Hacker News new | past | comments | ask | show | jobs | submit login

To begin with, yes. It also depends on how heavily I denormalize the data for recall. It could be many times that down the road. I'm not looking to implement something heavily distributed right now, but I know that day is coming with my datasets and I want to ensure I know what needs to be built when the time comes.



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: