RSS Converter for MastoFeed

In this post I’d like to share a way I’m streamlining sharing of my podcast recommendations to Mastodon, following the innovative lead of Alan Levine. (CogDog) Striving to imitate (some of) Alan’s digital breadcrumbs on Mastodon, I’m now using using Pinboard to collect and organize links, including a dedicated tag for podcasts I recommend. It’s been a LONG time since I used Diigo and Del.icio.us for social bookmarking! I learned about MastoFeed from ChatGPT, as a way to automate posts to Mastodon from an RSS feed, but when I tried it with my Pinboard.in RSS tag link I learned it uses an older RSS 1.0 (RDF) format. Unfortunately, that RSS feed format isn’t compatible with MastoFeed.

After researching different approaches — including Feedburner (which, sadly, isn’t as flexible as it used to be) — I decided to try building a solution myself using Python. Spoiler alert: It worked, and I learned a lot along the way. You can find my project on GitHub, where I’ve named it, “rss-to-rss2-converter.”

AI “Vibe Coding”

“Vibe Coding” is a term for using AI tools (like ChatGPT 4o in this case) to write and help with the writing of computer code. I’ve documented a few of my past experiments with vibe coding on my “Learning AI with Wes” website. In this case, I used ChatGPT 4o to generate a short Python script using two libraries:

• feedparser — to read the original RSS feed.
• rfeed — to generate the RSS 2.0 feed.

The script fetches my Pinboard (RSS 1.0) feed, processes each entry, and writes a new RSS 2.0 .xml file to my web server.

Lessons Learned

This project reminded me just how powerful — and sometimes complex — working with a VPS can be. It also highlighted (yet again) how powerful it is to “vibe code” with AI! There is NO WAY I would have been able to successfully create this project / accomplish these goals without the help of an AI platform for vibe coding. Here are a few of my key takeaways from these experiences.

• Terminal Access Matters: I needed SSH access to my VPS to install Python packages and run the script.
• File Permissions Count: At one point, my feed wasn’t accessible because the output folder was owned by root instead of my hosting user. Fixing ownership solved that.
• .htaccess Confusion: WordPress .htaccess rules appeared to be blocking access to my .xml file. I experimented with rewrite rules, but the real fix was proper file permissions.
• Learning by Doing: I’m no Python expert, but this project pushed me to stretch my skills with scripting, cron jobs, and server setup. It felt good to get it working!
• Beware of Security Issues: Initially the AI / ChatGPT created code which included direct links to my Pinboard RSS feed and my web server directories. When I realized that I asked for updated, “cleaned” versions… but this was NOT something ChatGPT did on its own. I had to prompt for these changes.

Sharing the Process

To help others facing similar challenges, I am documenting my process through this blog post and some “cleaned up” script files (so no private links or server details are exposed) on GitHub:

RSS 1.0 to RSS 2.0 Feed Converter on GitHub

Feel free to use it, adapt it, or suggest improvements!

Final Thoughts

The open web is built on tools like RSS, and it’s satisfying to stitch together older and newer technologies to make them work for your needs. If you’re exploring decentralized platforms like Mastodon and want to automate link sharing, I hope this project helps.

For those interested in AI tools that can support creative problem-solving like this, I’ve also been documenting my AI experiments and presentations at ai.wesfryer.com.

As always, I’d love to hear your thoughts or questions! And MANY thanks to the ever creative and innovative Alan Levine!