uppy-media

Images and videos for the uppy.io blog and docs. The website itself lives in transloadit/uppy under website/.

Files are served through Transloadit's Smart CDN, which imports them from this repository's main branch:

https://uppy-io-static.tlcdn.com/static/<path in this repo>

For example, img/blog/2.0/uppy-performance-after.gif is https://uppy-io-static.tlcdn.com/static/img/blog/2.0/uppy-performance-after.gif.

Adding media

1. Optimize first: re-encode videos and GIFs (e.g. to MP4 or WebM at web quality), compress images. The CDN serves files exactly as committed.
2. Add the file under img/blog/<post>/ with a new, unique filename, by PR (GitHub's web upload works fine, no clone needed).
3. Once merged, open the CDN URL once to check it works, then merge the website PR that uses it.

Never overwrite, rename or delete a file that is in use: the CDN caches files for a year, and existing pages link to them. Add a new file instead.

CDN setup

Transloadit Smart CDN workspace uppy-io-static, template static:

{
  "steps": {
    "imported": {
      "robot": "/http/import",
      "url": "https://raw.githubusercontent.com/transloadit/uppy-media/refs/heads/main/${fields.input}"
    },
    "filtered": {
      "use": "imported",
      "robot": "/file/filter",
      "accepts": [["${file.mime}", "regex", "^(image/(jpeg|png|gif|webp|avif)|video/(mp4|webm|quicktime))$"]],
      "declines": [["${fields.input}", "includes", ".."]],
      "error_on_decline": true
    },
    "served": {
      "use": "filtered",
      "robot": "/file/serve",
      "cache_duration": 31536000,
      "headers": {
        "X-Content-Type-Options": "nosniff",
        "Content-Security-Policy": "default-src 'none'; sandbox"
      }
    }
  }
}

• /file/filter only serves images and videos, so a merged HTML or SVG file can't become a page on the CDN domain. It also rejects .. in the path, so only files from this repository are served.
• /file/serve caches for a year and sends headers that stop browsers from running anything it serves.

Keep this section in sync when changing the template.

History

The initial files were copied from the archived transloadit/uppy.io repository at commit 67dd419 by scripts/import-from-uppy-io.sh, keeping only the media the website still links to.