Right AC. Was a typo, corrected now.

As for running nominally 12V stuff directly of a battery… you got lucky with your router then. I had the same idea with various mini PCs and ARM SBCs, and the majority absolutely should not be run directly off a battery where the voltage fluctuates so much. And it gets even worse when you try to solar charge the batteries.

Probably best to use NUT to shut down the servers gracefully if the power outage is too long.

When charging, 12V lead acid batteries have significant higher voltages (>14V), so you basically need a DC to DC converter to make sure you get a stable 12V supply. Some (typically industrial/automotive grade) systems have that built into their internal power supply and can work directly off a 12V battery, but with consumer routers I would assume that not to be the case.

While there is DC AC power the battery in a UPS isn’t used. It’s not a constant AC-DC-AC conversion, that conversion only happens when there is a power outage.

The problem is that lead-acid batteries are a really shitty technology with high self discharge rates and short overall life span (their only advantage is high burst capacity that’s good for starting motors).

I would look for 12V LiFePO4 batteries in a standard lead-acid form factor that are meant as drop in replacements. Those are somewhat more expensive (like 2-3x the price) but will likely last a quarter century or more.

I am currently using an interim solution where my solar inverter has an seperate emergeny power outlet that works even when power is cut. Sadly it doesn’t switch to battery fast enough (like half a second delay 😓) for it to work as a UPS directly. But in combination with some old regular UPS that have spend batteries that only last a few minutes it works ok. But yeah, not the most efficient option conversion wise, but solar power is free 🤷

The holy scripture should only be read in the original German /s

https://snikket.org/ specifically for a small family server. Much lower maintenance and breakages than Matrix and thus less issues with the family.

A water cooling block will fit 1U, but you will need to attach the other side of the water cooler outside of the rack most likely. Furthermore, it is unlikely that other people using the same rack space will appreciate you adding water cooling due to spillage risks.

I believe raised beds for kids have a lot of space beneath them for putting servers. Just make sure they can’t pull the plugs 😜

Yes, I am aware. But that is not very helpful if you want to host something publicly or even just want it accessible from the random wifi hotspot you might use that more likely than not is IPv4 only.

Like literally all mobile (4/5G) internet connections are behind CGNAT, as is Starlink for example.

For landline internet connections it is indeed uncommon.

If it is really only a private server for your friends, and they are not completely computer illiterate, then a Wireguard VPN or similar seems like the safest option.

But generally speaking it isn’t that big of a deal to temporarily open a port and forward it to a game server if you want to play a round online.

A VPS is not really cost effective for that kind of usage. Maybe one with hourly billing like this: https://tierhive.com/

The exact issue that was finally fixed in the above release caused Conduwuinity to fail federation all the time. Synapse never had that issue. Yes, Synapse sucks in other ways, but be careful to think the grass is greener on the other side.

I mean yes, Microsoft and slopcode probably plays a role, but it only got really bad once the LLM mass scraping started, for which I assume Github is also a major target.

And running your own public git forge is a real PITA these days for similar reasons.

If I remember correctly, Apple makes it really hard to bridge iMessage and they implemented some sort of hardware check and even disallowed older versions of MacOS or so. As usual, they really like to lock people into their ecosystem.

P.S.: something is wrong with that link you shared.

Seems not too bad. It seems to boot with mainline Linux (or at least Ubuntu ISOs) and the vendor releases patches for recent kernel versions, which is much better than most ARM SoCs.

Cool, let us know how it goes and maybe also see if the power consumption has improved with more recent kernel versions.

Seems indeed relatively good value for money, but how is the mainline Linux kernel support for that SoC?

Edit: also in the early reviews people found some strangely high idle power consumption, which really isn’t good for a home server.

That is not my understanding of the timeline. Yes, they asked their members to vote on it a month ago. But these votes take time, and once the vote was settled and the ToS adapted they immediatly put up the banner.

It is not my argument, but the official stance of the US copyright office and some other similar bodies around the world.

But yes, it poses some interesting philosophical questions, but those are not of very high relevance to copyright law.

Basically the transformative abilities of LLMs can make copyright obsolete.

That is wishful thinking. Copyright was originally designed to protect human creativity, and a lot of the laws and precedence cases involve human effort specifically to distinguish between plagiats and original works.

As long as courts don’t start to rule otherwise, LLM generated code is not considered to be copyrightable itself, thus it can not be original work as defined under copyright laws. This in turn lowers the threashold for such code to be considered a derivative regardless of if there is a clear pathway from one to the other. All it takes is someone to claim that it is derivative, and the burden of proof that it isn’t largly falls onto the person accused. And if the code is functionally identical and with no significant human involvement, that is very hard to proove.

IANAL of course.