5 karma · joined September 9, 2024
Currently building a keyboard-driven RSS reader: https://botscope.org
Interests: Art, Blockchain, Data Science, DevOps, Entrepreneurship, Hacking, Hardware, Media, Music, Open Source, Philosophy, Programming, Remote Work, Research, Startups, Technology, UI/UX Design, Web Development, Writing
Once I was able to make this mental connection, I tweaked my browser automation so that for that particular web property I would use a USA-based IP address. After making this fix, the browser automation was able to run successfully!
I think there may be other uses for BotScope, but that's how I started going down this road. Thx for asking!
Azure Front Door, ModSecurity, Fastly Next-Gen WAF to name a few.
In total, there's about 30+ different products that BotScope can detect. The entire catalog is here: https://botscope.org/abaam
A few technical details you might find interesting:
- Scans are full browser webpage loads. Not just `curl` requests.
- Scans are executed from datacenter IP addresses to closely mimic automated bot/agent webpage access.
- There isn't any kind of pentesting or destructive stuff happening... just pure recon!
- The product has API keys, and a hosted MCP server so scans can fit into agent/CI/internal-tool workflows too (I realize how meta this is).
Curious for your thoughts & feedback!
Does elfeed work well enough that you don't get anxiety that you might be missing a good article?
If we can advance portable specs so different open source projects can interop... that would be better time spent.
Re: Cards. I just posted a tweet on our account showing how you can change from cards, to a regular list view. See: https://x.com/coretext_io/status/1836034921028325615
Our motivation: Using a mouse is the de facto device for the web, but it requires fine motor skills that feels incongruent for RSS readers. Why? RSS is primarily text-driven media, and the human-computer interface should be text-driven too. There's something about using a keyboard that feels more efficient and, consequently, more pleasurable!
Our biz model: After the beta, we'll be releasing a Long-Term Support (LTS) version of the app so that people can self-host CoreText. Our "cloud managed" version will be closed sourced but have bleeding edge features slightly upstream from the LTS release. The cloud managed version will be ~$3USD/month. We want to be as FOSSy as possible, but also want to bootstrap the business so that CoreText can have at least 1 full-time employee. I think this is a good plan, but willing to be persuaded otherwise.
Loose thoughts: CoreText will be staunchly anti-algorithm so articles will always be chronologically-sorted by latest/oldest articles. "Suggested articles for you" is one of the lamest things to happen to RSS readers. We also have a healthy disdain for advertising & government so we won't be doing anything weird with user data. We also have a FISA Warrant Canary in case we're forced to comply to any warrants (https://coretext.io/warrant-canary). Lastly, there will never be any kind of platform lock-in so exiting CoreText will always be frictionless whether you're a free/paid customer. We want to honor Aaron Swartz's original vision of RSS as the most popular decentralized & portable layer of the web.