Hacker Newsnew | past | comments | ask | show | jobs | submit | Actor's commentslogin

https://metrics.torproject.org/bandwidth.html

It's gotten much faster over time, I have no trouble watching YouTube.

>I don't really know what most people get up to on Tor admittedly.

While there's a decent bit of cyber crime. I would guess a disgusting amount of that happens on Telegram and Discord.

Journalists use Tor. I read Richard Medhurst a British journalist talking about whonix iirc. But maybe that was just qubes.

Many people use it for censorship circumvention. People do not like to be censored especially in countries where they weren't at a time they remember. And the great thing about Tor for censorship circumvention is that it's completely free.

So while a VPN might be faster and Tor might not be needed for the circumventing censorship, Tor's accessible and free, so that makes it popular.

Another interesting use of it I saw recently is bridging securely and privately out of your local network with onion services. You need to install the Tor browser on whatever computer you want to access your website from, but other than that, you can get out of your local network for free.


Have been doing this for a decade and it works great. Tor punches through any firewalls, I don't need a dns name or one of the free dns-providers, but use the .onion address. I use it for ssh and backups and have successfully used it from lots of different countries.

We have Nym mixnet which is supposedly better but it's way slower! Although technically less exciting and not a Tor replacement, their VPN is a cool improvement on top of regular VPNs.

Tor has been getting way faster over time. I can watch YouTube in high quality.

https://metrics.torproject.org/bandwidth.html


If your threat model includes your ISP, which I assume most people using Tor include. A mitigation is to use a trustworthy VPN before Tor.

https://www.privacyguides.org/en/advanced/tor-overview/#safe...

It's amazing how fast Tor has gotten so I really notice no slowdown.


I was always wondering how it became so fast recently ... maybe an indicator for a wide-compromised network, otherwise I can't really explain how something volunteer dependent that was quite 'slow' for most of it's existence gained such speed increase.

I recall reading somewhere (perhaps some presentation?) that Tor had massive speed gains from implementing a better congestion control algorithm. A quick web search found https://blog.torproject.org/congestion-contrl-047/ which seems to have some details about it.

Understanding the origins of the network might help you understand its current network participation: we’re in the middle of a World War, and it’s a network very beneficial for spooks and criminals.

I started running a non-exit node a few months ago and after a little while it became trusted and became a guard node. Guard nodes seems more important than amiddle node so I was happy to run one I know is legitimate. That makes at least one legitimate guard node for the network.

I noticed the attack when my mom said the wi-fi wasn't working well and I checked it out and the download speed was down 99% from normal. And eventually after various other troubleshooting attempts I turned off my mini pc where the relay and other self-hosted services run and the internet speed returned to normal. Yeah I'm not a very adept sys-admin, AI set up all the docker compose stacks for me.

I then looked at my relay on the relay search and it said overloaded. So I checked the Tor forum and sure enough others were reporting the same thing.

So at first I tried just lowering the bandwidth let through, although I'd thought it should already not be letting that much through. That did not work.

So I made it not become a guard node and now it works fine. But I also read something about malicious guards and exits being how Tor users have been compromised before. So it seems important guards are trustworthy.

I'm not gonna risk running a guard again (letting my node become a guard) at least for a while until this attack stabilizes.


I'm not sure why this comment was flagged. Vouched.

My comment is blocked because I signed up on an IP that is blacklisted.

How was 'Actor' username not taken... Are usernames not unique?

It's a secret.

nobody gaf about og usernames on hn be real

Apparently not..

Thanks for the input, ‘rapist’

Was it a Tor IP

Why would the HN algorithm auto-flag comments from Tor IPs when it can just auto-kill them

To be accurate, flagging does not "block" comments

We can still read them

If flagging is supposed to be a mechanism for "alerting moderators" about a submission or comment of which they are unaware (there may be thousands of active submissions or comments spread across 100, 200 or more pages on HN at any one time) then perhaps flagging submissions or comments on the /front page (cf. pages generally no one is looking at) is not being used for that purpose

Perhaps HN users who flag submissions or comments on /front are reacting "emotionally" as opposed to "rationally"

HN readers are usually left to speculate the reason behind flagging

Good example from yesterday:

https://news.ycombinator.com/item?id=50496921

The reasons might be legitimate or they might be nonsense

I couldn't see a "rational" basis to flag that submission. It's a blogger's opinion


The AI detector was all over it.. this didn't need to happen.

It didn't flag my brain's AI detector, and it didn't flag pangram either.

No LLMs but AI was in use—I am quite enjoying parakeet tdt v3.

the ai detector is literally stupid on hn

the mods are out of their depth in this moment unfortunately. hopefully they'll figure it out soon.


I think it's pretty decent normally, it sounds like in this instance it was an IP reputation issue anyway.

As far as we all know there is no auto-flagging of content on the backend based on pangram (or similar) detection.

If you see a comment greyed out or dead comment, it's because fellow human users (or presumably third party bot farms) decided that it was ai/crass/boring/unrelated/wrong.


Can also happen because of bans/shadowbans. My account somehow got automatically banned but after a couple of emails they cleared it up. I (and they) have no idea how or why it got banned but ... apparently there's some automation somewhere!

can't blame them though, the amount of utterly useless arseholes farming karma with ChatGPT comments is unbelievable.


HN karma, the rock bottom of the cryptocurrency barrel.

impressive you ran a guard on your isp service

I run an always on commercial VPN on my router so it affects all the other devices. So the ISP given IP's reputation doesn't matter to me.

I would let it become an exit node if not for the physical risks.

I wouldn't really consider it impressive. I mean, yeah, they're pretty attacked services, as I found out first hand. But nothing compared to an exit node.

If you do need your home IP to be in good shape reputation wise, then don't do it. It's unfair because guard and middle nodes don't hide malicious users to websites, but they still get put on the blacklists, lumped in with exit nodes.

If I need a better IP than the commercial VPN on my router, I can just use cellular and a hotspot.

I really don't find it bad running everything on a commercial VPN though. For a lot of people it's that one site, though.


Although it's not very good. I still want to recommend something I've recently developed: https://github.com/DennySORA/vpn-manager

This allows you to isolate specific apps and specific web pages (you need to install a proxy extension, which I developed myself).

At the same time, it can start multiple isolated VPNs, allowing you to customize incoming and outgoing traffic.

It works on Mac and Linux, and makes Mullvad and AirVPN easier to use (P2P Port forwarding is more convenient).

I previously also wanted multi-layer nesting and routing traffic from different apps to different VPNs, but it was never convenient enough, so I developed this.

More importantly, it's super easy to use. Basically, you don't need much knowledge, and you can even directly ask an AI agent to help install it.

By the way, are there any other highly private and secure VPNs you can recommend?

I actually want to integrate more VPN providers, and then do multi-layer nesting or more detailed traffic splitting.


>> are there any other highly private and secure VPNs you can recommend

Yes, but running a Tor node through a VPN violates many terms of service agreements. It is one thing to operate a server via a VPN, another to operate a public communications pathway. Many would argue doing so is effectively sharing one's VPN subscription with the general public.

A VPN-owned IP address becoming a known/listed Tor server, even if only a guard, is not a great thing for a VPN.


its heavily suggested and known to NEVER run a Tor exit node from your residential ISP. please do not do this.

Thanks.

This is hyperbolic. I've done it for years. It's been fine. Obviously that IP is "tainted" and on a number of lists but that's about it. If you always use VPN to connect out anyway, that doesn't make much of a difference. Nobody knocked on the door yet.

These days I guess tor exit traffic is dwarfed by various residential proxies bundled in consumer devices and malware. They'd be raiding a bunch of families with Superboxes and stuff like that before they get to my little tor node.

We need more relay and exit node operators, IMO.


Guard is fine. Any relay with reasonable uptime becomes a guard. Relays with low uptime are confined to middle position only.

I think you were confusing guards and exits.

no. surprised someone would burn their isp upload on a guard.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: