I don't actually disagree with that post-mortem. Databases (and similarly, SDKs, widget libraries, and all other sorts of code-only deliverables) are rough to m

I don't actually disagree with that post-mortem. Databases (and similarly, SDKs, widget libraries, and all other sorts of code-only deliverables) are rough to monetize, precisely for the reason that the RethinkDB founders believe. In addition, if you need some capability to build your product, purchasing it always means buying more capability than you need and doing more integration work than you thought you would have to. On the other hand, workflow tools, especially if hosted, do really well and get to charge money. Jira, Github, New Relic, PagerDuty, and if you stretch it far enough, AWS do a rocking business selling to developers, because (imho) they are tied to the business process of delivering a product, not the developer need to build things. The market for selling to developers is huge, because the market for things created by developers is huge and developers are expensive to hire. You can't sell just anything because of the supply problems cited above, and I think Trello's user base is the big story here that RethinkDB didn't have.