Spyke

Replies

linux

Comment on

Bazzite Auto Updates

This is a fair question to ask given recent events. I don't run Fedora currently, so others could probably give a much more exact answer, but from what I understand of it:

Bazzite is built on top of Fedora with uBlue. To compromise one of the packages, the attacker would have to bypass the Fedora enterprise team who are rage filled roid-driven experts who don't take kindly to that sort of thing. They heavily secure their stuff. Even if an attack was successful, it would have little lasting effect because of immutability and having access to easy rollbacks.

It's not impossible (like somehow stealing Bazzite's keys), but it's incredibly unlikely. AUR/NPM package sketchiness is not anywhere on the same level as compromising Fedora's keys.

Comment on

Are ISPs responsible for bots having residential IPs or is this a user problem?

I worked as a network analyst for a provider for several years and during that time I'd say ~90% of the issue stemmed from sketchy apps/services that the user loaded from their end.

A lot of "free" VPN services will basically allow bad actors (the paid tier) to use your connection. A lot of IoT devices are also just openly available on the Internet to route through.

From the ISP perspective, we managed the roads, not your car. There is a push to blame the ISP as it's their network, but realistically how are they meant to provide security (in the context that is being asked) to any device that gets plugged into that network? We even had business customers demand we add clauses to contracts where we would accept responsibility for any malware they sent between sites over an MPLS setup.

In the end, a lot of people seem to want this impossible scenario of the ISP managing security for them but also not inspecting their traffic.

Comment on

[META] Are paid for, closed source projects, being advertised on this community, appropriate?

At it's heart, this is what @selfhosted is meant for:

A place to share alternatives to popular online services that can be self-hosted without giving up privacy or locking you into a service you don't control.

I would say that members talking about paid/closed products they use (ex. "I connect to this via Tailscale" or "I use company ABC for hosted VPS") to accomplish something is fine, but marketing or job boarding (ex. "Looking for QA on my commercial product") is not.

Comment on

Are ISPs responsible for bots having residential IPs or is this a user problem?

Reply in thread

4G/5G cellular? So, in some ways you're actually easier to find. Your cell gateway is connecting to a tower which is logged and includes cell strength metrics. That gets compared to other towers and via trilateralization your location is determined.

Again, going back to what I previously said: there is a path back to you even if only for either billing or connectivity purposes.

Comment on

Are ISPs responsible for bots having residential IPs or is this a user problem?

Reply in thread

You mention crappy security practices from the ISP but then mention the user's action (installing "free" VPNs). Why is the ISP on the hook for the user making terrible decisions?

What is the correct security practice in that instance? Fire the customer for being an idiot? Maybe just DENY IP ANY ANY on outbound traffic?

How do you protect somebody who is intent on running themselves off a cliff?

Comment on

Are ISPs responsible for bots having residential IPs or is this a user problem?

Reply in thread

Ok, this I can answer personally as we did multiple cases of this happening (CSAM, bomb threats, etc) at work.

So, anonymity on the Internet is not actually a thing. Whether its an IP address or telecom switch or whatever, there is a path back to you even if only for either billing or connectivity purposes. So, for IP, we would receive a subpoena signed by a judge to hand over any and all information regarding the identify of the a given IP address (they include a long list of things whether applicable or not in the order so every potential base is covered). Once legal was able to review and handed it off to us, we take that and look at the DHCP logs to see that on a given date at a given time that the IP address was assigned as part of shelf A / slot B / port C. That shelf/slot/port combination is tied physically to an address/account. We provide the relevant logs and personal information of that user to law enforcement.

For bomb threats over the phone, telecom switches love to tell every other telecom switch who they are (again, connectivity purposes). So, when you make a call to a business/school doing that, their PBX is going to log to the millisecond when that call occurred and who the switch was. Again, subpoena and we pull the SIP logs. We can even provide the RTP/RTCP packets and reconstruct the phone call audio if the subpoena asks for that.

Comment on

Are ISPs responsible for bots having residential IPs or is this a user problem?

Reply in thread

You can just look at the testimonies from others who have run exit nodes. The cost of your "free" VPN is that law enforcement will constantly be in contact and investigating you because your network/machine is being used to download CSAM.

There is no "oh don't worry, A.B.C.D is just a tor node, we can give it a pass". Every time that happens, it has to be treated with a full investigation.