The Bitwarden clients cache your data locally. So even if your Vaultwarden goes down, you’ll still be able to access your passwords. Just not sync new ones or make changes.

I’d throw in option 3: use a KeePass2 database, sync it using whatever sync tool you like (SyncThing, iCloud, NextCloud, WebDAV, …) and use compatible apps (KeepassXC, Strongbox, etc.)

Might need some

if (ob_get_level()) ob_end_clean();

before the readfile. 😉

And if you want some customisation, e.g. some repeating string over and over, you can use something like this:

yes "b0M" | tr -d '\n' | head -c 10G | gzip -c > 10GB.gz

yes repeats the given string (followed by a line feed) indefinitely - originally meant to type “yes” + ENTER into prompts. tr then removes the line breaks again and head makes sure to only take 10GB and not have it run indefinitely.

If you want to be really fancy, you can even add some HTML header and footer to some files like header and footer and then run it like this:

yes "b0M" | tr -d '\n' | head -c 10G | cat header - footer | gzip -c > 10GB.gz

I was going by this thread and this comment. Those other websites are mostly selling SIMs for industrial IoT or CCTV stuff and are either acting as a MVNO (where I don’t know how much influence they have on the network configuration) or probably selling foreign cards with free roaming or something like that.

Fixed IP might be a problem as basically all carriers use CGNAT. From what I could gathergoogle, Three seems to be the only one where you can manually change the APN to 3internet to get assigned a proper IP address.

https://terminaltrove.com/new.xml

This gives basically no headaches at all. I am running this schema on all my Linux devices. And swap is done using a swapfile instead of a partition. This way, you can easily increase it later on.

Ohh, SQLite isn’t “one” db. SQLite is file-based. I.e. a database in e.g. PostgreSQL (containing several tables, views, indexes, etc.) would translate to one SQLite file (e.g. mydatabase.db3 or myappdata.sqlite). And each app has its own file/database. If the file corrupts, then it’s only affecting that specific app. (However, SQLite is pretty robust.) And since these are just files, you can backup them together with the application. No need to export data or shutdown the database first.

There are smart valves. You can just unscrew/unclip the existing dumb ones and replace them with smart ones. They usually even report the temperature (well, near the radiator - so you might want to get extra temp sensors for the other end of the room).

Unless you share the boiler’s output with someone, I’d be surprised if you didn’t have access.

But still, if you don’t want to mess with the electrical connections, manipulating the radio waves works just fine. :)

Whenever it is supported, yes. I’m a huge fan of SQLite.

Very happy with it. I’m running it with an SQLite database, so no extra database server is needed. And I’m using Elk as a frontend for when I don’t feel like using an app.

On the receiving end of this controller, there’s probably just a relay shorting 2 wires from the boiler together - like the Honeywell BDR91.

So, you could just replace that by some Shelly switch and wouldn’t have to use the proprietary radio protocol.

On a sidenote: Those room controllers usually “learn” how long it takes to reach the target temperature (as this can be vastly different between summer and winter) and move the starting time accordingly. So they reach the target temperature at the programmed time. At least the more expensive ones do. You’d have to do something similar in your automation if you want that comfort. (Unless HA already has some code for that…)

This is what I’ll replace my DS415+ with, too, once it dies. The TerraMasters are basically bog-standard Intel NUCs with a storage adapter. And there’s HDMI output and an internal USB drive which you can just replace (or overwrite) and install OMV or TrueNAS or whatever.

While a non-profit sounds very honourable, I personally think Gitea has the more future-proof model. In fact, Gitea Actions was something they got commissioned for and were allowed to open source it as well.

This is basically how I think about it. (Not my comment.)

But we’ll see…

And Gitea has Gitea Actions. The big difference - and also the sole reason ForgeJo was forked - is that Gitea has a business behind it that allows them to sell support hours to other businesses. And then use the money to progress Gitea. Similar to how Zabbix and others do it. But the community decided that this is a bad thing and moved to ForgeJo.

how so?

I’d like to get educated as well, please. So far, ForgeJo doesn’t seem to have any specific advantage over Gitea. (And the name couldn’t be worse…)

If you spin up a Lemmy instance and subscribe to a community, all new posts and comments inside that community will be mirrored to your instance. As I’m subscribed to around 100 different communities, that was a LOT of traffic without me doing anything. That’s why I’ve given up on self-hosting Lemmy just for myself and went back to using lemmy.ml.

However, I do self-host a GoToSocial server just for myself. It’s probably not necessary as mastodon.social isn’t going anywhere anytime soon, but if you’re on a smaller instance, it might be worth it. Also, you get to show off your own domain name. And, while other instances may block yours, your content stays online as long as YOU want it to. There’s no way for an external moderator to delete posts on your own server.

People can find you via Boosts from others or by searching for your @username@domain.com.

Maintaining my GoToSocial so far consisted of simply getting WatchTower to update the Docker container. Migration of data to a new version happens automatically. (Well, there was one accident where some pre-release version got released under the latest tag and I had to use the development branch for a few days … but that was an accident from the GtS-team and shouldn’t happen again.)

I’ve recently watched this video on YT about some guy that replaced all his social interactions with AI models for for a while. And he, too, said that his ChatGPT “best friend” was too friendly and later in the test preferred some Character.AI “dominatrix” model instead.