Not in IPv4 or IPv6. I believe that to handle truly portable addressing requires a whole new way to think about addressing that IPv6 simply didn't try to do.
The problem of multi-homing in integrated in the design of both IPv4 and IPv6. The flaw is in the address concept itself. To fix this, you cannot just retrofit something on top of the existing IPv6. I do have an idea I call "layered addressing". It pretty much eliminates the core routing tables (it would most likely be way fewer than 1000 entries, perhaps just 200). But it also requires a whole new way to think about addresses. It has some similarities to "loose source routing", but works on the basis of autonomous secured zones. And that just isn't part of the IPv6 design. I highly doubt the multi6 working group has the authority to scrap the whole IPv6 addressing scheme and start over, so there would be no point in trying to do anything in that group.
So do you know where to reach the IPv7 working group?
Have you tried hiring a new CEO for your rapidly growing company lately? So far almost all that can be found are total idiots that have a resume of dot.com failures. Then when you do find someone who knows his stuff and didn't have to execute a bankruptcy, and knows how to go to Wall Street to line of your next $350 million of 2nd round financing, they end up wanting half a million a year and 50% of all the stock options the company has. Do these guys think they run the company or something? Just hire some bum off the street.
It's even rarer to find someone that is a designer, and a coder, and willing to start with someone else's design.
Designers tend to be creative people, and creative people for the most part want to build something new from scratch. What this means is you had damned well better keep your original designer. If the original guy leaves, find out why and correct that reason. It might be money in which case you better be paying more now that you learned your lesson. Whatever it is, if people leave for some other job, that means the other job was considered better than the job you now have open by at least the one good person you used to have and it would not be a surprise if a lot more thought so. If you want to make it easier to find good people, then you better make your job more attractive to good people than other managers competing for the same good people. There isn't (and never was) a shortage of people for the really good jobs worth having.
There's use for your style, but not on large from-the-ground-up systems that run a business. But they may need you to come in and fix what the system failed to do as a quick fix (before the client goes off to someone else to get a better system).
That's why the design can be re-thought during the development cycles. Methodologies like Extreme Programming describe how this is done. I suggest you read it (even though I don't necessarily suggest adopting it in whole). There's nothing wrong with changing the design. But it does need to be thought out carefully every time it is worked on, and it needs to be a clean, clear, well specified design.
BTW, those "pointy-haired-clients who know nothing about software and have deep pockets that will keep your company in business" will also dump your ass fast, and send in the lawyers, if your software does not do the job and do it right. If they lose money, or even come up short on projections, you can be sure they will come to collect it from you.
IPv6 is fundamentally flawed. It has the same fundamental flaw that IPv4 has. That flaw is that it does not support universally portable IP space. Just like IPv4, IPv6 requires a massive routing table space to be able to route to different address spaces. The only advantage of IPv6 over IPv4 is more addresses. It is NOT going to provide you with your own portable address block.
The Internet is going to end up splitting into a commercial version and a free (as in speech) version, anyway, so who cares. The latter will never need more than the IPv4 space, so IPv6 just isn't needed.
It might be better suited as just a commonly known hash. Given some piece of data, such as your login password, reduce it to a number and index into pi that many bits in, and generate N bits of hash. Somehow I doubt it will be as fast (especially for large numbers) as say MD5. But it could be interesting because it can be extended quite easily to as many bits as you want.
Now if that algorithm were faster than MD5 for indexes through 2^128, or faster than SHA1 and RIPEMD160 for indexes through 2^160, then I think we might have a winner.
These things are not cheap to produce, at least the good ones. If they can be legally copied and traded all over the net, it will happen, and there will be no payback of all the costs involved in producing it. A song is one thing, but a major production is something else entirely. If copyright is dumped, you might as well dump the entire film industry and all the movie theatres that play them. Then there wouldn't be anything to trade and the need for bandwidth would vanish, and ISPs would fold. Oh wait, they're folding anyway. Nevermind.
The cost of producing the CD, printing the inserts, the jewel case (which many are trying to eliminate), packaging, bulk boxing, and shipping to the music stores is probably not more than about $3 per unit on the large scale the top 1000 are done with. The artist gets no more than about $1. The store adds a markup which if a normal markup amounts to about $1.
I do see some CDs produced and sold for $5 or less, and these are not overstocks or clearances. So where does the other $10 for a $15 CD go? Obviously some of that is marketing as the top 1000 CDs from the top labels by the top artists are heavily marketed all the time. But how much can that really be? $10 per unit? I don't think so.
$8 per CD would be a reasonable price. Of course free market economics is supposed to allow the supplier to set the price where they believe people will still buy it. And they do at $15 per unit. So in that economic sense, $15 is a reasonable price if enough people are willing to pay that price for it.
Why are you still sending your name queries to ICANN controlled root/gtld servers? You know you don't have to.
Of course there have been some failed efforts to create alternate roots. Part of the reason for such failures is because the people involved were small timers trying to think they were big corporations. Either way, profit motive was their downfall. It's not wrong to make money at something, but if you think profit first, product second, you're most likely not going to have either one. Once we have an alternate root started without a profit motive in its control (but not excluding businesses), then I think it can succeed.
Imagine, if you will, two separate networks. One is like these corporate suits are proposing. The other is like the internet used to be back in the day (e.g. no banner ads, no spam). Now how will these networks differ in terms of things like domain names (assuming this technology is still used)? That's right, they will be different name spaces. And what is so wrong with that? Separate name spaces would help separate the information network from the commercial network.
It's the end points that are failing. The "in-cum-bent-on-raping-the-customer" telcos can't get DSL deployed right, even though the technology is sound. And the cable companies are no better.
For starters, let's take a look away from the internet for a moment and see how another of the operations of these large corporations is working: tech support. For the most part it just totally sucks. And the bigger the company, the worse it gets. The telcos and cablecos are among the worst. If they can't get that operation right, they are hardly going to be able to get something as complex as the internet right. And they haven't yet.
Yes, the internet has problems. But you're not going to get problems solved by letting them do it. Sure, things like massive broadcasts over the net to millions of people, or maybe even a billion in a few years, are going to totally overload it. A good broadcast infrastructure needs to be deployed within the net. All the suits needs to do is ask the jeans to design it, supply the cash, and it will get done. But they won't have it that way because the whole issue comes down to just one thing: control.
He who controls the internet controls information. He who controls information controls the world.
This is one of the reasons why so many of us have been fighting the likes of Microsoft. I could care less if they make high quality software or totally shitty software (I happen to believe they fall somewhere in the middle). I oppose Microsoft only because of their attempt to control information. But they aren't the only ones with their eye on the big cloud. We need to watch out for the likes of AT&T (yeah, the guys that tried to squelch it all in the first place), AOL/TW, and others. Watch what they are doing with content and see what I really mean.
We need some kind of ranking system for these companies that shows how well they do what they do, and what risk they pose in terms of trying to control information.
I have a 5/70 and a 5/85. Slackware is on the 5/70 and Solaris is on the 5/85. Rabbit and tortoise. If it weren't for the fact that I use Solaris to test portability of my code, I wouldn't have it running on here. Actually I need to get me some more of these so I can get Debian, NetBSD, and OpenBSD going on Sparc, too.
So what if there's a little bit of quality loss from ripping the analog path. We already have imperfection due to lossy compression. Once the loss is incurred, it will be finite and no more will be lost in the digital path thereafter. The concern will be the two accumulated loses of both analog ripping and compression. What MP3 (and probably others) depend on losing tends to be what will be lost in analog anyway. So the compression will probably see easier to compress audio anyway. The biggest concern remaining will be quantization errors from resampling at an unsyncronized rate.
More often the problem lies with management that won't allow the engineers to carry out best practices. This is because the best practices involve things that take extra time. Since the sales people usually commit product delivery often even before the development department ever heard of it, management gets really cranky about delivery times. Quality just goes out the window because that isn't what sales committed the company to.
Let's rake some managers and marketers over the coals first.
Most likely they will soon, if they do not already, restrict port 25 outbound to just their own servers. Earthlink/Mindspring does it, and it substantially reduced the sourcing of relayed SPAM from their network. I believe Verizon will end up doing this, too, because what they are doing now will have only minimal effect or SPAM reduction, and when people start running their own mail servers, there will be the new exposure to having relays within their network (not everyone who wants to run a mail server to host a vanity domain will be running something secure). And it won't further their marketing goals that probably prompted this particular restriction.
I understand perfectly. I don't know what Verizon literally does at a given moment (I don't get any service from them, fortunately), but I do see the ways they generally do things, and it's not good. This is only the first step, and a bad misstep, too.
Hosting the domain with Verizon isn't always an option. Some people have their vanity e-mail addresses in other domains where they don't own the whole domain. For example I'm working on setting up an email forwarding system for the domain ham.org for ham radio operators. If you were a ham and wanted to use callsign@ham.org, you could receive your mail once the email address is registered, but you can't use it in the FROM: field for outbound mail as a Verizon customer. And it's not a domain you can host with them, either.
This is a perfectly valid story, given the complications it really does cause for people. Do you really want to let a company like Verizon manage your domain? I wouldn't.
His point was that you don't need to make an SMTP connection to some other mail server for sending outbound mail with your own private e-mail address. I just tested it on a mindspring dialup to be certain. It works fine.
There are two ways to send mail out with your own private e-mail address. Mindspring blocks one of them (the one most abused by spammers because it lets them do the relaying) but not the other. They may have volume throttling on their servers to keep any one IP address from sending more than some limit through at once. That would stop most spamming through their own servers.
You do need to check your attitude problem. While ckuhtz wasn't specifically addressing the point you made, he was pointing out the alternative you have for sending e-mail. You might have a preference for not using his solution, but there is no evidence whatsoever in his posting that he is ignorant about how SMTP and such work. In fact he seems rather well informed to me. Your personal flame on him was uncalled for. Lighten up. If you don't like being narrowed to using the ISP mail server for outbound mail, talk about that without flaming people. Or make the capitalist move.
You trace it back to the first party that won't cooperate. They are then responsible. If that was an anonymous remailing service, then they should have implemented and deployed a mechanism to prevent bulk UCE and/or had a means to track down who sent it if it violated their policy. OTOH, anonymous remailers are hardly ever the source of spam. It's usually an ISP so gung-ho to get more accounts that they don't care who or what. These are the guys that will need to pay.
Giving up privacy is not necessary if things are implemented correctly. What you'd give up is the ability to send huge volumes of mail privately, and the ability to send to anyone who does not want mail coming from private/anonymous sources.
Slashdot Mirror
Not in IPv4 or IPv6. I believe that to handle truly portable addressing requires a whole new way to think about addressing that IPv6 simply didn't try to do.
The problem of multi-homing in integrated in the design of both IPv4 and IPv6. The flaw is in the address concept itself. To fix this, you cannot just retrofit something on top of the existing IPv6. I do have an idea I call "layered addressing". It pretty much eliminates the core routing tables (it would most likely be way fewer than 1000 entries, perhaps just 200). But it also requires a whole new way to think about addresses. It has some similarities to "loose source routing", but works on the basis of autonomous secured zones. And that just isn't part of the IPv6 design. I highly doubt the multi6 working group has the authority to scrap the whole IPv6 addressing scheme and start over, so there would be no point in trying to do anything in that group.
So do you know where to reach the IPv7 working group?
Have you tried hiring a new CEO for your rapidly growing company lately? So far almost all that can be found are total idiots that have a resume of dot.com failures. Then when you do find someone who knows his stuff and didn't have to execute a bankruptcy, and knows how to go to Wall Street to line of your next $350 million of 2nd round financing, they end up wanting half a million a year and 50% of all the stock options the company has. Do these guys think they run the company or something? Just hire some bum off the street.
It's even rarer to find someone that is a designer, and a coder, and willing to start with someone else's design.
Designers tend to be creative people, and creative people for the most part want to build something new from scratch. What this means is you had damned well better keep your original designer. If the original guy leaves, find out why and correct that reason. It might be money in which case you better be paying more now that you learned your lesson. Whatever it is, if people leave for some other job, that means the other job was considered better than the job you now have open by at least the one good person you used to have and it would not be a surprise if a lot more thought so. If you want to make it easier to find good people, then you better make your job more attractive to good people than other managers competing for the same good people. There isn't (and never was) a shortage of people for the really good jobs worth having.
Try posting in HTML.
There's use for your style, but not on large from-the-ground-up systems that run a business. But they may need you to come in and fix what the system failed to do as a quick fix (before the client goes off to someone else to get a better system).
That's why the design can be re-thought during the development cycles. Methodologies like Extreme Programming describe how this is done. I suggest you read it (even though I don't necessarily suggest adopting it in whole). There's nothing wrong with changing the design. But it does need to be thought out carefully every time it is worked on, and it needs to be a clean, clear, well specified design.
BTW, those "pointy-haired-clients who know nothing about software and have deep pockets that will keep your company in business" will also dump your ass fast, and send in the lawyers, if your software does not do the job and do it right. If they lose money, or even come up short on projections, you can be sure they will come to collect it from you.
If there isn't time to create solid design documentation, then for any project longer than about 2 days, the project is fucked!
IPv6 is fundamentally flawed. It has the same fundamental flaw that IPv4 has. That flaw is that it does not support universally portable IP space. Just like IPv4, IPv6 requires a massive routing table space to be able to route to different address spaces. The only advantage of IPv6 over IPv4 is more addresses. It is NOT going to provide you with your own portable address block.
The Internet is going to end up splitting into a commercial version and a free (as in speech) version, anyway, so who cares. The latter will never need more than the IPv4 space, so IPv6 just isn't needed.
It might be better suited as just a commonly known hash. Given some piece of data, such as your login password, reduce it to a number and index into pi that many bits in, and generate N bits of hash. Somehow I doubt it will be as fast (especially for large numbers) as say MD5. But it could be interesting because it can be extended quite easily to as many bits as you want.
Now if that algorithm were faster than MD5 for indexes through 2^128, or faster than SHA1 and RIPEMD160 for indexes through 2^160, then I think we might have a winner.
These things are not cheap to produce, at least the good ones. If they can be legally copied and traded all over the net, it will happen, and there will be no payback of all the costs involved in producing it. A song is one thing, but a major production is something else entirely. If copyright is dumped, you might as well dump the entire film industry and all the movie theatres that play them. Then there wouldn't be anything to trade and the need for bandwidth would vanish, and ISPs would fold. Oh wait, they're folding anyway. Nevermind.
The cost of producing the CD, printing the inserts, the jewel case (which many are trying to eliminate), packaging, bulk boxing, and shipping to the music stores is probably not more than about $3 per unit on the large scale the top 1000 are done with. The artist gets no more than about $1. The store adds a markup which if a normal markup amounts to about $1.
I do see some CDs produced and sold for $5 or less, and these are not overstocks or clearances. So where does the other $10 for a $15 CD go? Obviously some of that is marketing as the top 1000 CDs from the top labels by the top artists are heavily marketed all the time. But how much can that really be? $10 per unit? I don't think so.
$8 per CD would be a reasonable price. Of course free market economics is supposed to allow the supplier to set the price where they believe people will still buy it. And they do at $15 per unit. So in that economic sense, $15 is a reasonable price if enough people are willing to pay that price for it.
Why are you still sending your name queries to ICANN controlled root/gtld servers? You know you don't have to.
Of course there have been some failed efforts to create alternate roots. Part of the reason for such failures is because the people involved were small timers trying to think they were big corporations. Either way, profit motive was their downfall. It's not wrong to make money at something, but if you think profit first, product second, you're most likely not going to have either one. Once we have an alternate root started without a profit motive in its control (but not excluding businesses), then I think it can succeed.
Imagine, if you will, two separate networks. One is like these corporate suits are proposing. The other is like the internet used to be back in the day (e.g. no banner ads, no spam). Now how will these networks differ in terms of things like domain names (assuming this technology is still used)? That's right, they will be different name spaces. And what is so wrong with that? Separate name spaces would help separate the information network from the commercial network.
It's the end points that are failing. The "in-cum-bent-on-raping-the-customer" telcos can't get DSL deployed right, even though the technology is sound. And the cable companies are no better.
For starters, let's take a look away from the internet for a moment and see how another of the operations of these large corporations is working: tech support. For the most part it just totally sucks. And the bigger the company, the worse it gets. The telcos and cablecos are among the worst. If they can't get that operation right, they are hardly going to be able to get something as complex as the internet right. And they haven't yet.
Yes, the internet has problems. But you're not going to get problems solved by letting them do it. Sure, things like massive broadcasts over the net to millions of people, or maybe even a billion in a few years, are going to totally overload it. A good broadcast infrastructure needs to be deployed within the net. All the suits needs to do is ask the jeans to design it, supply the cash, and it will get done. But they won't have it that way because the whole issue comes down to just one thing: control.
He who controls the internet controls information. He who controls information controls the world.
This is one of the reasons why so many of us have been fighting the likes of Microsoft. I could care less if they make high quality software or totally shitty software (I happen to believe they fall somewhere in the middle). I oppose Microsoft only because of their attempt to control information. But they aren't the only ones with their eye on the big cloud. We need to watch out for the likes of AT&T (yeah, the guys that tried to squelch it all in the first place), AOL/TW, and others. Watch what they are doing with content and see what I really mean.
We need some kind of ranking system for these companies that shows how well they do what they do, and what risk they pose in terms of trying to control information.
I have a 5/70 and a 5/85. Slackware is on the 5/70 and Solaris is on the 5/85. Rabbit and tortoise. If it weren't for the fact that I use Solaris to test portability of my code, I wouldn't have it running on here. Actually I need to get me some more of these so I can get Debian, NetBSD, and OpenBSD going on Sparc, too.
So what if there's a little bit of quality loss from ripping the analog path. We already have imperfection due to lossy compression. Once the loss is incurred, it will be finite and no more will be lost in the digital path thereafter. The concern will be the two accumulated loses of both analog ripping and compression. What MP3 (and probably others) depend on losing tends to be what will be lost in analog anyway. So the compression will probably see easier to compress audio anyway. The biggest concern remaining will be quantization errors from resampling at an unsyncronized rate.
More often the problem lies with management that won't allow the engineers to carry out best practices. This is because the best practices involve things that take extra time. Since the sales people usually commit product delivery often even before the development department ever heard of it, management gets really cranky about delivery times. Quality just goes out the window because that isn't what sales committed the company to.
Let's rake some managers and marketers over the coals first.
Most likely they will soon, if they do not already, restrict port 25 outbound to just their own servers. Earthlink/Mindspring does it, and it substantially reduced the sourcing of relayed SPAM from their network. I believe Verizon will end up doing this, too, because what they are doing now will have only minimal effect or SPAM reduction, and when people start running their own mail servers, there will be the new exposure to having relays within their network (not everyone who wants to run a mail server to host a vanity domain will be running something secure). And it won't further their marketing goals that probably prompted this particular restriction.
I understand perfectly. I don't know what Verizon literally does at a given moment (I don't get any service from them, fortunately), but I do see the ways they generally do things, and it's not good. This is only the first step, and a bad misstep, too.
Hosting the domain with Verizon isn't always an option. Some people have their vanity e-mail addresses in other domains where they don't own the whole domain. For example I'm working on setting up an email forwarding system for the domain ham.org for ham radio operators. If you were a ham and wanted to use callsign @ham.org, you could receive your mail once the email address is registered, but you can't use it in the FROM: field for outbound mail as a Verizon customer. And it's not a domain you can host with them, either.
This is a perfectly valid story, given the complications it really does cause for people. Do you really want to let a company like Verizon manage your domain? I wouldn't.
His point was that you don't need to make an SMTP connection to some other mail server for sending outbound mail with your own private e-mail address. I just tested it on a mindspring dialup to be certain. It works fine.
There are two ways to send mail out with your own private e-mail address. Mindspring blocks one of them (the one most abused by spammers because it lets them do the relaying) but not the other. They may have volume throttling on their servers to keep any one IP address from sending more than some limit through at once. That would stop most spamming through their own servers.
You do need to check your attitude problem. While ckuhtz wasn't specifically addressing the point you made, he was pointing out the alternative you have for sending e-mail. You might have a preference for not using his solution, but there is no evidence whatsoever in his posting that he is ignorant about how SMTP and such work. In fact he seems rather well informed to me. Your personal flame on him was uncalled for. Lighten up. If you don't like being narrowed to using the ISP mail server for outbound mail, talk about that without flaming people. Or make the capitalist move.
It's a socialist state, what do you expect?
You trace it back to the first party that won't cooperate. They are then responsible. If that was an anonymous remailing service, then they should have implemented and deployed a mechanism to prevent bulk UCE and/or had a means to track down who sent it if it violated their policy. OTOH, anonymous remailers are hardly ever the source of spam. It's usually an ISP so gung-ho to get more accounts that they don't care who or what. These are the guys that will need to pay.
Giving up privacy is not necessary if things are implemented correctly. What you'd give up is the ability to send huge volumes of mail privately, and the ability to send to anyone who does not want mail coming from private/anonymous sources.
True, but we get to smack people for not supporting it.
Change ISP. What town do you live in?
Using incompetent ISPs only encourages them.