@pkncoder0Just finished fixing the issue with the shadows: I was compiling the wrong shader code and linking it 😭.
Currently, uploading, compiling, checking ids, and outputing debug shaders are dealt with with a quick and easy copy-paste method, and I just overload the constructors for the shader class. However when I did this for the new geometry shader it was late, so the copyed code was posting very wrong stuff. 🥀
Here's the git commit, look in the Shader.cpp file: github.com/pkncoder/Engine/commit/a74511d6f27f…
@pkncoder0----------------- Devlog XIX -----------------
\- 🌲 Textuers:female_fairy: - :female-technologist:*Shadoowwsss* 💀-
> To recap, my project is a custom made rendering engine that focuses on diverse rendering options along with high stylistic control (very high-level explanation).
✨ What did I do? ☀️
To recap on what I did:
- Got the program to launch to black screen
- I needed to inject the Asset Manager to a class
- Got objects loading correctly without shadows
- Got textures to load
- Started work on fixing the shadow map
Nothing too crazy has been changed, the texture issues were about some compatibility problems and not uploading to the same index.
The shadow values are uploaded to a 3d texture, where to calculate if an object is in sight or not, we commit triangles based on their position. The problem now is about the far plane, and hitting triangles that weren't in shadow.
I'm not sure what the problem is, but I'm going to do a bit more looking into it.
---
:miku:Thank You for Reading:3-blahaj-spinning:
I still have to fix the shadows issue (obviuosly) but then I can move onto formatting the data (putting it into buffers, creating wrappers, etc).!!
Stardance Devlog: stardance.hackclub.com/projects/15129/devlogs/43917
Github repo: github.com/pkncoder/Engine/tree/Phase-IV
Project landing page: pkncoder.github.io/Engine
---
❓ What's next?:octagonal_sign:
1. Write the shadow cubemap geometry shader & ad it to the OpaquePass (temp main pass)
2. Make and debug UBOs replacing base uniforms
3. Refactor / clean
4. Create the real Render Graph?
---
🥀 Git Commits (shortened)🗿
Commits: github.com/pkncoder/Engine/commits/Phase-IV
📈✈️@QuackX0📔 New Devlog BoostedAuthor:@_got.DDR5.RAM_Title: Omni-Core Summary:
You are browsing as a guest.
Sign up
(or log in) to start making projects! Progress:
• I made a fully-local AI-powered student assistant. Like it has EVERYTHING from Textbook Analysis, a Priority Matrix and Web Research to Music Streaming and Music Downloads to Auto-Meeting Transcription and Observer Mode and MORE. Like, building this project was a nightmare in debugging as it used Tauri, Rust and React, and shifting between debugging one to the other is hell as they are VERY DIFFERENT. And the most powerful thing is that this app is like that one teacher who refuses to stop observing you guys, so that you guys work. This constantly records your screen (bit like PiecesOS or Windows Recall BUT I don’t ship all your data to a server unlike Pieces) and gives you statistics on what you’ve been doing, and your entire data including Statistics, Goals, Tasks, Textbooks and Notebooks and MORE are sent to the local AI models (btw, yeah, you have to have Ollama set up with the 4 models mentioned in setup) so that it has your entire context. And all this with no telemetry, privacy risk or internet needs, it can run COMPLETELY LOCALLY (music too is downloadable hehe). Also, please read the GitHub README as it holds all the features, I have only teased it here. Also, btw, this wasn’t made in just 10 hours, that is just the time I had hackatime on (and I’m fine with getting 10 hours worth of Stardust as it is my fault for not having hackatime), I just want you to know that 60+ hours (3-4 hours per day including research) of MY work was put to this and that this is NOT just a scrapped project that over-claims and under-delivers.
• Yea, I just added a Start-up page, and also finished up with all the core features. (I spent way more than 3 hours lol, don’t think I just sat and built everything in 3 hrs). Anyways, try it out at: github.com/Koundinya-Git/omni-dashboardNext Steps:
• All devlogs, followers, and associated data will be removed. View Devlog button🚢🔬@Kaplin867+Day 222 / 365
Yeah i got nothin lol. I really didnt do much. I slept in, played a bit, worked a tad bit on Router Rebellion, and talked with online friends for most of the day. Im also contemplating how i want to code the attack system in Router Rebellion. Specifically with moves, and their secondary effects. I have the base move stuff setup, so you can target, have multiple targets, target allies, etc etc. things that are on every move. But now im considering how i want to code the additional effects. Theres 3 main ways: 1. Every needed combo has a class that inherits base move. The issue with this is that it might get complicated quick, and i wouldnt have a good way of checking things like "if move applies burn". 2. The move class has a lot of variables. Like there would be a damage variable, then a burn boolean, then burn count, etc. This would probably make giant scripts. Easy to check if a move does a certain thing though. 3. The moves have an array of "application" objects. It would run through each application object and apply it to the target. For example, damage would be one, alongside burn, etc etc. but each would be different objects with unique code. Easy to check for specific actions in a move. Theres not too many issues with this one except the fact it would probably take longer. and maybe harder debugging.
@pkncoder0I have been working the past week to get the Event system working & trying to fix pointer / referance issues (there are still a few I haven't figured out).
But now I have a working event system. To recap, my project is a custom made rendering engine that focuses on diverse rendering options along with high stylistic control (very high-level explanation). In my previous devlog, I explained the Layer Stack, a top-down structure for controlling object ownership and main loop execution. Every frame, the layer stack runs the first layer to poll camera movement inputs (temp), and then next layer is ran to render out the frame through the RendererManager.
This is great, but where the layer stack really shines is the event system. Events are reactions to when anything happens, keys being pressed / released, the window resizing, etc. When an event is dispatched, the layerstack stops, loops through each layer running it's "onEvent" function to let that layer react (ex. the final render needing it's texture resized). If a layer doesn't want below layers to react to the event, it can consume it, stopping the event dispatching and continuing execution.
I struggled a ton with the event state not updating correctly, and the camera not giving the right values to render the triangles. It turned out to be objects/classes duplicating, and not being able to update. There is so many debug messages in the terminal 😭
This was all front-end changes (no change in the render), but now there is a better system put in place (back end changes).
Github repo: github.com/pkncoder/Engine/tree/Phase-IV
Landing page / Website: <https://g|https://>pkncoder.github.io/Engine
@SuzyQCable0okay so i found an lg webtv app (in dutch) and looking into it its very janky but they left in... a debug console!!
after a bit of finagling i got it working, for nothing to show up
turns out even if i set the debug vars to true, it still wouldnt let it log anything, so i just changed the function to have it always log to the element
yes, it uses green courier new as the font and font color.. eh, ive seen worse.. but it works!!
kinda pointless but i like the UI and everything in it, gonna see about repurposing its CSS and images in it for a jellyfin UI mod, idk i like how it looks okay
unfortunately, the UI elements are mostly all images but its nothing a bit of photoshop magic cant fix.... maybe ill log it with hackatime
thatd be a good idea
💖@Kaplin867+Day 134 / 365
I lowkey didnt do all that much today since I slept in + had to immediately go tutor at the local church. this is the last day before the group we tutor has summer, so that means last day. I coded one hour on my horizons game to keep the streak (i want the holographic stickers they have for doing a 5 day streak right now) and in that hour i was sorta just debugging the custom dialogue editor. I was having some strange issue where godot wouldnt let me edit an array of a loaded resource? strange. anyways the way i fixed it was just creating a new empty dialogue holder resource and loading all the information in manually. Seems to do the trick!
🎮☀️@Henry Wauzivuff018 April 2026, Saturday, Day 108/365. Sabatho. (Today)
---
9 PM
• Slept late, woke up late :sadge:
• Im tired
• Made :oh-my-god-bruh-jjk:
• Went on a walk until 5
• I saw a scooter on top of a bus stop this afternoon and I cant tell if its normal
• On card game:
◦ Fixed debug menu
◦ Fixed pause menu
◦ Fixed z_index ordering
◦ Trying to add difficulty
Enjoyment: 2/5
Productivity: 3.5/5
Streaks:
1. Coding: Yes
2. Art: No :<
3. Documentation: Yes
`██ boundaries remain unset, at old friends ██ upset🚌🎨🎮@n0o0b090lv0DAY 36/:50days::
• FIRST DJing :DD 🎧
• Made new pfp :noobhead:
• Fixed a lot in MUT4rr#
- Made hp bar (may or may not be debug) - Ammo - Fixed item use not working - Made raycasting better so cursor stops (it aws bit bugged XD) - Use cooldown - some other bs and bug fixxing XDD
• Got 3d fillement (whole 5 rolls + starter kit XD)
• Started looking for a dj thingy :))
• Some more art in #C09RCEV6YNR|
• ALSO 57/60 IN #C09BE8QF1V5| (2 people joined :yayayayayay:, sad that wokflows are down ;-;)
• Distributed Christmas mail in my school :)) 🎁
• Some more things i did today, but these are of top of my head XD
BIG OH SO BIG DAY IDK HOW I WILL SLEEP NOW :xxdd:
🎨🎄🎒💤🐛@ajhalili20060DAY 31 (2025-12-12):
• Literally tweaking around Atlassian Home app for project and goal updates on both personally and at Recap Time Squad (#C07H1R2PW9W|), although I’ll be sticking with Obsidian in the future.
◦ Guess we’ll going to have a field day with Confluence and Jira (💀) during the holidays break
• Actually replied back to Doppler Support regarding Cloudflare Access SSO related errors
• More debugging hell at github.com/recaptime-dev/devenvkit why the build fails (more GitHub Copilot chat threads, I guess?)
@SakG0Day 22:
• Just forgot about AOC, will do!
• Did coding for 5 hrs
• Got rejected for #C09DLGM81E1|
• Got approved for #C0A1GEUM8R3|! Looking for participants and team.
• Did some debugging and found out the problem and quick fixed it..
• For complex formulas, relied on LLM, not anymore..
• Didn't study
Day 23:
• Code for 10 hrs(I promise i will)
• will fix out the bug in uconnect today and make the cursor effect and wrp up homepage
• Minimize external help for debugging
• work on visisphere for atleast 30 min
• Do AOC fasttt
📚🐛@glitch_purge0Log #3 @ Scrapbook about Mosaic!!
So far so I have
• Got my website to save the projects you have added to your slack account (aka when you sign in, it loads the projects you added last time/session)
• Spent 3.5hrs debugging Hackatime to get to display the hours on dashboard once again because of the saving feature I added tdy
And for those of you following Mosaic, DW theres gonna be a huge CSS revamp on this - this is just my prototype for development!
Next steps:
• Leaderboard System
• Submission System
• Admin System
@SakG0Day 1️⃣
• Was making uConnect hero but encountered an err that took my day(debugging even now :cryin:)
• :ms-cross: Complete scholarImpact (for hackathon)
Day 1️⃣8️⃣:
• Make uConnect Hero section and tour part
• Study 1 chapter
• Code for 6 hackatimes hrsssss
📚@hex40(re-posted from bsky.app/profile/hex4.xyz/post/3m6ghh46jyc2g)
:50days: day 14!
hack club slack is down :( but that means I got a lot done for mixtape! i set up a sveltekit route to hit my AI proxy, wrote a system prompt, did some frontend debugging, and Mixtape is now able to write Strudel code all by itself!
🖋️@KittyCat0:50days: day #6:
• I did some more godot debugging and I’ve decided to just start over with a new project but make it more organized and enhanced (the current one is a mess: github.com/AlanikREDAWN/moonshot-dragonpedia) :heavysob:
• I did some crochet with the yarn I bought
• I finished a DE English assignment thats due tomorrow
• I’m going to read more of Shatter Me before bed! I’m liking it so far, its pretty fast paced and hard to put down
🎒@freddie0Ducky Diagnostics is a project designed for the PicoDucky (but just python for now)!
It runs a series of commands and saves output to debug.txt with some formatting and voilà, a detailed list of spec's and more information about your device. You can download it from this GitHub link. (this was made for YSWS #C08QZ0P7QP6|)
🚢🗣️@snej550Finally got IBL (mostly) working in my OpenGL renderer. I spent most of yesterday debugging some weird artifacts I was getting, but I'm pretty sure it was just some crazy high values from the HDR map I was sampling, because it works fine now after I switched to the hdr map used in the learnopengl.com tutorials. (So I just need to add some clamping to the values I think).@MinglangDu0Used a stress-tester to debug my solution to the Closest-Pair Problem!
@py6600I finally found the registry, after 10 hours of continuous debugging (well, debugging for an absolute beginner at Java)! 🙂
@Irtaza0After digging through Sprig’s schematics, figuring out the right CircuitPython libraries, and spending countless hours debugging errors in Thonny’s serial shell, I finally got my <https://github.com/Irtaza2009/Hackamon|Hackamon code> (submission to #C0809PN4TPE|:hackapet:) running on an *Orpheus Pico* on *Sprig*:sprig-dino:!:yay:
🦕🦕@Irtaza0After digging through Sprig’s schematics, figuring out the right CircuitPython libraries, and spending countless hours debugging errors in Thonny’s serial shell, I finally got my Hackamon code (submission to #C0809PN4TPE|:hackapet:) running on an *Orpheus Pico* on *Sprig*:sprig-dino:!:yay:
🦕@ChrisB0Well, that was a nightmare.
I have spent the last hour and a half debugging what turned out to be a single line of code, but other than that, today i have started work on the gpu conversion of my walkers thingy.
Basic rundown of how it works.
I store the data for each walker in a texture, I keep their position and time alive, but not much else. In my case, I store the x position in the top left of an image, (see below), encoded as a 32 bit float (using all four channels because webgl 1.0 only supports 8 bit color channels), the top right keeps the y position. In the bottom left, I keep the time alive (in frames) in the r channel, and whether to render or not (as a bool, 0 or 1) in the g channel. The bottom right is at this point unused.
Each frame, I take the previous's frames texture, and render over top of it, updating using a fragment shader (which is not yet written :frown-wx:)
Then, I pass that texture to another vertex shader, which generates a bunch of points based of of each pixel in the texture.
In theory, its super easy, but in practice... not so much.
Encoding the float over the four color channels was not fun, as I kept getting a normal distribution of points even when the initial texture was completely random (and uniform, I checked). I spent a couple hours trying to work that one out only to discover I hadn't unbound one of the webgl globals, so that was fun. Calculating the uv coordinates based only on the point index was ok, but implementing a ping pong frame buffer for the texture update frag shader was a bit tricky. Tomorrow I'll try and get this working, so I've got them functional and moving.
Below:
Computed texture
Rendered walkers (from computed texture)
Colored renderered walkers (r: posx, g:posy, b:index)
@jollyroger1820First Scrapbook post! Today I worked on this auction website for my school. Of course that .prevent right there will take me 1 hour of debug time :pain-gold:
@Nethul0Day 15 of #C045S4393CY| I really wanted to show the redesigned clock app today, I worked on it a lot, but my code completely stopped working around 2 days ago and its showing no errors, I spent a few hours today trying to fix it adding debug prints everywhere and more, but I can't seem to fix it, but I'll keep working on it, but there will not be a big reveal. I have been using try and except blocks but somewhere along in the code something has to be happening, and if it's really small, I might have to add back code function by function and see what happens. I did work on the app some amount in a different program for testing so I could integrate it but again, it needs other functions and all so I am focusing on making sure the phone does not stop right after starting.
@Nethul0Day 14 of #C045S4393CY| I am expecting to finish the clock app redesign tomorrow! Today I worked on functionality and logic. But I have a huge error right now that stops my code from opening, the code does not even show a single debug statement which is a huge problem, it suddenly started happening and I have to fix it so I have been trying. Don't have much to show though.
@HridyaAgrawal0Day 7 #C045S4393CY|
very WIP site, added ADS on uzi, fixed muzzle flash, also i tried to debug hands more, but its just too hard, i might try another method or just give up w hands@Nethul0Day 6 of #C045S4393CY| Today again like yesterday was a debugging day, I don't have much to show visually but I was able to make the code a bit faster, not crazy faster but a bit faster, and I am still working on connecting apps and building prototypes of code to integrate the AI with the clock app. This should help everything work better when I do implement new features, because sometimes if the setup is not nice everything breaks down and it becomes a mess
@HridyaAgrawal0Day 5 #C045S4393CY|
added Uzi, no animations yet, but configured it so it works + SMG ammo, also tried to debug a bit with why the IK wasnt working, also found a bug where zombies take time to register that they're dead🐛@AkshayShah0Day 4 of #C045S4393CY|. Spent around 1hr debugging and moving through code error by error. Started to figure out what it is analyzing. The code is now analyzing the waves but still working on the decision making. github.com/Ventengo1/EllioAnalysis
@AkshayShah0Day 3 of #C045S4393CY|. Today I worked on continuing to debug the code and make changes so that now I at least get to where it says that it is loading and going through the waves. Still experiencing errors after that though. github.com/Ventengo1/EllioAnalysis/tree/main
@MinglangDu0Added a little debug window to my neural network test. It is going to help me fix many issues! youtu.be/k6WEUDrnXCc?t=8
Edit: added an image to placate scrappy
@snej550Exam week is finally over! Today I added some coins to my game, with some more vfx, some more sounds, font rendering (with SDL_ttf), some ui improvements, leaves (particles that fall from the trees and stuff, kinda hard to see with all of the coins), and did some debugging (another seg fault waah). Next I am planning to add some lava, music and a main menu (finally).
github.com/snej55/paper-world
📚🍂🎮@SkyfallWasTaken0launched wakatime for figma in #C07TNAZGMHS|!
also started debugging an issue that a firefox user was having (which i could not replicate at all :pf:) (also there are no errors :heavysob:)
@dj1ch0Started working on some method to automatically fill in some blanks for structs in case I run into stack exceptions. Development is overall going well but I hate having to go through a file that's a thousand lines long to find a specific variable, function call, etc. It's getting less complicated, and debugging everything isn't too arduous of a task.
Link: github.com/dj1ch/HomeNet
@dj1ch0Did some more debugging today, but I've mostly just been resting. I think I can get this done by the end of the month hopefully.
Link: github.com/dj1ch/HomeNet
🚢
💖
📚
🖋️
🎒
🦕
🦕
