4 ms·

Let's not act like using a "probably not in-use IPv4 but we can't really be sure" is a crime against humanity. If you're designing any kind of large scale syste

by simias 9y ago

Let's not act like using a "probably not in-use IPv4 but we can't really be sure" is a crime against humanity. If you're designing any kind of large scale system over the internet you end up hitting the problem sooner or later (like how some VPN solutions started using 5.x.y.z to be sure not to clash with LAN IPs for instance). The real solution of course would be to switch to IPv6 where any vendor can claim some private address anywhere without any realistic risk of collision but we all know that we're not ready for that yet. By their own admission CF receives a ridiculous amount of garbage traffic at this IP, it was not absurd for AT&T engineers in the past to thing "well, we need an IP that we can be reasonably sure nobody is going to use and is never going to conflict with anything on any network, 1.1.1.1 seems reasonable". Seeing everybody in this thread jumping into conspiracy theories instead of the much more likely configuration issue is a bit disappointing for a community that's supposed to understand technology.

arbitrage 9y ago
> it was not absurd for AT&T engineers in the past to thing That is an utterly unreasonable conclusion. The same logic resulted in Y2K, which was generally a huge waste of time, money, and resources. The same logic has resulted in the anemic adoption of IPv6, which is NOT a correct solution, because it doesn't work properly for large swaths of the public. The correct answer always was, and will continue to be, to use internal routes for internal routing, and external routes for external. Clashes with your LAN? Too fucking bad. This sort of pushing out of externalizes onto the customer results in the same exact outcome anyway: customer gets screwed. The customer always gets screwed. Don't rationalize the incompetence of engineers who should know better, and corporate execs who don't give a fuck.
denkmoon 9y ago
Would you feel the same way if AT&T started blasting static on "unused" frequencies they don't have the right to? Do that and the FCC will come down on you like a tonne of bricks, and I feel that is absolutely justified.
eeeeeeeeeeeee 9y ago
It's not reasonable. We have RFCs for a reason, which define which subnets can be used for public use, and which can be used internally. This has been written down for a long time and anyone working at ATT that can make these kind of decisions should know better.

by simias 9y ago

arbitrage 9y ago

> it was not absurd for AT&T engineers in the past to thing That is an utterly unreasonable conclusion. The same logic resulted in Y2K, which was generally a huge waste of time, money, and resources. The same logic has resulted in the anemic adoption of IPv6, which is NOT a correct solution, because it doesn't work properly for large swaths of the public. The correct answer always was, and will continue to be, to use internal routes for internal routing, and external routes for external. Clashes with your LAN? Too fucking bad. This sort of pushing out of externalizes onto the customer results in the same exact outcome anyway: customer gets screwed. The customer always gets screwed. Don't rationalize the incompetence of engineers who should know better, and corporate execs who don't give a fuck.

denkmoon 9y ago

Would you feel the same way if AT&T started blasting static on "unused" frequencies they don't have the right to? Do that and the FCC will come down on you like a tonne of bricks, and I feel that is absolutely justified.

eeeeeeeeeeeee 9y ago

It's not reasonable. We have RFCs for a reason, which define which subnets can be used for public use, and which can be used internally. This has been written down for a long time and anyone working at ATT that can make these kind of decisions should know better.