The Big Switch: How We Rebuilt Wanelo from Scratch and Lived to Tell About It

We launched the rewrite a few months ago, the site is working wonderfully :).

The facts of this case seem to tilt things in favor of a rewrite more than other other examples I've seen. - reasonable confidence that large portions of the current codebase are truly dead - lack of tests in the existing codebase - a fair amount of real world experience showing higher productivity in rails over java web stacks - a problem domain that the decision makers can actually wrap their heads around (vs a legacy system that manages commercial mortgage repackaging or something where there is no one person who actually understands everything important that the system does) - a new development team who's going to have learning curve anyway

this seems like about the perfect situation for starting over. congrats!

I'd say that if anything could be formalized as a set of skills or as a process that could be applied to multiple projects, it would be extreme discipline around what gets built. For example, we were fairly militant about cutting features and keeping things as simple as possible, which is always easier said than done.

Also, this rewrite wouldn't be possible without each team member's personal productivity which was also aided by pairing. Curiously, we worked fairly regular days (although some nights and weekends, but those were by personal choice).