This is why a government needs an actual technical expert. What UK has done is make icloud less safe for everyone, while the “criminal” activity they seem to want to be privy to will force criminals to use another method of data transfer/securing.

Yep, that’s what moved me to OpenOffice and now libreoffice

I use Newpipe since I’m not going to login. No ads

You could scp them from server out, you can setup simple http and grab via web, you can use a phone and take a picture, you can read them on the server and type them here. Lots of ways besides ssh into the machine from a client

People are replying like they are because you have asked for help but aren’t able to provide details on the steps, the yml files, or other commands you executed. Essential it is like when my wife calls me on my phone and says “I’m lost, where am I?” And I have no clue where she is because she can’t provide me street names or what direction she is travelling. Basically nobody here knows where you are.

It has been working well for us, we have a central work server and most use WebGUU, but some of us have the Sync or Drive Client and Phone Apps. I only had to pause sync and restart for a file that wasn’t syncing once in about 9 years

Have you tried Seafile. https://www.seafile.com/en/home/

It has server and client. And you can choose what to sync, so you don’t have to pull down entire folders.
Also has a webgui. Client side has Sync client or Drive client for Windows/Linux/Mac.

Drive client works more like a network drive so you don’t need to sync the files.

Also has Android/ IOS app

I personally believe you are overbuilding. For example my OpenMediaVault Samba Server and DLNA server runs on a SingleBoard that has 256 megabytes of RAM. Yes MB. And it still has RAM free without swap. And I should alter my clock.

You can have more than one drive on a pi, either USB, or they sell sata boards

So their CEO is clueless to how his API changes drove people to the fediverse?

Not sure on MergerFS…I haven’t gotten deep into ZFS pools.

Also MS, let’s add it to everything.

What ZFS are you using?

“As of November 2023, this feature was merged in main and is scheduled for release in OpenZFS 2.3:”

'The OpenZFS project (ZFS on Linux, ZFS on FreeBSD) is working on a feature to allow the addition of new physical devices to existing RAID-Z vdevs. This will allow, for instance, the expansion of a 6-drive RAID-Z2 vdev into a 7-drive RAID-Z2 vdev. This will happen while the filesystem is online, and will be repeatable once the expansion is complete (e.g., 7-drive vdev → 8-drive vdev)."

Awesome! Interesting with MTU. This was a first hit google search, seems MTU in wireguard setup plays a role also. https://gist.github.com/nitred/f16850ca48c48c79bf422e90ee5b9d95

It may not even be the issue, I’m just spitballing, and it maybe DNS issues like others mentioned or cell provider blocking some aspect. As you probably know Wireguard can be set for all traffic to route through the tunnel, or some outside of it. Maybe cell network is not routing DNS through the tunnel, but using its own, or maybe they used a cached Domain name lookup that doesn’t have your domain IP. Again, networking/DNS etc is my blind spot. I.e. I have setup openmedia vault on a 256MB RAM arm board to serve my music and SMB shares, but I don’t understand reverse proxy LOL.

I don’t have a lot of networking skill here, but could it be your WiFi connections (anywhere) are IPv4 and on mobile it forces IPv6, thus why you can connect via IP:port? Typically your wireguard host machine has IP forwarding/masquarading setup so you can reach your home LAN. Could this be an issue through the router / proxy?

I never tried for YouTube. I mainly use for technical info, in which it gets me to the results. Google has an advantage for YouTube search since they own both platforma and share meta data. Stract is probably oldschool webcrawling. I did read somewhere that it takes time for it to build up its hit results. Did you try the content filters?

Stract.com free and no ads. Some guy built it in his basement. He suggests it may need paid tier or ads in future to support it. It gives results like google did in 2010 era before all the ad nonsense

It would still have happened at some point, chatgpt is not the only AI. I hate we called it AI.

I bought a new XBOX drive off amazon once. Stickers says void if opened so I didn’t open the case till I found a larger drive and was going to swap the drive in the case. The drive inside was dirty as hell. Like it spent years in a dirty factory. SMART report looked good, but the drive itself was from 2011, but the controller board was from 2017. Some retailer was grabbing old drives and swapping the control board, I assume as repair or likely to falsify drive use hours.
I reported to Amazon as seller fraud, initially they said you are way out of your return period. So I had to explain it doesn’t when I discover a seller has commited fraud, it is fraud.