Major updates across html2rss, html2rss-web, the documentation website and configs
#1118
I wanted to share a quick project update, because a lot has changed across html2rss recently.
On the core side, html2rss v0.18.0 shipped with a lot of meaningful progress.
On the html2rss-web side, there has also been a substantial step forward since the last phase. It has grown into a much more complete product, with:
- a versioned API including OpenAPI YAML documentation
- a dedicated frontend
- better feed-creation UX
- clearer Docker/runtime behavior
- improved health/auth/observability handling
- deployment standardized on docker image:
html2rss/web
The docs site at html2rss.github.io was updated alongside this, so deployment, Browserless usage, troubleshooting, and product behavior are much better aligned than before.
I also spent time improving the curated config set, including a new Browserless-backed subset for a few more mainstream sites.
A few practical notes if you use the project:
- if you run
html2rss-webin Docker, please migrate tohtml2rss/web - if you want to try it quickly, start here: Getting started 🚀
- if you self-host, it is worth checking the current deployment docs again: Deployment & Production
- if you use automatic feed generation, the updated guide is here: Use automatic feed generation
Helpful links:
- Core gem release:
html2rssv0.18.0 - Docs site: https://html2rss.github.io/
- Web application docs: https://html2rss.github.io/web-application/
- Ruby gem docs: https://html2rss.github.io/ruby-gem/
- Feed directory: https://html2rss.github.io/feed-directory/
- Deployment & Production: https://html2rss.github.io/web-application/how-to/deployment/
- Docker image:
html2rss/web
The community around this project is still small, so every bit of usage, feedback, bug reports, weird site examples, or self-hosting experience is genuinely useful.
If you’ve tried html2rss recently, I’d be very interested in hearing: 💬
- what worked well
- what still feels rough
- which sites or workflows you most want better support for
Cheers, and thanks for following along,
Gil
All reactions
Replies: 3 comments
Hi,
Thanks for making this awesome project!
I am trying to setup my own instance and going through the instructions for the first time here is some things you might be able to clear up for new not so technical in this side of self hosting like myself:
First generating tokens you do mention a few times in the docs that the tokens are self made but maybe if you added the command to run in the compose file or in step one mentioned to make a few keys.
Second this might just be me but when I had just setup my web instance the page showed up white with no error and I was confused because the logs were saying that I was accessing it via HTTPS when I was over HTTP I do not know what caused this but I had to enable this flag in chrome to get it to work:
image
Third is where I am currently stuck when generating a custom feed via the web UI it asks for a token in the stock docker config and I have no idea where this comes from I have tried all the tokens I put in the docker compose and they are all marked as invalid.
Please let me know if I am just missing something or if my observations are misguided in some way just wanted to add my 2 cents on how the user flow might be improved in the future.
Thanks again for making this amazing project and I am exited to be able to use it soon!
All reactions
I can also provide some feedback after trying this for the first time:
Firstly: I also had problems accessing the Getting Started configuration. I got the following error in logs:
HTTP parse error, malformed request: #<Puma::HttpParserError: Invalid HTTP format, parsing fails. Are you trying to open an SSL connection to a non-SSL Puma? I guess this was caused, because I was not accessing the page from localhost, but deployed it on a remote machine. After setting up a Caddy reverse proxy, it worked like a charm.
Secondly: I'm also confused about the different tokens. I enabled the "automatic feed generation", but when I try to add a page in the web ui, it asks for an access token and I have no idea where to get it from.
Thirdly: I'm confused about the difference of the web application and the ruby gem application. I guess the ruby gem application can be used locally to create the respective yaml config files. How can I then add these config files to my instance?
I think this is a very nice project with potential, but its lacking some overall concepts documentation (even though I really like the existing docs).
Thank you for this project 👍
All reactions
Thanks for the detailed reports, @fabolhak & @Digitalpeer1 ! Both friction points have been resolved in html2rss/web >= 1.12.2:
- Plain HTTP / LAN IP white screen & Puma SSL errors: HSTS and CSP
upgrade-insecure-requestsare now strictly conditional on HTTPS. Accessing an instance viahttp://<LAN_IP>:4000works out of the box without browser flags. - Token clarity: The UI now explicitly prompts for
HTML2RSS_ACCESS_TOKEN(avoiding confusion with the health-check token).
If you set this aside earlier, it is definitely worth giving another spin. Beyond massive scraping and anti-bot performance upgrades, the app now ships with:
- Zero-config auto-source: Just paste any webpage URL to get an instant feed.
- Pre-built catalog: Over 200 ready-to-use feeds included directly in the UI.
- AI MCP server support: Curate and test feeds natively with AI coding assistants.
Pull the latest image (docker compose pull) and check out the revamped guide:
👉 https://html2rss.github.io/getting-started
(A full .env.example and a 30-second docker-compose.quickstart.yml are also available in the html2rss-web repo).