Recent comments in /f/Privacy

BlackWinnerYoshi wrote

Those are the essentials I think that any VPN provider, and any provider of any service, really, should have to be even considered by me:

  • Tor support - that means having an onion mirror for the VPN site itself; Tor with the VPN itself is not necessary, as Tor is kind of easy to set up, at least on Linux
  • If it's paid, Monero acceptance - obviously, one might not want to be using some kind of anonymity-hating VPN, like with the point above
  • No personal data for registrations - that means confirming your phone number is out
  • Compatibility with standards - in the case of VPNs, that's OpenVPN or possibly WireGuard
  • No Cloudflare - MitM, especially Clownflare, likes being a honeypot: https://codeberg.org/baobab/cloudflare-tor (clear net only)
  • not a privacy-related issue, but if the service likes to go down often, it goes out too

Okay, I don't actually pay with Monero or any other cryptocurrency, but other people might, so if I'm going to recommend a service, I want to check that.

4

Rambler wrote

Wordpress CORE is relatively secure. They've got a giant team that is always pushing out updates. The trouble usually comes from a variety of plugins and users not updating their Wordpress or not changing default values.

I'm not sure what a good alternative would be unfortunately. I have a HTMLy flat-file blog instance running on I2P that I never update and it's... okay. Pretty simple.

2

riddler wrote

Many years ago I needed a face database for some AI work. Everything I could find for free was crap and it wasn't worth the price to pay for any of the better ones. For a few weeks I took a two hour lunch in the tourist area of town at the busiest fast food place. I parked right by the sidewalk and had a camcorder record HD from my car. I filtered out people who didn't look at the camera or who's face was partially covered. Lighting was consistent because it was always recorded at the same time of day. I probably had 5-10k faces of usable quality though I never tried training with the full data set. A few of the other grad students in my lab asked to use it when they found out about it. As far as I know my filtered version of that database is still floating around at the university.

In short, there is way more data than can even be identified floating around. That being said, the Facebook dataset is likely a treasure trove of interesting stuff. I can't imagine what deep learning on millions of pictures with their associated metadata could figure out. It's sad that data isn't being used for good.

3

DcscZx5idox wrote (edited )

Search Engines - which one to choose? (onion service mirror) (eepsite mirror) - DigDeeper website

DigDeeper website tells that there search engines are weak of privacy or usability. The parsons another pages tells that Email providers and browsers instance a case of evil. However, alternatives are better than Google.

1

Rambler OP wrote

By confused, I mean for example input boxes for forums disappearing, unless you switch to a user-agent they recognize and "support" or similar shenanigans.

Or sometimes getting mobile versions of sites when browsing on the desktop, because the site thinks you're coming from an Android or iPhone device.

1

Wahaha wrote

I'm using "Random User-Agent", which looks to be doing pretty much the same thing.

I went that route after discovering how you can be tracked based on your user-agent and that there's no way to make a meaningful change to it to prevent this. Thus I just change it every couple of minutes.

Some sites get confused by this, but overall it's a seamless experience. Not sure if it's actually preventing tracking, but it sure will defeat naive approaches.

By confused, I mean for example input boxes for forums disappearing, unless you switch to a user-agent they recognize and "support" or similar shenanigans.

2