This book would sell X,000 copies pretty quickly and create substantial value, so if anyone wants to take a stab at writing it... Much of it is not rocket science, but having every SaaS entrepreneur independently reinvent not-rocket-science as table stakes is clearly suboptimal.
Simple things like "You're basically going to inevitably eventually have web servers serving live users and the API, worker processes doing aynchronous heavy lifting, and a job queue. You probably *will not* implement a message bus in the way large software shops would."
"Monoliths versus microservices: Make it a monolith. Done. You won't have sufficient organizational complexity to justify the engineering tax of microservices until you have multiple engineering teams, and plausibly your company never gets there."
"You will use cloud hosting on one provider for the lifetime of the company. Portability is overrated.
You will automate deploys. You will automate testing, including continuous integration relatively early but probably not continuous deployment."
You will automate deploys. You will automate testing, including continuous integration relatively early but probably not continuous deployment."
"You're going to send a lot of email but will never operate your own email server. There are another forty things that could substitute for email in that sentence but that's probably the one you'll hit first and hardest. For all of them you have plenty of options; satisfice."
