I don't understand this phenomenon; even I want to make one for my application domain, but at the same time, all of this seems intuitively wrong. Perhaps some of you have some insight, thanks!
Because writing a NoSQL database is easy. MongoDB is just a thin wrapper around mmap() that makes it look like JSON for example. Whereas writing a new and better SQL query optimizer is hard, and testing to ensure no data loss is tedious, the MongoDB guys just skipped it.
I guess it's because people don't buy the "one size fits all" approach to databases anymore...
15 years ago, you had to choose between Oracle, db2 or Sybase... The SQL offer grew significantly too since then.<p>You're probably right: nosql DB developers spread their efforts too thin. I'm currently only interested in new DBs if they solve a new problem (statistics and behavior DBs look promising).