For sure, I’ll also edit them into the post

Example 1:

• My post: https://lemmy.ca/comment/19153240
• How they see it: https://sfba.social/@otters_raft@lemmy.ca/115267196743748430

Example 2:

• My post: https://lemmy.ca/comment/19202083
• How they see it: https://mas.to/@otters_raft@lemmy.ca/115283224174559468

IMO it’s partially because investors are willing to throw money at anything “AI” related, and so people are throwing everything at the wall in case something sticks

Unless they mean something like the Respondus rootkit

https://discuss.privacyguides.net/t/best-way-to-use-respondus-lockdown-browser-for-school/26098

IMO some exams should just be proctored in person

Is this because it’s getting difficult for students to mess with the boot options?

Have you tried the third party ones? I’ve seen recommendations for Swiftfin

https://github.com/jellyfin/Swiftfin

I’m not sure which guides to recommend, but in case it helps narrow down your search, you could share more about your situation:

• Do you have any existing hardware or are you planning to buy? If so, what is the budget for the equipment and where in the world are you approximately?

• What did you want to self host? Some services would benefit from a certain type of setup. For example, if you’re serving lots of media, if you need redundancy and uptime, if you’re running AI models or something that needs a GPU

General tips:

• take lots of notes on your process. This will help you iteratively improve your setup. If it all falls apart, it will be easier to quickly get going again since you know what options you picked at each step.
• Make more posts here when you get stuck on something :)

For Linux, a lot of people go with Ubuntu server because there are a lot of existing guides for it. You don’t need much Linux knowledge to start self hosting since you can learn by doing over time. Some concepts to explore before getting started might be cron, the Linux file system, and user permissions.

For Docker, you should be fine if you know the basics. I’d recommend using Docker Compose since it’s easier to understand what’s happening when its written out in a nice yaml file. Install Docker and Docker Compose on the server, and then install something like DockGE to manage the compose files. When you want to run a service, copy the Docker compose file and then swap the port to what port you want to use, and the volume to the location you tend to use.

For a very basic setup, I’d find a video guide for

• installing Ubuntu server on the machine
• basic setup of the Ubuntu server (file permissions, docker, docker compose)

Assuming you need to keep your account for work, here are the direct links:

1. Go to this page and turn it off: https://www.linkedin.com/mypreferences/d/settings/data-for-ai-improvement
2. Submit this form: https://www.linkedin.com/help/linkedin/ask/TS-DPRO

In addition:

• check this page for other privacy settings: https://www.linkedin.com/mypreferences/d/categories/privacy
• Delete unnecessary personal information:
  • Profile image > Manage > Posts and Activity
  • Profile image > View Profile > remove anything you don’t need to include
  • Go to this page and delete old resumes: https://www.linkedin.com/jobs/application-settings/

The AI tool that I saw on there was to give users advice on “how to make your profile better”. The tips were generic garbage, so maybe after they train the AI on member profiles, the tips can be even more generic garbage.

crashes

Maybe, but could it also be an intentional dark pattern to make it difficult to cancel?

Just do like 10 minutes of research before you buy

The average person might not know what’s reputable and what’s not. A lot of the VPN review sites are also secretly run by the VPN companies, or get paid off by them. Someone might even mix up the bad PrivacyTools for the legitimate PrivacyGuides

I haven’t heard of any learning curve with Jellyfin. It seems easy to set up, and the apps are about as user friendly as you can get (especially the third party ones)

On a related note, this story from 2024:

https://www.techradar.com/computing/cyber-security/facebooks-onavo-vpn-used-to-wiretap-competitor-data-court-filings-reveal

Even ignoring any ownership issues, it’s an awful service. If you look up “ExpressVPN cancel” you’ll find lots of posts and comments from people complaining that they were charged again after cancelling, that the interface makes it difficult to cancel, that they needed to reach out to support multiple times, etc.

Instead:

• Use a different VPN
• Check if you actually need a VPN for your situation

I check this guide from time to time to see if anything has changed with the usual recommendations:

https://www.privacyguides.org/en/vpn/

Most sites run as well, if not better, on Firefox for me.

If you’re running a quick and dirty test, you might not get an accurate picture of the performance differences. For example:

• Your usual browser might have cached some content from the last time you used it
• Unless you kill them properly, your computer might not have the RAM/processing to be running two browsers at the same time with the best performance
• One browser might be bogged down with extensions / issues that built up over time

You could try giving Firefox a clean install, or opening it in safe mode (it’s now called troubleshoot mode), to see if there’s any difference

I think they’re referring to Firefox’s funding, a lot of which was through search deals

An article from 5 years ago: https://www.pcmag.com/news/mozilla-signs-lucrative-3-year-google-search-deal-for-firefox

edit: seems like that hasn’t changed by this ruling either

United States District Judge Amit Mehta has ruled that Google can continue to pay other companies, including browser makers like Mozilla, to be their default search engine.

https://www.omgubuntu.co.uk/2025/09/google-antitrust-ruling-firefox-search-deal

An excerpt

Jen asked:

There’s a lot of secure software, that probably people in this room work on, that is developed in the open, and that is used primarily by at-risk users, including things like Tor, Signal, SecureDrop. That’s great, because it makes it easy for folks to contribute. Maybe you don’t want that, I understand that can be hard. But it also makes it easier for people to audit and gain assurance that the app is doing what you claim without having to have, you know, EFF reverse engineer it. Would you be open to making the app open source?

His answer: “Absolutely not.”

Why? “I don’t want anybody from the government to have their hooks in how I’m doing what I’m doing. Once you go open source, everybody has access to it. So I’m just going to keep the codebase private at this time.”

He also claimed that the government can’t learn everything about how an app works by reverse engineering it, which isn’t true.

I agree with Jen. His answers are very concerning.

It depends on how you want to sync. If you’re selfhosting freshrss for example, my personal preference is Capy RSS

In one of the most popular presentations at 37C3, the three hackers uncovered something monstrous: Newag trains went into hibernation using a sophisticated game of hide-and-seek if they were parked for too long within the geocoordinates of competitors‘ or customers’ workshops or were left in conditions that indicated they underwent an unregistered repair. Only by calling in a Newag technician could such deactivated trains be ‘rescued’. All of this was uncovered without the potentially illegal replacement of train components which would require certifications.

What.

Streisand Effect in 3, 2, 1…

It should be possible to grab the license key before you wipe it. You could also copy it into an external drive and store it away as is

I meant in general, is there anyone maintaining it and keeping it compatible with newer versions of Lemmy?