That makes a lot of sense and where I’m leaning towards as well

While my homeserver still has plenty of resources to spare, I see a lot of them going towards multiple DB containers. It’s nice for “segregating” the containers, but backups are also a pain, gotta plan backups/restores for multiple DBs

Same story with an s3 (well, minio) instance running. Seems like it would make more sense to centralize DB and file operations and having different services talk to them. Then if I ever needed to move them into separate servers, it wouldn’t be as big a move.

Thanks!

I use Linux on my personal laptop, my work laptop is a Mac, but my desktop (main computer) is still Windows largely cause of video games. Lot of the games I like to play don’t work or require more tweaking than I’m willing to invest to get them running on Linux. I also play flight sim and racing sim games with peripherals a lot, and if the game support on Linux seems bad, the support for those peripherals is even worse lol.

Immich for personal photos and pictures I take

Lychee as an image host for funny pictures, memes, and publically sharing photos

Seems like a good chef/recipe writer/whatever he is, but I dislike his personality/TV presence and can’t stand watching him.

I don’t, I really wish I did more, but I’ve found it a difficult habit to develop and keep up with for some reason, although I have tried several times.

One of the most successful methods I’ve had was with my sous vide cooker. I would go to Costco (/Sam’s Club/BJ’s/whatever bulk goods store) and buy a variety of meat, couple packs of steaks, pork chops, chicken, etc. Once home I would season and portion the meat out into individual servings, then vacuum seal and freeze. Before work every morning I’d throw a frozen protein in the water bath and go to work, which was only 5 minutes away from home, at lunch time I’d clock out, rush home, quickly sear my whatever was cooking and add a can of vegetables or some other leftover for a side. Was a phenomenal system, but only lasted a few months before the job ended, and just haven’t tried to pick up the habit again.

At one point I set up Mealie, a self-hosted recipe tracker/scraper that worked well and helped to generate some grocery lists. That was nice because I was able to select different things I wanted to make over the course of a week and have it generate a list of what I need ingredient wise.

However I never really “stuck with it” as a habit. Mealie is really cool in that it can (theoretically) scrape recipes out of other websites, so it centralizes your recipes and strips the “blog” fluff out of them. But in practice it wasn’t great at doing that, it relied on very specific metadata tags that just aren’t present/formatted properly in a lot of recipe blogs, so it wound up being more trouble to use than it’s worth. If I were more dedicated I might be willing to manually transcribe the recipes, but I ain’t lol.

Anyways, apologies, I realize that’s kinda ranty and doesn’t really answer your question. I’m posting partially cause I hope other people will share their meal-prep-planning and I can steal ideas haha.

Not entirely. There’s a couple larger subreddits I enjoy and tbh it’s still my go-to for doomscrolling.

However my desire to interact with reddit couldn’t be any less. Zero desire to post or comment anything due to the sheer hostility of the site to it’s users.

I think if lemmy continues to grow in userbase it could completely replace reddit, but not at this immediate moment.

I see a couple ways you could do this. For what it’s worth, I think Lemmy may be close enough to what you want out of the box. I run my own private instance similar to how you’re laying out, although I don’t “disable” registration, I do require account approval by an admin. I’m sure at some point I’ll start getting spammed with registrations, but thus far it hasn’t been an issue. Similarly, federation can be disabled as a setting in the admin section of Lemmy.

What you’re laying out on the technical side is also absolutely possible. Lemmy has a JS client you can pull in as a dependency to manage the calls from Javascript, but it also describes the HTTP endpoint for said call, so you could make your own calls using a separate back end service (ie: after your form submits it makes the call over to your Lemmy instance to register the user).

Here’s the register call specifically, which registers a new user on your instance: https://join-lemmy.org/api/classes/LemmyHttp.html#register

However, note that they would still not be “approved” (if your instance required approval to join), but you could just make another API call to complete that process: https://join-lemmy.org/api/classes/LemmyHttp.html#approveRegistrationApplication

As another option, I haven’t dove into the actual code, but to the best of my knowledge Lemmy is ultimately just interacting with a Postgres database, my guess is you could likely also connect to this database and directly insert the appropriate rows to create your new user accounts. Definitely use some caution with this approach though, ensure you’re using separate database users with appropriate permissions for each application. Additionally, know that if you write code to handle custom user registrations by writing straight to the DB, you’ll have to be mindful of updates to Lemmy that change the DB schema for that table will (most likely) break your registration script/code.

Lastly, if you really wanted to, you could just fork Lemmy itself! The UI and API are completely separate services, there’s nothing stopping you from forking the UI part of the project and applying your own custom changes to the registration flow for your own instance!

Enjoying it a lot. Set up my own instance to try out, and exploring a lot of others. I enjoy the community a ton!