Show HN: My Tizen multiplayer drawing game flopped, but then hit 100M drawings

25 points • lombarovic • about 16 hours ago • 3 comments

Hi HN,

I built the first version of Drawize back in late 2016 specifically for a Samsung Tizen OS app contest. I crunched and built the whole thing (including the real-time multiplayer engine) in under 4 weeks.

It didn’t win anything in the contest.

Since it was built with web tech anyway, I published it on the open web in early 2017 just to see what would happen. It started living its own life, and today — 8 years later — the database processed the 100,000,000th drawing.

On the busiest days it’s been 30k+ active users, and storing 100M drawings currently sits at ~3.16 TB.

The milestone moment: I was watching live logs today, terrified the 100Mth drawing would be NSFW. Luckily, the RNG gods smiled and it turned out to be a Red Balloon (You can see the 100Mth drawing here: https://www.drawize.com/blog/100-million-drawings-milestone)

Tech stack (boring but fast):

Backend: .NET + WebSockets (real-time sync)

Frontend: hand-coded HTML/JS + jQuery (no React, no bundlers)

Data: PostgreSQL & MongoDB

Storage: Wasabi Cloud (moved there to save on S3 costs)

Scaling as a solo dev: real-time lobbies + reconnection edge cases + moderation/content filtering. I use content classification models trained in 2021 to filter bad content, and the real-time multiplayer side is mostly highly optimized .NET code.

Happy to answer questions about the “failed” Tizen origin, real-time multiplayer on the web, moderation, or how .NET handles the load.