. You can re-use the most specific answer you still have in your cache about authority)
CC
Much like it is today for the roots. DNS servers don't need massive cpu power anyway, they are more memory intensive.
The system you describe is active for the root. Verisign runs the master server for that too. And the other organisations get the zone from a hidden master. And even for
It simply comes down to wether or not you think it is easy to maintain the
CC
Not really, as the load on the root and tld servers is lessened by caching the information (for a ttl). So if you do a query for www.bar.com to a root server, you are getting a referral to the servers authoritative for
The only way in which it would work is if you set the TTL of the records extremely low, which would increase the level of traffic well over manageable levels.
Oh, and just so you know, most root servers run regular versions of Bind and NSD, not specific changed versions. My company runs a root server and it has no special software. That would increase the possibility of special bugs too much.
CC
You mean the current setup? Where ICANN has a company (registry) manage the master server, which gets updated through registrars, who are multiple companies handling the registration of domain names under the particular TLD that the company/registry is chosen for. (in this case .com)
There is a reason that ICANN has outsourced the infrastructure management of the TLDs. It simply does not have the resources to do it itself, nor the legal background to handle ccTLDs.
CC
I am not a DNS guru, but yes, you are totally wrong. DNS is embedded in so many machines and is already such a complex standard, that it is impossible to change something so fundamental and have it work. Even adding something like DNSSEC took more than 10 years (and some broken attempts) before we reached the standard (just happened).
Basically, it has to be 1 company. Since delegations are on label boundaries. There is no way to get around that, unless you want to develop a new DNS-alike protocol and get everyone to switch over to it. Oh, and if you, build security into it while you're at it!
CC
Cue the comments about shortage of IP addresses in China and how the US has almost all of the addresses. This is simply not true.
Currently about a third of the IPv4 addresses are still available, with about a sixth of the addresses used up in the last 12 years, world wide.
Current estimates (based on historical usage) give us about 20 to 35 years before we run out of IPv4 addresses, although this can be much less if new technologies get widely adopted (china and india becoming as connected as the western world is, mobile phones all getting addresses, voip taking off)
Big companies that are not even making sure that new hardware with a 5 year or more lifespan is IPv6 capable (or firmware upgradable) are shortsighted. On the other hand, companies that feel they need IPv6 now, and don't use it to pump up their tech credibility (about the only thing you can do currently with IPv6 that can't be done otherwise) are throwing away money.
CC
I know a lot of people say this is too big, to much this or that, but really, if you have over 20GB of storage, you are not really targeting the casual music listener (other than the gullible ones, who think bigger is always better), but people with an interest in having their music collections in a good quality with them.
Of course, the iPod doesn't support lossless compressed formats, but this is about a harddisk that could also be used by better audioplayer manufacturers.
Anyway, a record, ripped in good quality, or even lossless will run between 100MB and 300MB. Let's be conservative and say 150MB per album. That means that on this disc will have space for around 500 albums. (rounded down to be on the save side, if you have only mp3 playback this number might grow to be around 800-900 albums)
500 albums is a medium sized collection for music lovers. (and 800-900 is not excessive) Personally, I would really like to see players with 80GB that are small and have good battery life. I don't care for colour screens and video, image and other capabilities (apart maybe from recording or digital in/out) and I would really like to design a menu for a music player. (is it so hard to have different random modes: artist, album, year, genre? or the ability to schedule songs to play next without generating a playlist?)
Oh well, I guess I'm not a good target market, I want to control how I listen to the music I love...
CC
For me, the Mac interface has always been an example of how not to do it. I want my menu's where my program window is, I don't want to drag cd's to the garbage to get them out, I don't want to have programs running after I click to close them.
I seem to be in the minority, but I don't like to work that way. So, unless they port KDE to the x86 OSX port, I won't be using it, no matter how streamlined it all works together.
CC
relative to the two main US parties, it might be thought of as left wing, possibly though.
CC
Not all artists are prefab performers. Some are musicians, actually. There are plenty of artists I like that regularly completely overhaul the arrangements of songs, or play songs they haven't released yet, or play covers that you simply cannot hear studio versions of.
Besides, it also happens that the studio mix simply isn't very good but that live recordings sound astounding.
It all comes down to preference, live recordings can add excitement, overview of an artists growth and they can show you songs in a different light, as live artists are prone to let a song breath some more, giving it the room to blossom that simply isn't always possible in studio recordings!
the myth about address shortage in Asia is long standing, but does not get truer the more people repeat it.
/48, of course, so over 65000 subnets! Yay!
The RIR system makes sure everyone that needs addresses can get them (regardless of who has the legacy address space from when we used classes). So when Asia or Europe runs out, the US runs out as well. The only ones who have addresses left are the ones who have the old legacy classes and they are already slowly getting more and more pressure to give back unused networks.
IPv6 has more addresses, yes, but that is mostly on the end-user side of things. I can't wait to have more then 1 address that I can have visible from the outside world if I choose it to be so.
From an ISP perspective though, IPv6 is a bit of work (=money) with no current returns. Let's hope more people ask for it. Luckily, here in the Netherlands we already have 1 ADSL provider that can give a native IPv6 connection. With a
CC
Sounds like this is a perfect candidate for a machine that can be a VoIP gateway while still keeping backward compatibility with pots.
Hopefully the thing is powerful enough for a home with a few cheap SIP phones.
CC
Personally, I prefer CoolMon as a small lightweight desktop display program. It is for W2K and XP mostly, with client/server options and a lot of plugins possible.
Actually, in trying to find something similar for Linux, I found Karamba, although I found it bit complicated to set up correctly.
CC
It is strange that there are hardly any truly silent powersupplies, when there is almost no laptop with a powersupply that makes noise..
If powersupply manufacturers simply take out the heat generating part so that you can have it well ventilated and only let the dc wires go into the case, then the completely silent pc would be so much closer.
CC
the current system of handing out addresses (which have no value as such and should not be charged for, although an administrative fee can be charged for setting it up, which quite a few ISP's do) is essentially global through RIPE NCC, APNIC, LACNIC and ARIN. So if China runs out, it means we have all run out.
/48 according to current policy, no extra charge)
Admittedly, the US has quite a bit of legacy space, but I'm sure that large chunks of it will be reclaimed for everyone, should the need arise.
In the US, the idea is still that the Internet is American, so the US will be ok. That is exactly why the ARIN region is (too) slow to pick up on IPv6.
Ofcourse, IPv6 may not happen in the end (there are still quite a few bugs to be ironed out by the IETF et al), but I hope it does, because NAT is getting old real fast. Port forwarding helps a little, but remains a hack at best. The pain of having several machines do the same things behind one IP address (ICQ, webserver, netmeeting) is simply not worth it when I can get over 65000 subnets (with billions of addresses in each one) assigned to me with IPv6.(Everyone that could subnet, should receive a
We could then finally do all the things that we should have been able to now.
And currently, IPv6 is totally free. Everyone gives free transit to everyone, IPv6 is not taken into account with the fee that the RIR's charge their members (at least in the RIPE region, I think the other regions too).
This will change ofcourse, but IPv6 is already a major improvement over IPv4, the US will feel the pain of coming late everywhere if they don't prepare.
CC
Like someone has pointed out before, address space is a global resource, and about 30% is available.
Asia (through APNIC) can use these as well as the US (through ARIN) or Europe (through RIPE NCC) can.
If 1 region runs out, we all run out.
CC
Does this mean that in a few years I can copy my thinking, work from home and have the computer do my work?
Time to invest in a better couch!
CC
In this case it would ofcourse relate to acts performed in Oz that are illegal under Oz law..
CC