Separate names with a comma.
Discussion in 'Off Topic [BG]' started by colcifer, Jun 6, 2012.
That's about the long and short of it for me too.
I bet we'll get a lot of projects once networks stop working and no one knows how to configure them.
*tumbleweed rolls by*
so i guess its supposed to speed up the transfer of information from my IP?
Interesting, but surprisingly un-promoted/shoved down your throat IMO.
how are we going to remember IP addresses now?
Good luck phasing out all the machines that run on older OSs that don't support it. Aside from actually implementing it, that will be the biggest issue.
Too many security risks to implement in the business world is my opinion. We won't be moving to it here anytime soon. Maybe later.
How about enlightening those of us who have no idea what IPv6 is?
the primary difference is that ipv6 changes the standard from a 32 bit ip address to a 128 bit address.
ipv4 only supports 2^32 = ~4 billion different unique ip addresses. they have been exhausted for about 6 years now, and has required various hacks and workarounds like CIDR and NAT translation to expose sufficient unique addresses to users to enable connectivity to continue.
believe it or not, ipv6 has been around for a little while (~20 years), but because of necessity it is only recently that folks have been making it a priority to switch to it.
ipv6 supports 2^128 = 340,000,000,000,000,000,000,000,000,000,000,000,000 different unique ip addresses. to put this another way, ipv4 doesn't even have enough ip addresses for 1 per person on the planet, while ipv6 enables many many unique ip addresses for every person on the planet (think on the order of the number of grams of matter that make up the entire earth (~6 x10^27). then multiply it by 10, and that's how many static ip addresses would be available to every individual on the planet). of course many of these are reserved for certain applications (just like many addresses in ipv4 are reserved and therefore unavailable for general use) but the difference in address space is still staggering.
the reason for this massively increased address space isn't specifically because folks wanted to give every cluster of 1000 atoms a static ip, it was to speed up routing of packets (by many orders of magnitude). currently, routers in the internet use complex routing algorithms based on CIDR to route packets from "e-here" to "e-there", and these can be inefficient, especially if not properly implemented.
the intended address space (usable addresses) of ipv6 is 2^64, which is the square of the current address space - the internet times itself (sorta like every webpage on the current net being replaced by a copy of the entire internet) = the new ipv6 internet. all the extra information in the ipv6 ip address is intended to improve routing efficiency and speed up throughput.
ipv6 also supports much larger packet sizes - on the order of 4 gigabytes vs 64kilobytes with ipv4. larger acceptable packets mean faster communication from end to end, and it also means that routers do not perform packet fragmentation, removing another speed-soak from the pipeline. in fact, the entire routing proceedure is vastly simplified from ipv4, and the router's job duties are simplified immensely, removing the biggest speed bump to the internet (all the computing and recomputing the routers are required to do to route a packet)
ipv6 also was built with security in mind - ipsec was originally built into ipv6 and then built into existing ipv4 protocols.
What he said.
windows has supported ipv6 since xp sp1 (2002). mac os 10.3 and up (2003) have also supported ipv6. even linux has officially supported it since 2005.
the biggest hardware roadblock is going to be router/infrastructure overhaul more than end user machines and servers.
from everything i've learned regarding ipv6, it's substantially more secure, natively, than ipv4, but i seem to remember that this kind of thing is what you do so clue me in .
It's 2 IPv more,innit?
you know, it's a shame - it took implementing the global ip network (in ipv4) to learn sufficiently of the quirks and issues inherent in the implementation to actually develop something suitable for the network.
ipv4 was an amazing achievement, especially if you consider where computing was at the time it was developed. the final spec on ipv4 was 32 years ago - that was before the largescale availability of commercially viable personal computers, let alone any kind of on-demand network outside of industry, military and academic environments.
if you consider that the majority of the spec has remained unchanged since then, and it still has been flexible and powerful enough to support the www and every other network in the world, it's sorta like folks figured out a way to fly to the moon in a wright brothers' airplane.
but it is definitely time to switch on over to '6 .
the internet - listen to it. can you hear it?
you would if it were ipv6.
IIRC, IPv4 was designed when the internet was still experimental.