YYSuni
cover

Review of August and September

I have spent two months in a daze. It feels like I have learned something, but it's still a bit unclear. I need to clarify things and get organized.

What I've done

In my last update, I mentioned that I would continue working on code-mind. However, my interest in the project has diminished. The main reason for this change is that the project react18-json-view has gained more popularity and was featured in Ruanyifeng's weekly. In response to user requests in the GitHub issues, I have added numerous new features, including a new website.

Another factor contributing to my shift away from code-mind is my realization that the reasons I wanted to promote it are not quite valid. The user base for mind maps is limited, and adding the code label would make the user base even smaller. Although this reason may seem weak, I find greater satisfaction working on json-view. Users come up with their own ideas and requirements, which I can then implement and receive feedback on. It provides a continuous and positive experience.

Additionally, I have been interested in learning about 3D space and exploring technologies such as three.js and three fiber, And observing how others create 3D websites.

To challenge myself, I set a small goal of porting beautiful blender rendering examples from the internet to three.js. This involves tasks such as modeling and adjusting in blender, applying colors, rendering, lighting, exporting the model to three.js, and then adjusting the colors, lighting, and rendering once again.

However, I have encountered a problem. The rendering that I adjusted nicely in blender is difficult, or even impossible, to reproduce in three.js. Lighting and materials in blender require constant adjustments and fine-tuning of rendering parameters. Unfortunately, manually adjusting these in three.js is extremely challenging. Each change comes at a cost, and it is not easy to achieve the desired perspective. Furthermore, different renderers make it feel like no matter how much I adjust, the result does not look right.

This realization has led me to conclude that rendering in three.js is not recommended. Instead, research should focus on skillfully applying it to interactive presentations, while avoiding the limitations of rendering.

GitHub exercises

Tonight, while casually reading some technical articles, I came across babylon.js, which also boasts impressive 3D capabilities. Perhaps I can explore that as well.

New Plan

I vaguely remember being busy with learning json-view and 3D in the last two months, while also constantly dealing with colds, getting sick, and having diarrhea. Thanks to this vacation, I feel better physically.

  1. Continue to work on completing some json-view functions, and maybe even contribute to some popular small projects.
  2. Study babylon.js , and how to better apply 3d on websites.
  3. Besides front-end development, there might be other areas to explore.
  4. RN and Electron. (Optional)
  5. Perhaps I can write some interesting and popular technical articles.

It seems like I still missed something, so I'll read some articles again and clear my own path.

TABLE OF CONTENTS