Two interesting posts, the first from Jeremy Wright, suggesting that web 2.0 companies need to make sure their products can scale, and the second from DHH, creator of Rails and Basecamp, suggesting that they don't need to worry about scaling, but instead should worry about getting users.
So which is right? The last one to speak?
Actually, in this case, yes. Within reason, web startups really don't need to worry about high-availability and scaling before they have users. DHH summarises the key takeaway of the discussion in a comment on his own post:
Not worrying too much about scaling doesn’t mean building Basecamp on an Access database with a spaghetti soup of PHP making 250 database calls per request. It simply means having a profitable price per request, chilling the f*ck out, and dealing with problems as they arise.
Ultimately, scaling, improving performance, increasing uptime, etc... all those are problems which have a whole catalogue of known solutions. Unless you build your app in a truly stupid way, you will be able to scale it. Building a product that users love and want to pay for, on the other hand, is much harder to achieve. So focus on that first.
If you read this far, you should follow my RSS feed here.