Isn't that generally what external auth providers are designed for?
I imagine it isn't much of a concern if Immich isn't exposed to the internet.
Isn't that generally what external auth providers are designed for?
I imagine it isn't much of a concern if Immich isn't exposed to the internet.
It's been like that since the Pi was around $150 tbh, at least when looking at used hardware as a comparison.
I think they're neat, and can make it easy to get started.
However, when something breaks it can be really difficult to fix, as you have no idea how things are set up under the hood. Often if an app isn't in the 'store' it can be much more difficult to run it as well.
So upsides and downsides, easier to get started, harder to maintain imo.
It'll stay connected to both laptop and desktop at the same time without issues? Does it auto-reconnect quickly when coming back onto your wifi from mobile data too?
Maybe I need to give it another try..
Yeah, and a modern x86 setup will be faster and likely even more efficient under load, with better software support too.
I believe AirVPN is a solid one, port forwarding support too.
The advice I use for myself is always assume a service will either turn shitty enough that you need to leave, or just shut down due to some reason, so I try and only use services that make it easy to move away from.
So instead of Proton or Tuta, use a normal email provider that provides IMAP and supports a custom domain name. There are good ones around for privacy and features like Mailbox.org or similar.
A domain is ~$12/year and ensures you (hopefully) never need to change a bunch of emails on accounts again even if moving to a different email provider.
I use it for sending files between devices occasionally. KDE connect has never been reliable when switching between PCs, it will only stay connected to 1 at a time and I have to often manually connect it to the others which is annoying to do.
Its file browser/send feature also feels a bit half-baked and takes more clicks than it should IMO.
Yes? The error should at least show basic info. Requiring checking the logs when the error message could just show some info is silly and clearly doesn't work well given how many people are asking for help with it.
Because there's no error code attached or any details about what specific error it was, what file it happened on, what mount it happened on, etc...
It'll stop a lot of them, some will slip through though.
Tracking pixels are just a 1x1 pixel 'image', the way they work is when your client loads the image URL it sends a requests to the server to do so, and the server logs the request and knows you've opened/viewed the email.