I realize my reaction is, in a sense, completely off-topic, but what I notice most is that they didn’t think Docker solves this problem for them. In fact, I bel

I realize my reaction is, in a sense, completely off-topic, but what I notice most is that they didn’t think Docker solves this problem for them. In fact, I believe this is the first time in 4 years where I’ve seen a company try to solve this problem without making Docker the center of their effort (though they bundle the nightly build in Docker, it seems a minor part of their strategy, and could have easily been a traditional VM instead). They are being creative. The crucial thing is they didn’t feel this problem was solved by Docker, so they did something original and, in some sense, this amounts to a vote-of-no-confidence on the question “Does Docker solve our development problems?” A question that, back in 2016, most people thought Docker would solve. It is possible that 5 years from now, when we look back, we will remember this as the moment when the momentum and focus and energy of the tech industry began to shift in new directions. Because, certainly, for the last 5 years most of the tech industry has been asking “How will Docker solve all the problems of development environments.”