[synotic]

I think that this article is talking almost entirely about shipping. If you consider the start of Facebook or Twitter or Tumblr, a lot of what they had going for them was first-mover advantage and building out a product that people actually used. If Twitter started out trying to build the architecture they have now (http://www.readwriteweb.com/cloud/2011/01/how-twitter-uses-n...), they probably would have never launched.

Re: refactoring and other comments. Of course optimizing and re-working your code for new use cases and scale issues is critical to support your product, but unless people are using it, you're not going to know what those use cases are and what parts of your product you need to scale.

[DougBTX]

He says that his latest creation with "minimal" architecture hasn't shipped yet either, so this is all hot air.