[mdlug] OT: the great IPv6 debate

[Aaron Kulkis]

Dan Pritts wrote:
> On Mon, Apr 26, 2010 at 07:07:36PM -0400, Aaron Kulkis wrote:
>> Dan Pritts wrote:
>>> the problem is not borked DNS; it is not at the server or at the client
>>> site.  The problem is some broken network, somewhere in the middle.
>>>
>>> Yes, someone somewhere screwed up; yes, they should fix it.  yes,
>>> i should complain (and i do!).  all that said, this is to be expected
>>> in this transitional phase.  Having application software that
>>> exacerbates the problem just makes users want to turn off v6.
>>>
>> And if IPv4 was screwed up in the middle, the same problems
>> would occur.  Otherwise, we could all just put random numbers
>> in, and networking would still work.
> 
> "this is to be expected in this transitional phase." 
> 
> back in 1989, IPv4 got screwed up in the middle a lot more than it
> does now.  We didn't have anything else then, but today, deploying
> ipv6, we have a pretty reliable ipv4 internet.

So we should just stick with IPv4 for another 3 years
until we have a crisis situation?

The solution is not to say "oh, the IPv6 doesn't work, lets
just be content with IPv4"... the solution is to tell the
ISP to get their IPv6 situation straightened out, or you will
not only change vendors, but you will also see them in
court for breech of contract, and you'll also be forming
a legal class for a class action suit (which distributes
your legal costs among all participants -- and many trial
lawyers will take on such a case for free if the class
turns out to be big enough).

Once it's cheaper to fix their IPv6 configuration than
to pay for lawyers, they'll fix their IPv6.