Stats

3
Projects
39
Devlogs
60
Votes
3
Ships

Coding Time

All Time: 507h 2m
Today: 0h 2m

Member Since

July 03, 2025

Badges

2
🚢
Maiden Voyage
you shipped your first project! the journey begins...
🕷️
Spider
this user has a pet!

Projects

3
NestHelper

NestHelper

Shipped
17 devlogs • over 1 year ago
Pindle (Guess a PIN)
12 devlogs • over 1 year ago
AtomChatDB

AtomChatDB

Shipped
9 devlogs • over 1 year ago

Activity

did a lot of bugfixes between the ship to get it to even compile for windows (unfortunately i don't have a very good windows test box only a windows vista crapbox) so i did fix the shipwright feedback !!!

Update attachment
Luckmuc
Luckmuc Shipwright gave you feedback
443d ago

NestHelper

Please update your project so it works completely and not only partially, nice project :)

from Luckmuc

Completely also added JavaScript tasks too! I also did a lot of other things, like removing all unnecessary logs and fixing the fact that flows and command history didn't save (due to a problem with saving arrays using tauri stores). The JavaScript tasks include a whole user manual to indicate all the special sauce that they include, such as log statements and fetch functions that can fetch even if CORS is an issue.

Now, users can make custom flows using SSH commands. While the ones I demoed in the video seem to be (and are reasonably) useless, this does mean you can automate more mundane tasks. For example, if you need to update your server from a git pull, instead of doing a systemctl --stop server && git pull && pnpm install && pnpm run build && systemctl --start server you can just make a flow and run it whenever needed. This usecase makes sense to me because it's the reason why I'm making NestHelper.

Finished writing commands, which was quick, since all the infrastructure already exists: it's just a matter of getting it all shown in a new interface on the app. Now, I'm probably going to work on the final feature before the first ship, which is the flows feature.

This is the next step: just running commands as needed. While I wanted to put a normal shell, realistically rewriting a program that already exists wouldn't really do much for the user experience.

Got the uploads working!!! Why did this take so long? A problem that I hadn't considered was that SSH will just not even try to take your request if you send like 40 or 80 kilobytes of data through arguments. This isn't a problem when you're editing text files, but when you start to upload binary data, this problem becomes apparent. To fix this, uploads now rely on DD (yes the Disk Destroyer) and write the byte content using STDIN. This wasn't compatible with the existing flow logic, and I was not going to rewrite the existing flow logic, so Uploads are now frontend tasks that just call a command designed to upload files like this. 👍

Basically finished up the filesystem portion of the app, added an image viewer and implemented downloading of files, and you can mark files for deletion. after updating the filesystem code with an upload function, this portion should pretty much be done.

sometimes the simplest things are the most annoying: so I spent the last about 3 hours trying to implement cancelling a flow or command while in progress. The solution I came up with was to just hook into a system api and perform an unsafe SIGKILL of the underlying SSH process. if there's any data in flow being overwritten, your data is probably gone, but I guess you wouldn't cancel unless you knew the process occuring is something that doesn't need to occur. And unlike SIGKILLing a normal process on the client side, SIGKILLing SSH doesn't do much and the server can still recover normally.

woah this is cool. Just got the ability to commit file changes working. You can now edit as many files as you need and flush them out all at once. You can create new files as well. This also has the support of making sure that if a file is deleted on the server it asks you before updating so you can save your data. SO that's cool.

I got the text editing working so now you can edit files !!! Also added syntax highlighting courtesy of Monaco. Monaco is a bit heavy but since NestHelper's a desktop app anyway it doesn't seem to matter anyway.

soo this is the Filesystem browser. it took a while but is basically functional in all that it is meant to do, which is browse the filesystem. I'm probably going to import the basic syntax highlighting javascript component from codium to get text editing working, and probably also add a hex editor. image viewing should be easier, just encode bytes as a base64 string.

Got the dashboard implemented, so now you can see some important information about your current nest configuration. the debug box will probably just be for seeing how many caddy domains you have set.

Refined the SSH flow to be more modular. This should now work better when users make their own flows. First though, I'll include a couple dashboards and editing areas.

Update attachment

i didn't expect to make a devlog about this but it did take me 3 hours: i added the ability for the client to do things with data as part of the flow (for example if you need to see existing data in your own flow). basically the ssh session pauses until the client's promise resolves.

(by the way why is rust so convoluted...)

Update attachment

Due to my significant inexperience with Rust, It's been a while. However, I did just get code execution and output streaming working!!!! This current flow as shown on screen is just a demo: real flows will most likely be less hardcoded so the user can do as pleased. But the flow works after hours of learning about stupid async and move.

Got the basic flow setup working, so you can get your username. Next, I'm going to include a startup command flow that says what to do on startup of the app (so ssh hackclub.app, nest caddy list, cat Caddyfile, etc.)

Started this new project, and I'm using Tauri for the first time. NestHelper is gonna connect via SSH to get tasks done in a way that you don't have to wait for each action to slowly complete.

Update attachment
atomtables
atomtables created a project
467d ago

NestHelper

A helper tool for hack club’s Nest that makes doing actions easier with the help of a GUI and caching and a separation of input from connection to make tasks go by faster

NestHelper
17 devlogs 3 followers Shipped

fixed up the readme and stuff and had to log the last 13 minutes.

Update attachment

took a lot of time and a lot of pain, but just got the leaderboard code working. there's now also an express server that handles this mess. !!!!!!! ready to ship soon.

so i lied... just added high scores to the game. this had a point though: now i have a function to compare scores which will help implement the leaderboard.

Update attachment

Implemented the different gamemodes and they seem to be fully functional. I also added some more statistics, so now you can see how much time you took to solve the puzzles. All that's left is the polishing and the Daily Leaderboard feature.

Update attachment

waow very cool so like check this out... RANDOM NUMBER GENERATOR JUST DROPPED!!!.

now the special thing about this random number generator is that you just call it via new RNG(difficulty).get() and it gets you the value. was all that syntax necessary? no. was it remotely useful? no. does it look cool? yea thats what matters lil bro 😤.

this game is actually starting to come together, especially since difficulties now work. after implementing the game modes, which might take a while, I also want to include a daily feature with a leaderboard so everyone can play together.

i completely forgot to put a devlog in 3 hours ago... so BIG DROP 💦 💦 (watch the video)

there is now a background that looks pretty decent (and is the default for the era of android being imitated here).

you can now select a certain difficulty. I even included helpful tips as to what should be shared.

there's also a functional sideswipe icon that looks cool and even has the expanding dome.

you can also forfeit a game. why? i don't know. but you can if you really really want to.

Added an instructions screen so you actually know how to play the game. I'm thinking it would be better to show this as a notification on the main screen that you can interact with.

Also, the clock might be too unnecessary. The game seems more like an Android simulator with the clock, so I'm thinking I might make the clock just say P1N:DL3 or something fancy like that

Update attachment

just got the entire gameplay flow running. entry screen => gameplay screen => result screen. so technically the game works. it's now just a matter of polishing, adding choice, and finishing up the remaining features planned like a leaderboard.

4 difficulties plus some toggles sounds good, but to ensure there's a way to measure leaderboard statistics, there will also be a Leaderboard game mode where everyone gets to guess a certain difficulty.

This is the gameplay screen fully functioning. Inputs, animations, and transitions now look the same (or at least very very similar) to how Android look(ed).

I also decided on the game logic. For the best result of difficulty, Pindle follows a similar system to Breaklock, with filled dots representing characters in the right place, open dots representing characters in the wrong place, and no dots representing wrong characters. This provides a good challenge, and adjustments can be made as more people play the game.

Just finished the gameplay screen. Unfortunately, there's a problem with mobile platforms where the web browsers don't seem to want to actually make an attempt at rendering transitions.

Had to use javascript to stop Safari from doing anything with touch start events, preventing zooming in, and also forcing the manual active state on the buttons. Desktop web browsers are a lot better at this, so they can just use the default of whatever is needed. Mobile browsers suck though.

Update attachment

Just added the unlock animation (this is very necessary). I guess the main aesthetic of the game is that it's like a real phone, so this is pretty good work for now. Once the PIN code pad gets animated in, gameplay mechanics should be pretty simple.

As it stands, there should be 4 difficulties. Easy, so 4 digit pin, Medium, so 6 digit pin, Hard, so 8 digit pin, Impossible, so any length of digits, and Timewaste, so no feedback and any length of digits.

There'll also be different modes, so maybe a free mode, where it's just guessing for as long as it takes, Timed mode, Tries mode, and whatever else seems like a good gameplay mechanic. It definitely has to be different from wordle, since feedback on placement for each individual digit would give away easy, make medium very easy, and make hard not that hard.

So Pindle is this concept for a game where instead of 26 letters in specific combinations you have 9 numbers in 4/6/8/impossible length. The main point is to make the game LOOK like it's fun. So far, this is what the starting screen looks like. It's themed after android.

Update attachment
atomtables
atomtables created a project
482d ago

Pindle (Guess a PIN)

Imagine wordle but instead you're trying to crack a PIN to unlock a secret phone. Congratulations. That's the app. Inspired by Breaklock and wordle. Check readme 🙏.

Pindle (Guess a PIN)
12 devlogs 1 follower Shipped

oh my god... turns out going from pretty much done to mass production ready takes a lot of time - Linus Sebastian. Server hosting was a nightmare. First, image hosting completely broke because the build folder had a different hierarchy. Then web sockets was broken for a week because Caddy wouldn't accept same-origin requests to connect to a websocket. Finally changing domains, this ship is finally ready for human consumption. Enjoy!

Update attachment

Funtime! Just got the basics of AtomChatDB fully working and fixed the stuff that gives the app proper usability and demo-ability. There is now a working bot that helps create groups, a bunch of text boxes now work when press enter, and I also made everything a lot more uniform. This version of the app is going to be shipped, so please check out this app because it's very cool!!!!

Update attachment

finished the chatrooms and newsgroups for the most part so almost ready to ship, but the images turned out to be a pain.

The way I was transmitting user images and message images before was stupid... User images were encoded in base64 and attached by the server to every message sent. This would obviously add up to way too much space taken by duplicate base64 images. Message images were also in base64, which made implementation easy but couldn't be cached like normal images.

I ended up redesigning the way images are uploaded, which is that display pictures go through a pipe that compresses images into a jpeg and saves them locally. If a user doesn't have a profile picture, the server hook will auto-redirect to a generated one on-the-fly. This way, users don't need a special key for images, seeing as usernames are supposed to be unique anyways.

For messages, there is a unique key. Images are thrown into the /parseimage API endpoint which checks for authentication before compressing and saving their image. It then returns a unique key that is saved along with the message, which can be easily accessed by the client to display images.

The plus with having files is that the database doesn't get slower and slower. Base64 strings are the same size as the images they take part with (around 20kb for the compressed files I'm using), so there isn't much of an advantage unless I literally cannot store files on a server. Also, web browsers have much better caching now.

Hopefully tomorrow will be the finishing touches day.

Update attachment

no common cold's gonna stop me from grinding that 💰 web socket implementation!!! just got messages sending and receiving so yea that's cool!!!

Update attachment

slight crash: so I unified the SQL tables so chatrooms and newsgroups share the same layout. surprisingly the beta version of sveltekit with websockiets is working pretty well, now I'm working on the chatgroup api for real-time communication.

Update attachment

finished up the forum implementation, now just have to add realtime chat using a pull-request version of sveltekit to get websockets support 😭

Update attachment

sooo just got the logic for deleting and auto-updating!!!! it seems like we got some good progress rolling, and we're 40% of the way done with the app

Update attachment

got some work done into creating forum posts. wasn't all good, had to completely redo a portion of the code to make the code less janky and reliant on the worst sql you'd have ever seen.

Update attachment

just set up guest accounts so you don't need to create an account to participate in chats, but gotta think about how to fix the spam 🙏

Update attachment
atomtables
atomtables created a project
504d ago

AtomChatDB

a usenet-like application where you get to talk with other people in rooms with four dot identifiers. SEE README

AtomChatDB
9 devlogs 0 followers Shipped
atomtables
atomtables joined Summer of Making
504d ago

This was widely regarded as a great move by everyone.