Hacker News Reader

Comment by daneel_w | Hacker News Reader

Happy "customer" here. I've been using their free 6in4 tunnel through OpenBSD for about five years and have had no mentionable problems. I configure mine solely with OpenBSD's network interface files, e.g. /etc/hostname.gif0:

  tunnel <my current IPv4> <HE's IPv4 endpoint>
  inet6 <my desired IPv6 address> 128 alias <HE's IPv6 gateway>
  !route -n add -inet6 default <HE's IPv6 gateway>

I use the connectivity to reach a cluster of VPSes in AWS deliberately set-up without public IPv4 addressing, which would otherwise represent a large part of the monthly costs because of buttholes like Jeff Bezos actively monetizing IPv4 address space.

View on HN

> because of buttholes like Jeff Bezos actively monetizing IPv4 address space.

IPV4 addresses are finite and rapidly being depleted. What other solution do you have to manage demand of a finite resource other than charging for it?

My stance is that common connectivity shouldn't cost an additional $3.70 a month on top of already egregious traffic costs. The price per IP today is about $30. The lifetime of the investment is infinite and upkeep is in the grand scheme of things nothing. The markup profit is insane. It's a new behavior, pure usury, seizing an opportunity to profit on a crisis. To offer some contrast (without getting into the sizes of their respective turfs) Oracle doesn't charge a dime.

We are in crisis precisely because nobody charged for IPv4 addresses in the past, and so overwhelming majority of those are wastefully allocated. What you want would exacerbate the crisis.

We're in this crisis because we failed to anticipate the explosive growth of the Internet. It took a bit into the 2000s until we stopped doling out generously oversized networks to everyone who asked. Vetting the need would've been the right requirement. Shutting the door for organizations with not enough money would've hampered progress.

Yes, and why did people ask for these oversized networks? That’s right, because addresses were free.

That's a depressingly shallow knee-jerk-y way of reasoning around something so fantastically open as the Internet... You're offering the deplorable solution of "let the money vote" instead of reason and restraint. The consequence if we had asked for money from the get-go would've been a corporate-ruled scenario where connectivity and Internet foothold were primarily in the hands of the businesses that had the most money. Smaller businesses, and non-profits in particular, would effectively have been shut out and innovation and growth in the Internet's most sensitive phase would have suffered greatly.

The fact of the matter is that “let the money vote” works much better than alternatives. “Reason and restraint” is precisely how we got to where we are.

Have you looked at the state of the world? "Let the money vote" works better than "let anyone just commandeer the finite resource, but only once", if nobody immoral enough to exploit it learns about vulnerabilities in your system.

If I can use my money to vote for "give me more money" at a profit, and I have no qualms about doing so, then I win – and if we play multiple such games with the same money, then we end up with a situation that's worse than "let anyone commandeer the resource".

The world is not black and white, and paid service can easily coexist with subsidized service. There are many examples of “it costs you $XX but ask us if you need it badly” policies. The best varieties probably do some degree of vetting (because otherwise would slightly defeat the point or make it impractical, especially while LLMs are cheap enough that anyone can use one to write a convincing tear-jerker) and have objective criteria.

I haven’t thought enough to say whether it makes sense for specific cases like IP address allocation, though.

Don't worry, we've learned nothing and will repeat the same mistake with IPv6:

https://news.ycombinator.com/item?id=42671847

https://www.theregister.com/2024/12/06/apnic_huawei_ipv6/

Reply on news.ycombinator.com