Aaah, the cube... I still have mine, and it's still chugging away. I've had multiple laptops come and go (#%^%^## gravity...) but the cube soldiers on. I sure do wish someone would make a pull-out upgrade to either a G5 or a core duo for it...
It was a single semester in high school, and it didn't really cover enlightenment history -it was more the story of the Mormon pioneers and their journeys.
Interestingly, I did take a semester of "Utah History" there, and a couple of friends of mine who went to school in Virginia said that there was a state curriculum there as well. Perhaps it's just Maryland...;)
You're right - your words were "not entirely unlike cruel and unusual punishment" (p29), which I had misremembered as torture. Sorry about the misquote.
I don't put a whole lot of faith in Shimv6, because it moves complexity from the network onto the hosts (i.e. a host with a single interface still has to have some element of routing knowledge), and I don't think that's a good long-term solution. Consider the case of a company with 30 webservers, which now all need to maintain N shim addresses each. In v4, each server has one address, and the network figures out how to get there. With Shim addressing, each server is topologically independant of the others - how do you do loadbalancing, etc when the IP address is not deterministic? I'm not saying that this couldn't be resolved, but from a network elegance point of view it adds a tremendous amount of ugliness and complexity, and makes something which was easy hard.
I'm not opposed to geographic aggregates, but what about those large companies with lots of offices? Or a medium size government office (US Forest service, for instance)? If I have two v6Net connections in LAX and NYC, do I need two independent address blocks? If so, how is that not more consumptive of space than just tying blocks to ASNs? I suspect far more enterprises have this type of geographical layout than are suspected.
Fair point on the complexity issue - it's not the IPv6 is necessarily MORE complex than IPv4, but rather that it mandates rethinking a whole lot of well-established approaches for a relatively nebulous benefit. The multihoming issue is a great example - what enterprises want to do is have two connections and provider-independent space. They understand how this behavior works today, and for those people who already have PI space, it works great. What's their incentive to rethink their whole routing strategy?
Yeah, ND is better than Arp, so that's a plus, but it's a marginal plus, because Arp does work pretty well most of the time...
What cracks me up is the US government mandate to migrate - they're advocating for v6 only as of 2008 (yeah, right), and given that all of the government agencies I've seen have their own PI address space, there's precious little benefit in them renumbering their own little non-Internet islands -most already use application-layer gateways, and don't WANT end-to-end reachability- other than compliance with the mandate. Sigh.
"Broken" was my take on the discussion of manually configuring DNS server addresses - I believe your phrase was "not entirely dissimilar from torture" - and that the autoconfiguration didn't include a way for a host to automatically learn the server addresses (thus prompting DHCPv6, etc). The original poster was saying that autoconfiguration would remove the need for DHCP etc, and that doesn't seem to be the case yet - I apologize if I overstated the case or misrepresented your opinion.
My theory on multihoming is this: give everyone with an ASN a/32. Make the requirements for an ASN include multihoming, and then we're golden. There are well under 40K unique ASNs in the wild today, and their growth is vastly slower than routing table growth, so if we can have more direct prefix ASN mapping, we can conserve the routing table better. I was in the minority before on this (and probably still am, but I hope not...)
I certainly think that we need to be concerned about running out of address space eventually, but I think that IPv6 as currently implemented tries to solve a lot of other problems as well, and I favor the guidance of RFC 1925 in the matter:
It is always possible to aglutenate multiple separate problems into a single complex interdependent solution. In most cases this is a bad idea.
brrr - IPv6 NAT: isn't that the worst of both worlds? complicated AND not-end-to-end?
I'm assuming you've never actually used IPv6, because you'd know that v6-autoconfiguration doesn't provide rather useful things like DNS server addresses - manually configuring those is "not entirely unlike torture," according to Van Bejinum (runningipv6.net).
What you're probably seeing is this: most things running IPv6 today are running dual stacks, and get the DNS server addresses through v4 DHCP. DHCPv6 isn't fully baked yet, but it's coming. So chalk one up for progress! We'll get to have more complicated configurations, and still require the same number of servers! excellent...
The list above thoroughly distorts the "benefits" of IPv6 - this list has become a troll which shows up during every debate. I challenge the author or anyone else to actually show how to configure all of those things.
For information about how broken routing is, take a look at NANOG - enterprises can no longer multihome.
For information about how broken autoconfiguration is, take a look at Running IPv6 by Iljitsch van Beijnum.
For information about how broken IPv6 is with regard to speed of routing and transmission, look at cisco - most IPv6 is software-forwarded, as opposed to hardware forwarded.
The other items in the list are things which IPv4 does AT LEAST as well as v6 (yeah, try getting AES-256 to work with IPv6 on an existing VAM2, without using IPv4 anywhere, and then talk to me about IPSec-v6...)
There are good and bad things about the protocol, but it's NOT the greatest thing since sliced bread, and that list is a heap of garbage.
DNS is just a technology, and it works pretty well.
The IETF's motto is "rough consensus and running code." When you've got a running-code alternative which does not use DNS, but still provides the hostname IP address mapping as effectively as DNS does, then I'll be interested. Until then, being "anti-DNS" is roughly like being "Anti-IP": not going to get you very far.
I always thought that it was related to "briar" - I had assumed that "Br'er Rabbit" and "Br'er Fox" were a rabbit and fox which lived in briar.
I first encountered Br'er Rabbit as folktales read to me by my Grandmother zt"l and my Aunt - but Wikipedia proves me wrong: thanks for the tip! I did hear my Aunt use the phrase just two days ago (having just gotten a job offer to write publications on an obscure area of tax law in which she worked for 15 years...), so at least I came by it honestly...
"Skin me Brer Fox," says he. "Snatch out my eyeballs, tear out my ears by the roots," says he, "But please, Brer Fox, don't fling me in that briar patch, " says he.
Why wouldn't netflix just open up their porn-rental service (netf--ks, perhaps?) I would think that the model of not going into a brick store would specifically appeal to the the porn consumer...
True story: I had a (gov) customer today tell me that they're going to have to deploy IPv6 in not just their DMZ/Internet edge, but also their private Intranet WAN. Why? Because that's what they're being told to do. None of the Intranet devices actually speak IPv6, and there has been exactly 0 demand for it there, but hey, they're your tax dollars at play...:/
Put the kool-aide down, and step away from pitcher...
IPv6 basically promises to be the "final 7337 protocol to rule them all" and tries really hard to roll all these functions into a single protocol. The Problem is that we've been there, and done that: think IPX.
IPv6 thoroughly breaks the idea of layering protocols, and in the spirit of trying to do too many things, accomplishes far too little.
Consider the fact that multihoming doesn't work any more, and the solution to the multihoming problem is Shim6, which is host-based instead of network-based (yuck!). That won't scale to any serious number of people - it's an N^2 problem rather than N(N-1)/N^2, the way multihoming today is.
Before you tell me about any more of these functions you listed above, let me ask: have you, yourself, personally used any of them? No? I didn't think so - most of the IPv6 stuff is vaporware, and it'll continue to be until the standards get worked out. Really, IPv6 is a solution in search of a problem.
If you wouldn't mind, please refrain from attempting to describe my internal mental state - your descriptions are very likely to be quite wrong (and in this case are thoroughly so). Thanks.
a "permissive attitude on homosexuality" is the opposite of a "restrictive attitude on homosexuality." Those are descriptive statements, not value judgements. One set of opinions might be more palatable than the other, but they are both descriptive (think "red hair" or "blonde hair").
"Homophobia" really *IS* just a buzzword - any rational understanding of the meaning of the word does not describe anyone I know, and I know a large number of people with varied attitudes toward homosexuality (from very right-wing religious folks to the gay couple around the corner).
With regard to "Insanity" which I said was not an appropriate word to describe Card or his beliefs, here's what Mirriam Webster's Medical Dictionary has to say (from dictionary.com):
Main Entry: insanity Pronunciation: in-'san-&t-E Function: noun Inflected Form: plural -ties 1 a : a deranged state of the mind usually occurring as a specific disorder (as schizophrenia) and usually excluding such states as mental retardation, psychoneurosis, and various character disorders b : a mental disorder 2 : such unsoundness of mind or lack of understanding as prevents one from having the mental capacity required by law to enter into a particular relationship, status, or transaction or as removes one from criminal or civil responsibility
Now exactly does that apply to Card, his work, or his belief systems? Remember, what I said which provoked your "You are using "sanity" in its most facile, oversimplified sense, which is not useful for any understanding of in/sanity - it's only good for vilification" response was the following sentence:
That you cannot relate to something does not make it insane. Perhaps what you mean to say is "I don't understand Card, and I think his opinions suck."
The English language is a powerful tool for communication. It is not necessary to use hyperbolic phrases to convey distaste.
Slashdot Mirror
Insightful, thank you.
Although I'd think the most useful bit of worked metal would actually be tongs...
Aaah, the cube... I still have mine, and it's still chugging away. I've had multiple laptops come and go (#%^%^## gravity...) but the cube soldiers on. I sure do wish someone would make a pull-out upgrade to either a G5 or a core duo for it...
sigh. and while I'm dreaming, could I get a pony?
It was a single semester in high school, and it didn't really cover enlightenment history -it was more the story of the Mormon pioneers and their journeys.
put the pipe down...
(I agree that this is a bad idea, but not for tinfoil-hat reasons)
Interestingly, I did take a semester of "Utah History" there, and a couple of friends of mine who went to school in Virginia said that there was a state curriculum there as well. Perhaps it's just Maryland... ;)
this is MY deal
quit beating a dead horse!
You're right - your words were "not entirely unlike cruel and unusual punishment" (p29), which I had misremembered as torture. Sorry about the misquote.
I don't put a whole lot of faith in Shimv6, because it moves complexity from the network onto the hosts (i.e. a host with a single interface still has to have some element of routing knowledge), and I don't think that's a good long-term solution. Consider the case of a company with 30 webservers, which now all need to maintain N shim addresses each. In v4, each server has one address, and the network figures out how to get there. With Shim addressing, each server is topologically independant of the others - how do you do loadbalancing, etc when the IP address is not deterministic? I'm not saying that this couldn't be resolved, but from a network elegance point of view it adds a tremendous amount of ugliness and complexity, and makes something which was easy hard.
I'm not opposed to geographic aggregates, but what about those large companies with lots of offices? Or a medium size government office (US Forest service, for instance)? If I have two v6Net connections in LAX and NYC, do I need two independent address blocks? If so, how is that not more consumptive of space than just tying blocks to ASNs? I suspect far more enterprises have this type of geographical layout than are suspected.
Fair point on the complexity issue - it's not the IPv6 is necessarily MORE complex than IPv4, but rather that it mandates rethinking a whole lot of well-established approaches for a relatively nebulous benefit. The multihoming issue is a great example - what enterprises want to do is have two connections and provider-independent space. They understand how this behavior works today, and for those people who already have PI space, it works great. What's their incentive to rethink their whole routing strategy?
Yeah, ND is better than Arp, so that's a plus, but it's a marginal plus, because Arp does work pretty well most of the time...
What cracks me up is the US government mandate to migrate - they're advocating for v6 only as of 2008 (yeah, right), and given that all of the government agencies I've seen have their own PI address space, there's precious little benefit in them renumbering their own little non-Internet islands -most already use application-layer gateways, and don't WANT end-to-end reachability- other than compliance with the mandate. Sigh.
-David
My theory on multihoming is this: give everyone with an ASN a
I certainly think that we need to be concerned about running out of address space eventually, but I think that IPv6 as currently implemented tries to solve a lot of other problems as well, and I favor the guidance of RFC 1925 in the matter:
brrr - IPv6 NAT: isn't that the worst of both worlds? complicated AND not-end-to-end?
-David
I'm assuming you've never actually used IPv6, because you'd know that v6-autoconfiguration doesn't provide rather useful things like DNS server addresses - manually configuring those is "not entirely unlike torture," according to Van Bejinum (runningipv6.net).
What you're probably seeing is this: most things running IPv6 today are running dual stacks, and get the DNS server addresses through v4 DHCP. DHCPv6 isn't fully baked yet, but it's coming. So chalk one up for progress! We'll get to have more complicated configurations, and still require the same number of servers! excellent...
The list above thoroughly distorts the "benefits" of IPv6 - this list has become a troll which shows up during every debate. I challenge the author or anyone else to actually show how to configure all of those things.
For information about how broken routing is, take a look at NANOG - enterprises can no longer multihome.
For information about how broken autoconfiguration is, take a look at Running IPv6 by Iljitsch van Beijnum.
For information about how broken IPv6 is with regard to speed of routing and transmission, look at cisco - most IPv6 is software-forwarded, as opposed to hardware forwarded.
The other items in the list are things which IPv4 does AT LEAST as well as v6 (yeah, try getting AES-256 to work with IPv6 on an existing VAM2, without using IPv4 anywhere, and then talk to me about IPSec-v6...)
There are good and bad things about the protocol, but it's NOT the greatest thing since sliced bread, and that list is a heap of garbage.
-David
DNS is just a technology, and it works pretty well.
The IETF's motto is "rough consensus and running code." When you've got a running-code alternative which does not use DNS, but still provides the hostname IP address mapping as effectively as DNS does, then I'll be interested. Until then, being "anti-DNS" is roughly like being "Anti-IP": not going to get you very far.
I always thought that it was related to "briar" - I had assumed that "Br'er Rabbit" and "Br'er Fox" were a rabbit and fox which lived in briar.
:)
I first encountered Br'er Rabbit as folktales read to me by my Grandmother zt"l and my Aunt - but Wikipedia proves me wrong: thanks for the tip! I did hear my Aunt use the phrase just two days ago (having just gotten a job offer to write publications on an obscure area of tax law in which she worked for 15 years...), so at least I came by it honestly...
this telling reads:
"Skin me Brer Fox," says he. "Snatch out my eyeballs, tear out my ears by the roots," says he, "But please, Brer Fox, don't fling me in that briar patch, " says he.
So thank you for making me look that up.
-David
"please don't throw me in that br'er patch, br'er fox!"
Bwaaa!
You're right - there's no market for porn at all. Now, there IS a market for V1@gr..a, but that's another story.
Why wouldn't netflix just open up their porn-rental service (netf--ks, perhaps?) I would think that the model of not going into a brick store would specifically appeal to the the porn consumer...
Green Tea may be healthful, but let me direct you to a discussion of the chemical composition of coffee and tea
BWAAAA! I haven't wished I had mod points this much in a long time...
and maybe we can keep the net free...
I appreciate the sentiment, thanks :)
:/
True story: I had a (gov) customer today tell me that they're going to have to deploy IPv6 in not just their DMZ/Internet edge, but also their private Intranet WAN. Why? Because that's what they're being told to do. None of the Intranet devices actually speak IPv6, and there has been exactly 0 demand for it there, but hey, they're your tax dollars at play...
Put the kool-aide down, and step away from pitcher...
IPv6 basically promises to be the "final 7337 protocol to rule them all" and tries really hard to roll all these functions into a single protocol. The Problem is that we've been there, and done that: think IPX.
IPv6 thoroughly breaks the idea of layering protocols, and in the spirit of trying to do too many things, accomplishes far too little.
Consider the fact that multihoming doesn't work any more, and the solution to the multihoming problem is Shim6, which is host-based instead of network-based (yuck!). That won't scale to any serious number of people - it's an N^2 problem rather than N(N-1)/N^2, the way multihoming today is.
Before you tell me about any more of these functions you listed above, let me ask: have you, yourself, personally used any of them? No? I didn't think so - most of the IPv6 stuff is vaporware, and it'll continue to be until the standards get worked out. Really, IPv6 is a solution in search of a problem.
-David
whoa there pardner - take a deep breath...
;)
and remember there are decaffinated brands with just as much flavor as the real thing*
-David
* - Val Kilmer in Real Genius
Have you ever read the Larry Niven & Stephen Barnes short story "The Locusts"? It's in Limits which is one of his better collections.
The premise is similar, but it takes it a bit further (if you haven't read it I don't want to spoil it for you...)
-David
See also the use in The President's Analyst - there it's by "TPC... The Phone Company"
-David
a "permissive attitude on homosexuality" is the opposite of a "restrictive attitude on homosexuality." Those are descriptive statements, not value judgements. One set of opinions might be more palatable than the other, but they are both descriptive (think "red hair" or "blonde hair").
"Homophobia" really *IS* just a buzzword - any rational understanding of the meaning of the word does not describe anyone I know, and I know a large number of people with varied attitudes toward homosexuality (from very right-wing religious folks to the gay couple around the corner).
With regard to "Insanity" which I said was not an appropriate word to describe Card or his beliefs, here's what Mirriam Webster's Medical Dictionary has to say (from dictionary.com):
Now exactly does that apply to Card, his work, or his belief systems? Remember, what I said which provoked your "You are using "sanity" in its most facile, oversimplified sense, which is not useful for any understanding of in/sanity - it's only good for vilification" response was the following sentence:
The English language is a powerful tool for communication. It is not necessary to use hyperbolic phrases to convey distaste.