📰 News

News is a minimalist RSS/Atom aggregator that saves to HTML files.

📂news
  ├📰 index.html
  ├📰 page2.html
  └📰 page3.html

That's it! No database, no configuration files, no HTTP server, no ads, no tracking and no JavaScript. Everything is stored in the HTML files which look like this:

Usage

Running news creates 📂news directory containing a sample 📰index.html file which you should edit with your own RSS/Atom feed sources.

Every 10 minutes it fetches news from your feeds and saves what's fresh to 📰index.html.

When 📰index.html grows large (1000 items by default), the oldest 500 items are moved to 📰page2.html.

📂news can reside in Google Drive or Dropbox for easy access everywhere.

This is how I use it:

news -wait 30 -dir "/mnt/d/gdrive/news"

Command-line arguments

news -h prints:

  -dir string
        directory to store html files. By default ./news is used and created if necessary
  -items int
        number of items per page.html file. A new page.html file is created whenever index.html contains 2x that number (default 500)
  -noflood int
        minium seconds between calls to same domain to avoid flooding (default 30)
  -opml string
        path to OPML file containing feed URLS to be imported. Existing feed URLs are ovewritten, not duplicated
  -template news/feed/template.go
        custom Go html/template file to use when generating .html files. See news/feed/template.go
  -timeout int
        timeout in seconds when fetching feeds (default 10)
  -verbose
        verbose mode outputs extra info when enabled
  -wait int
        minutes to wait between updates (default 10)

Running from source

git clone https://github.com/ww9/news

cd news

go run main

Downloading binaries

Windows, Linux and OSX binaries are available in Releases.

How to build releases

Install Taskfile (https://taskfile.dev/installation/). I did it like this:

sh -c "$(curl --location https://taskfile.dev/install.sh)" -- -d -b ~/.local/bin

Then run task release to build binaries for Windows, Linux and OSX in a directory called release.

Todo

It works great but we can always improve! Here are some ideas:

• Parallelize feed fetching
• Replace logrus with std log
• More tests
• Vendor
• Dockerfile
• Serve over HTTP if started with -http flag
• Colorize HTML lines based on retrieval age. So we know just be looking what is new.
• Put retrieval date in tooltip.
• Move <style> bellow body so important and editable info stays at the top of index.html
• Add this repository to the bottom of the page.