I use Iceraven with ublock, privacy badger, decentraleyes and canvasblocker.

Yes, the main thing is you don't have to pay for nice!nanos which are $25 each as I remember. XIAOs are only $10 each. The price I put on tarneo.fr is only as I remember it, might be a bit more depending on location and the shipping options you choose. But yeah I guess even with that it's cheap.

Thanks! Looks like I've reached my goal ;-)

Not really, you get used to the small number of keys if that's the question. It's really just muscle memory after some time (I've been using this layout for around a year, iterating occasionally)

Thanks! Some people find the monospace font hurts their eyes though, but I guess it's a tradeoff of the 90s theme

DIE

Framaforms + framacalc.

My servers have names of Spanish words humorist El Risitas says in his mythical video where he laughs with no real reason.

The biggest server is named "cocinero", because I can (jokingly) easily imagine a very fat cook.

Then there is plancha, a lenovo thinkcentre which has the size of a plank.

My raspberry pi's have names of tapas: chorizo, keso etc.

Yes. But p10k has many downsides:

• requires using oh my ZSH, which alone is quite bad because of how much slower it makes the shell.
• is a piece of software you'll have to either install on each new device or have the software in your dotfiles. Bad practice. I very much prefer having no additional dependencies or overhead, plus the way I do it I can do whatever I want without the limitations of a prompt made by someone else, for which I'd have to dig in a lot of documentation. Compared to this, I only spent half an hour making a prompt exactly how I like, which doesn't add overhead and doesn't require a third party piece of software which I'd have to install on every new device.

Free software tells you "do whatever you want, you're free" but open source completely misses the point: it means you can read the code, but not necessarily recompile, modify and redistribute. Plus the term was invented for the confusion that would come from it. For example, a lot of AI models like LLM's claim they are "open-source", which basically means nothing: it's far easier to say that than to claim it's a free model, because that would imply freedoms to modify, reuse, redistribute the training data, weight etc. (no AI model allows that for now, and there will probably never be one that does).

Unattended updates are 10x better because those programs allow you to only do security updates. Plus they are much more stable, and something like this would never happen on a stable distro.

Yup. gives nickel back