I love how YouTube approached this: Choose the simplest and most stable tools and use those.
It's a great strategy for developers that want to ship. Otherwise one can easily be lured by the siren calls of new tech.
I've certainly fallen victim to this temptation, but I've found that as I let go of the new pretty things and focus more on using the old boring workhorses of the interwebs, I get a ton more done.
Plain vanilla seems boring - but it's bad-ass in web architecture. We forget that YouTube had plenty of competitors that were well ahead them before they dominated.
They kept things simple and solid and that allowed them to scale. It's certainly not the only reason they won, but I'd bet it's a huge part of it.
It's a great strategy for developers that want to ship. Otherwise one can easily be lured by the siren calls of new tech.
I've certainly fallen victim to this temptation, but I've found that as I let go of the new pretty things and focus more on using the old boring workhorses of the interwebs, I get a ton more done.
Plain vanilla seems boring - but it's bad-ass in web architecture. We forget that YouTube had plenty of competitors that were well ahead them before they dominated.
They kept things simple and solid and that allowed them to scale. It's certainly not the only reason they won, but I'd bet it's a huge part of it.
---
For more on YouTube's stack, see: http://highscalability.com/blog/2012/3/26/7-years-of-youtube...