RethinkDB is still very young software and will take some time to mature. The worst issues that we had during the last release cycle did result in data corruption, one due to a bug in our code. Another due to a filesystem we weren't expecting (which we still consider to be a bug in our code.)
Right now we can't recommend RethinkDB for production ready code but we're working very hard to get it to that point. Part of this is beefing up our internal testing system which we've made a big push on during the last release cycle. Part of that is fixing all of the bugs that users submit until they stop finding serious ones.
What jdoliner said. A few were cosmetic, but I think I experienced some data corruption as well (it was a test dataset). The team was extremely responsive and fixed the bugs in a few hours, and RethinkDB is clearly marked on their website as still an alpha, so I can't blame them.
I'm sure it will be a fantastic product when it matures, and I can't wait for a stable release so I can try it out.
Oh, that's very true. We used an internal versioning scheme and it became exceedingly difficult to change all the repo tags, internal references, etc. so we decided to just bite the bullet and use the same scheme publicly and privately. I wouldn't have done this if we could go back in time, but you know what they say about hindsight :)
Right now we can't recommend RethinkDB for production ready code but we're working very hard to get it to that point. Part of this is beefing up our internal testing system which we've made a big push on during the last release cycle. Part of that is fixing all of the bugs that users submit until they stop finding serious ones.