Mastowall is a social wall application that displays posts from the [Mastodon](https://joinmastodon.org/) social network based on specified hashtags. It was written entirely by ChatGPT4, guided only by text prompts.
Try it live: [Mastowall for the BiblioCon conference]([https://rstockm.github.io/mastowall/?hashtags=111bibliocon,bibliocon,bibliocon23&server=https://openbiblio.social))
- **Display Posts:** The app fetches and displays posts from Mastodon based on the hashtags provided in the URL. If no hashtags are provided, it presents a form to enter up to three hashtags.
- **Relative Timestamps:** The timestamps of the posts are displayed relative to the current time, and are updated every minute to reflect the passing time.
- **Masonry Grid Layout:** The posts are displayed in a masonry grid layout for a visually pleasing experience.
- **Responsive Design:** The layout adjusts according to the screen size for better readability on different devices.
- **Navbar Hashtag Navigation:** Clicking on the hashtags in the navbar takes you to the form screen, allowing you to change the existing hashtags easily.
- **Navbar Color Customization:** The color of the navigation bar can now be customized via the `config.json` file.
- **Including Replies:** By default, replies are excluded from the wall. However, this behavior can be changed by setting includeReplies to true in the `config.json` file.
1. Load the application in a web browser. If no hashtags are specified in the URL, you will be presented with a form to enter up to three hashtags and a server URL.
2. After entering your hashtags and clicking 'Reload', the application will fetch and display posts from the specified Mastodon server that include those hashtags.
Mastowall supports URL parameters to easily share specific hashtag configurations and the Mastodon server. Simply append the desired hashtags and the server URL to the URL following this format: `?hashtags=hashtag1,hashtag2,hashtag3&server=serverUrl`
Mastowall may serve as an example of how artificial intelligence can aid and accelerate the software development process. The development of this version of the app was guided by OpenAI's GPT-4, a large language model.
In this process, the human developer posed problems, asked questions, and described the desired features and functionalities of the application. GPT-4 then provided solutions, answered queries, generated code snippets, and suggested optimal ways to implement these features.