4 ms·

> I want to emphasize that IP address leaks are a serious privacy problem. No, they are not. IP addresses aren’t meant to be private. Furthermore, in the world

> I want to emphasize that IP address leaks are a serious privacy problem. No, they are not. IP addresses aren’t meant to be private. Furthermore, in the world of CGNAT (and even more so with IPv6) it’s just a meaningless set of numbers.

tredre3 3y ago
IP addresses being meaningless to you is irrelevant to the discussion at hand. The job of a firewall is to not leak information. Little Snitch is leaking information. This is a problem. End of story.
lapcat 3y ago
> IP addresses aren’t meant to be private. That's precisely the problem. The internet wasn't originally designed with privacy in mind. It didn't even have encrypted connections. > Furthermore, in the world of CGNAT (and even more so with IPv6) it’s just a meaningless set of numbers. This feels like reality denial, a sticking of the head in the sand. IP addresses often can be and indeed have been used for the purposes of personal identification. This especially true if you have multiple sources of data and can cross-reference them. Regardless, if you have a network content filter extension rule that specifies "disallow all outgoing connections from process X", then you would expect that the developer of process X should not be able to detect that your computer is running process X.
- sgjohnson 3y ago
  If it’s really that important, you should have a hardware firewall on the edge with deny-by-default on outgoing. You cannot trust the security framework on the endpoint alone.
  - lapcat 3y ago
    A hardware firewall can't do per-process blocking. That's the whole point of Little Snitch. Besides, I wasn't soliciting advice, I was offering criticism. For some reason, a lot of people always confuse those two. Don't say "You cannot trust the security framework" if you really mean that the security framework should be exempt from criticism. And if you don't mean that, then the hardware firewall is a red herring. It's a red herring anyway, because it's not suitable for the use case.

tredre3 3y ago

IP addresses being meaningless to you is irrelevant to the discussion at hand. The job of a firewall is to not leak information. Little Snitch is leaking information. This is a problem. End of story.

lapcat 3y ago

> IP addresses aren’t meant to be private. That's precisely the problem. The internet wasn't originally designed with privacy in mind. It didn't even have encrypted connections. > Furthermore, in the world of CGNAT (and even more so with IPv6) it’s just a meaningless set of numbers. This feels like reality denial, a sticking of the head in the sand. IP addresses often can be and indeed have been used for the purposes of personal identification. This especially true if you have multiple sources of data and can cross-reference them. Regardless, if you have a network content filter extension rule that specifies "disallow all outgoing connections from process X", then you would expect that the developer of process X should not be able to detect that your computer is running process X.

sgjohnson 3y ago
If it’s really that important, you should have a hardware firewall on the edge with deny-by-default on outgoing. You cannot trust the security framework on the endpoint alone.
- lapcat 3y ago
  A hardware firewall can't do per-process blocking. That's the whole point of Little Snitch. Besides, I wasn't soliciting advice, I was offering criticism. For some reason, a lot of people always confuse those two. Don't say "You cannot trust the security framework" if you really mean that the security framework should be exempt from criticism. And if you don't mean that, then the hardware firewall is a red herring. It's a red herring anyway, because it's not suitable for the use case.