I think most information is passed through the global discovery api: https://docs.syncthing.net/specs/globaldisco-v3.html

STUN is only used to figure out public IPs and port and should store no information.

Doesn’t have to be SQL. But most of the time that quote refers to a relational database.

Nowadays there are graphical tools that are alright, such that you don’t have to learn a query language. For example (1), (2) or more commercial (3).

But what’s still important is doing good relational database design. Learning to look at the world as entities and relationships between them. Constraints, keys, indices. There’s books and courses on that. While you’re at that, you’ll probably learn SQL along the way, as it’s so convenient.

Meaning they still don’t understand that it’s morally disgusting to scan everyone’s private communications. They’re just annoyed that people care.

Their proposal should be explicitly outlawed, privacy should be a human right.

Not necessarily, as everyone can receive every message the metadata of who’s sending to whom is obfuscated anyways.

Nostr is a twitter clown/clone.

That’s confidently incorrect. It’s a signed distributed trustless message protocol (1). You can use it to build whatever. The most well known application is a twitter clone.

Whittaker notes that AWS, Microsoft Azure, and Google’s cloud services are the only viable options that Signal can use to provide reliable service on a global scale without spending billions of dollars to build its own.

As long as they insist on being man-in-the-middle. Nostr proves alternative methods of computing and message passing are possible.

I rate 9 out of 7

if a tool exists which can easily and reliably detect AI generated content then they’d just be using that tool for their training

Generative Adversarial Networks are an example of that idea in action.

One of these projects might be of interest to you:

• https://github.com/Mintplex-Labs/anything-llm

• https://github.com/mudler/LocalAI

Do note that CPU inference is quite a lot slower than GPU or the well known SAAS providers. I currently like the quantized deepseek models as the best balance between quality of replies and inference time when not using GPU.

That’s what Palantir Gotham does: “predictive policing”.

When voting do read the political manifestos

I’d argue against that. Instead look at past voting behaviour. What politicians say, and what politicians do, tend to be very different!

Voters in:

• Austria.
• Belgium.
• Czech Republic.
• Finland.
• Germany.
• Luxemburg.
• Netherlands.
• Poland.

Your members of european parliament have made a good choice!

Voters in:

• Bulgaria
• Croatia
• Cyprus
• Denmark (!)
• France
• Hungary
• Ireland
• Italy
• Latvia
• Lithuania
• Malta
• Portugal
• Slovakia
• Spain
• Sweden

In 2029 there’s new EU parliament elections. You can list your representatives on https://fightchatcontrol.eu/#delegates and make sure not to vote for them ever (again) 👍

The incumbents in EU are very scared as politicians from outside the traditional political families are getting popular votes. And instead of looking into to mirror as to why that is happening, they blame “the internet” and go authoritarian.

Another example: EU parliament also accepted a last minute amendment, mandating age verification for pornographic (whatever that is) content online, punishable with up to one year prison sentence.

This was a last minute addition to a directive concerning CSAM. Because adults accessing porn need to be de-anonymised to avoid child exploitation?

The EU apparat is self-serving, enjoying their private internal turmoil and drama. Annoyed by those pesky voters and their stupid wants and rights. If only there was software we could use to get them to shut up?!

(1), (2), (3)

For those who enjoy self-hosting, there’s coboldcpp and lamacpp.

Get outta here with your facts.

I’ve worked with bookstack. Found it easy and intuitive. It’s wiki software not specific to family history.

I think it depends on the rate of change, rather than the amount of containers.

At home I do things manually as things change maybe 3 or 4 times a year.

Professionally I usually do setup automated devops because updates and deployments happen almost daily.

At one of my clients, who wants everything on-prem, I use gitlab CI with ansible. It took 3 days to setup, and requires thinkering. But all in all, I like the versitility, consistency and transparency of this approach.

If I’d start over again, I’d use pyinfra instead of ansible, but that’s a minor difference.

Those benefits existed long before in VPS-es.

What I understand that was new as the marketing terminology shifted to “cloud” is that the provider not solely provided you the virtualized hardware with a stock OS. They would also manage typical software like nodejs, database, elasticsearch, queue, … Along with the annoying stuff like updates and backups.

Basically reduce the amount of work invested in the ops part of devops.

Funny enough, that website uses the illegal “consent or pay” dark pattern:

Accept cookies, including advertisement cookies, to read for free. Or become a paid subscriber and decline cookies.

They’re lucky GDPR is only enforced against large US companies - almost as if they’re specifically designed as a tarrif against them :D