lycos

But it is fun to sometimes program something small like this for yourself (and friends/family), without having to think about the many things you would think about in a business environment. If you can let go of those mental constraints, that is. I recently made a little mobile remote control app for my e-ink picture frame and the code is terrible. Everyone who sees it is impressed and it works without any issues 90% of the time, took me about a day and I get happy every time I use it.

Hot code updates for most applications aren't really worth it in my opinion, assuming you do something like blue/green rollover deployments. It's cool that it's possible though. But it requires appup files and afaik Distillery is one of the release tools that has support for it built-in.

I recently built my own e-ink picture frame because I wanted it to feel like a regular picture frame (and it has fooled people already), because it is basically just a printed photo in a nice normal frame. And it looks amazing in direct sunlight, and obviously in the dark you see it less but that's the same with any normal picture frame.

If color e-ink was ready for this use case it would be pretty cool, but I think we're still a while away from that. It might also be too expensive.

LiveView itself is an interesting concept and while we do use it in production daily for almost 2 years, I'm not yet convinced it's really better than having the frontend be React (or whichever you prefer). It has grown quickly and become much more useful than it started out as.

Ecto in my opinion is the thing I would miss most were I to switch to something else.

I'm coming from homebrew + asdf, but so far Nix appears to be working fine for me. I think asdf has a lot of custom config to try and isolate version-specific shared files/modules etc. I can replicate it by setting some environment variables myself in nix shell file though. I've got my basic home manager setup (still need to migrate some manually managed files to nix config) and every project I work on is now running with their own shell.nix file. No huge problems yet, aside from the mentioned watchexec failing.

Not yet sure if I'll keep it in my regular workflow, it's been an interesting experience and I would be happy to not need homebrew anymore. (Might just try ports too). I'm also curious to see how fast new versions of software will be added/supported, so I'll have to at least keep this running until that happens.

The main problem I have is when I try to install a package that fails to install, usually it has to compile. Like today I tried watchexec on my M1, doesn't work. Luckily I can just download a binary watchexec build to get around it but it would be nice to know how to get it working. Worked with homebrew. (but homebrew has it's own problems)

The other thing is that I like the idea of using nix-shell for projects, but I hoped they would be more isolated than they appear to be. It's possible for example to have files in a user directory which will be accessible to all projects and package versions I'm using. It hasn't become a problem yet in my testing so far though.

Reading other comments I decided I will also play with running things in a VM with NixOS. Maybe I'll end up with a mix of local Nix for some basic stuff and then a development VM.

That's just what I wanted to read having my first ever surgery planned for next week, another thing to be afraid of happening!