that in theory is correct but in practice a sentence or few lines of code don’t let me fight for them in court.

Pls dont bother to answer if you do not have a proof that a simple sentence or a simple line of code has any meaning for copyright in the court. And no i dont mean the few sentences that are actually copyrighted since they are catchphrases to specifics. No i mean a general sentence, or line of code.

Try to copyright “i = i + 1”. Then we can continue this talk.

When i write a sentence then do i have a copyright upon it which i can defend in law? No. Yes copyright is nothing you get, that in theory is correct but in practice a sentence or few lines of code don’t let me fight for them in court.

The current models are all trained with data which at least some larger creator (book writer or whatever) does not want a model to be trained on. So the discussion is a complete other, not how ppl will be compensated on their unlawful use of their data but on how the payment works in the future and here comes my first comment back.

The current models are under the current law unsustainable. And more and more courts rule in this direction.

Ah ha! You see, that’s the problem.

No. Meaningful in the sense of copyright. Data you get copyright

There you got your distinction.

Class action lawsuits. How would something like this work for anything else? Like Netflix subscription violations?

And we are talking about meaningful data not your shitpost, like licensed code or a book you wrote and anthropic pirated.

I don’t think it’s really possible to do this.

Why not? Both parties have to figure out how much they want to pay or to get payed. Similar like it works for any other license like on Netflix or Spotify. It is not the AI companies job to determine how much they will get payed but how much the data is worth to them and the data creator how much they want.

If i am at a bakery and the bread they offer does not look like its for that much to me then i have not to buy it.

Ah ok. Was confused. Cause that would have made the community edition mostly useless.

Maybe you mixed that up with the Tenant System in the enterprise Edition?

I do not think that this is correct. Otherwise multiple permission features that are part of the community branch would not make sense. Like “impersonate”.

Do you have a link to the part in the documentation?

As of now, i keep them basically for ever, but granularity gets less over time. (restic forget)

Before you go nuts about files you want to exculde the “churn” as you said it, go ahead and actually check how much space this actually takes up. Text and especially logs are highly compressible and will most likely end up taking up almost no space.

Be sure you backup everything you actually need, rather backup a file more then a file too less.

Lock in is more about open standards and protocols then it is about an opinionated implementation. Yes there is not simple “export to whatever” and an alternative service that dies an “import from home assistant”

But you cannot blame a service for not having any valid alternative or other options.

Its either 20 different vendor logins and apps or home assistant.

The charging cirtuit that also tries to guess the battery percentage can be the problem. The ‘guessed percentage’ will depending on the chip and programming go down even if you solder in a dc power supply. When it hits the 0% it will shutdown.

Thats the case for some smart phones.

Consumer Hardware has feature enterprise/datacenter Hardware does not have and the otherway around.

High End Consumer CPUs for example clock way higher and therefore have better single core performance, which can be a benefit.

Non ECC memory and overclocked memory like on Consumer Hardware is often faster, especially when you compare price to performance.

Especially when things like a minecraft server is not optimized for a 128-core CPU or even 32-core CPU.

Firstly, wildcards are convenient but they break your functionality of to which you are connecting which is half of what TLS does. That may or may not be important to you.

You are implying that https is supposed to be done by the individual server and not reverse proxy. Since otherwise you could not “authenticating the server”

Short answer, it should work as you describe, everything local, no need to expose anything.

More information is needed to help.

Its either the caddy setup, maybe proxmox internal networking? Or most likely DNS, its almost always DNS.

Its good practice to handle https on a reverse proxy. Also who actually does cert pinning?

My guess, auto updates for major versions is not good practice. Installations potentially corrupt and users come screaming to the project maintainers instead of searching the blame by themselfe.

forgejo defined the lifecycle for every version so just add a calendar entry for when the supports runs out and then update the 3 characters in the docker compose file.

netbird uses wireguard as vpn protocol

To my knowledge:

• File Versions only 30 days available.
• -> Everything older and deleted than 30 days gone.
• Slow and manual Restore Process
• -> Web downloads max. 500 GB manual selection.

B2 is data storage, Backblaze Personal is a service. 2 Completely different things.

You can share albums with other users, or even with a public link.

That is not what i have asked.

I doubt that this applies to files smaller than the pack size or a for example bitwise unique 17MB file would result in a (overly simplified) 16 MB and a 1 MB chunk or am i wrong here?