A couple weeks ago on a Wednesday morning I posted my current project PaperTalk on HackerNews here.
To my suprise, it went kind of viral and made it to the front page, racking up 319 points and 208 comments.
Here’s what the analytics looked like:
My system wasn’t prepared for the traffic; I didn’t realize Supabase (my DB and auth provider) rate limited sign ups to the shockingly low number of ~3 users every hour. I was able to fix this relatively easy by integrating Resend as an SMTP server. Lesson learned for next time. Despite being bottlenecked in this area during peak traffic, I got about 60 people to sign up to PaperTalk. The feedback was surprisingly positive.
Since that post, I’ve fixed a lot of the issues people brought up:
- broken mobile UX
- sign up / login issues
- various UI improvements
- sign up rate limiting
I also added a few things:
- search
- tags
- more papers
- more categories
- server-side pagination
I don’t know if PaperTalk is going to actually take off. Solving the chicken and egg problem of community-building is hard. But even if it doesn’t go anywhere, taking off on HN was an encouraging signal that I can identify problems people resonate, and I can build systems on my own that are good enough to get some traction and validation.
I’ll focus my time and energy on this project for the next 2-3 months and see where it goes. I have other things I want to build and work on but for now this is the priority.