"Our existing codebase sucks. So let's buy some cool companies in the wild and let them help fixing our codebase" Yup, it sounds like the perfect plan to me.

"Our existing codebase sucks. So let's buy some cool companies in the wild and let them help fixing our codebase" Yup, it sounds like the perfect plan to me. What likely happens is this. The current codebase is too ugly to improve. But since they have a lot of users, it has value. So, the engineers from Keybase started from scratch, try to implement all of the functions in the existing codebase, plus secure. The plan is, after it has been developed, replace the existing codebase. But unfortunately, they miss the planed deadline by years and when it's finally working, they couldn't implement all of the existing codebase because nobody knows how to implement it. No documents and original implementers were left the company long ago. But they spent so much effort on the new project and all the new features are implemented just on the new one. Resulting the chimera of old and new code base both running at the same time. Oh and by that time, the user is rapidly decreasing for they failed to improve the service for years while the competitors offer the better service now. The same story repeated countless times.