Bug description
The documentation site at https://docs.nextcloudpi.com is broken in two ways. The following analysis was done by Claude in Chrome, but I have verified the problem myself (as a human haha).
1. Tags shown as raw JSON objects
Under "Filter by tag", each tag is rendered as a full JSON object instead of a readable tag name. For example:
{ "id": 6, "name": "ncp", "slug": "ncp" }
The frontend code retrieves the tags correctly from the API (89 tags found), but fails to extract the .name property and instead dumps the entire object as text into each <li> element.
2. Articles stuck on "Loading..."
The articles list never loads. The browser console shows sections Array(0) — the sections array is empty, meaning the article data is either not fetched or not correctly processed. The page remains in a permanent loading state.
Technical details
- The site fetches data from:
https://docs.nextcloudpi.com/discourse/c/ncpdocs.json
-
- The API returns data correctly (topics and tags are present in the response)
-
-
- The frontend JS bundle (
assets/index.1bd6da33.js) appears to expect a different data format than what the Discourse API currently returns
-
Steps to reproduce
- Visit https://docs.nextcloudpi.com
-
- Observe that all tags in "Filter by tag" show raw JSON strings instead of tag names
-
- Observe that the articles section shows "Loading..." indefinitely
Expected behavior
Tags should display as clickable labels (e.g. "ncp", "installation"), and articles should load and be listed correctly.