olio

I worked on small features and made one big one. This time I tried out properly vibecoding (gh copilot agent mode) a whole feature, namely file uploading (I know, not the best one to use AI for, but I did check it mostly :P.) It was an ok experience btw, I used the claude sonnet 4, cuz that's the one I've heard the most good things about. It did understand what I wanted, and it did produce something workable at the first try, but I had to go through quite a bit of iterating to get a properly working function. I followed the styling pretty abysmally the first time but it understood most of what I asked it do do. All in all I would somewhat recommend at least trying it out, but be wary of all the crap it sometimes decides to make.
The integration is quite good, and it did find all the important stuff to know about the project, despite me creating the ugliest architecture ever to exist.

I have been scouring the interwebs to learn c and find libraries for my usecase. But I will probably only really begin coding starting now. I don't even really know, how wakatime counted this much time.

Been fighting every kind of deployment method, because I did not want to deal with https for now, and cors doesn't let me mix them. Anyway, in the end I just decided to implement html serving into the backend.

I have successfully finished the app with all of the originally planned features, I still have some time to spend with something, I might just write some documentation.

I have finally seen success, and was able to send and receive the lines properly, but it's jank af right now. Refactoring into separate components was a good idea for readability, but it posed it's own challenges. I am now working on getting the game launch as smooth as possible.

I have been fighting with sockets only to realize I haven't even reloaded the backend. Whatever... It's a bit difficult to write devlogs as the app needs to work to show off anything at all, and that has been a challenge. Anyway the coder side seems to be right, but the next devlog will definitely go into more detail.

I have been messing around with socketio in vue, and an express backend. Some (more than you think) of my time was spent using the wrong option in a function and not realizing pinia is not reactive.

I was working on smaller features and some ux tweaks.

I have made workshops publicly discoverable by anyone who confirms their email. You can join public workshops or view the ones you will participate in.

I made the invite system send out correct emails (if the app is shipped). I will ship a release as soon as it is in a more proper working state.

Mostly backend work, invite system

Deployed the app properly with nginx, and https on a proper port. Made an invite stub, but no content yet for it.

I've added boost pads to the game, and made the controls remappable and saved them in local storage.

I made the frontend backend connection work. This is a bit late, I recorded it on saturdaym but forgot to upload

This time I implemented proper shooting, the real assents and tuned the gameplay a little more including a floating health bar.

Base of the game written in pixijs, familiarizing myself with the engine.

Porting is mostly going well; new stuff.. not so much.

I wasted some time; in a nutshell. But anyway, I made the backend for listing and creating workshops, and started the migration to tailwing unsuccessfully. I'm really in the figuring out phase right now.

I transfered the project to nuxt, finally going full-stack, made the login work with cookie based auth, reworked the routing

I made signing in and up possible, but please don't set real passwords, cause it logs them still. The dashboard also had some work done on it, it now has a layout at least.

I've started learning vue.js, and figured out the basic style for the website.