—————– Devlog XVI —————–
> 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? ⭐️
For the quick overview:
• Dealt with an issue in Obsidian, as apparently there was “minor rendering performance improvements.” This is FALSE :( - I just downgraded the version
• Created the new BufferManager to handle UBOs, SSBOs, VBOs, EBOs, etc. + Dynamic, Static, and Stream buffers.
• Integrated the RenderTarget & ShaderPass system from PathTracer straight into IRenderer + added base utilities to it.
• Did some formatting
Overall the two big things that I did was the new Buffer system & the updated IRenderer interface.
For the new Buffers, previously I had UBOs (dynamic values) & SSBOs (big mostly static information blocks) set up in seperate files, that BufferManager returned. Now, every buffer is a single struct called GPUBuffer, and GPUBuffer has a type, and a usage (go look at the images in stardance). Now everything is centralized on that.
For the new IRenderer interface, now there are RenderTargets - textures that shaders write to as output - and ShaderPasses - A pass of a shader. These can be added to a registry, keyed by a UUID (big number, in this case).
:miku: Thanks for reading!!! :3-blahaj-spinning:
Thanks so much @AVD for the Superstar!
I’m excited for when I can finally start writing my shaders, it’s getting closer & closer!
The fancy Path Tracer will be cool and all, but I have big plans for the Ray Marcher!!
Stardance devlog: stardance.hackclub.com/projects/15129/devlogs/37468
Github repo: github.com/pkncoder/Engine/tree/Phase-IV
Project landing page: pkncoder.github.io/EngineDay 221 / 365
Today i felt considerably tired. I even took a nap (which i rarely do). I tried to get myself to work on One Turret, but made barely any progress. I then tried to work on Router Rebellion, which i will say i did find myself putting more effort into. I was trying to calculate some stuff to get the camera to always look good when your choosing a move, but i still need to workshop it a bit. Its funny because im at the level of knowledge with 3d vector maths where I know enough to do simple things and how to execute ideas, but not enough knowledge to instantly now how to achieve the end goal, so i end up just playing around with a bunch of stuff. I even tried this one thing where i had a target angle, and then i calculated look-at vectors (like positions for the camera to look at) based on the current selected character and the middle of the battle screen, then i just iterated through -1 to 1 by 0.01, getting the inbetween value using that (0 = middle of map position, 1 = selected character pos, 0.5 = inbetween, and anything past that is overshoot or inbetween) and tracked which one was the closest to the desired outcome. It still didnt turn out good but it did get me a tiny bit closer. odds are im just overcomplicating it and i just need to set the angle to predefined things and i dont need to calculate anything :)
Today, I started revamping my personal portfolio website. My old portfolio worked perfectly fine. But lately, every time I opened it, a voice in my head screamed: “That’s not you anymore.” It is a hard thing to let go of past versions of yourself. It is even harder to let go of projects you poured your blood, sweat, and tears into. My old portfolio took me months to perfect, but it simply no longer fits the direction I am heading. Growth requires making space for the future. Outgrowing your own creations is not a failure; it is proof of progress. Hopefully, the new portfolio ends up looking even better.Day 220 / 365
Today I held another hack club meeting!! yay!! My club members finally got their IDs so I helped them get the authentication stuff in, and also teaching them about slack. After that, we tried our hand again at doing the 3d modeling for our hackpads we are going to submit to stardance. Its... not going good lol. Its weird, at first there was a guide for it on stardance, but then they made it to where you have to use the one on the hackpad site. only issue is that, for starters, its a different guide so theres some inconsistencies. And the second issue is that the one on the hackpad website is just worse lol. idk like i feel like theres way less content and explanation. Like we are struggling on this one step because we dont know how to move an autodesk sketch, but i suppose thats my fault since i didnt learn it before the meeting. We ended up just moving the rest of the model in a way the sketch is in the right place. Honestly, im considering having us restart. or perhaps switching project. i dunno, just something about hardware feels way more difficult compared to software. especially the time tracking. we didnt even try to track the time today since we didnt do anything substaintial (we literally spent one hour trying to move a sketch). theres also this hack club ysws called greeble which i might try. make a procedural pattern get it printed. Only issue with me participating is i already have like 4-ish projects on my plate (One Turret, Yomi Hustle mod, Stacklands Mod, Router Rebellion) due to my burnout(?) seeking to try new stuff quick, so idk how long id stick to it right now if i were to even start it.
