Pricing virtual assets is hard
//
imo the thing to look out for is actual vendor / architecture lock-in. So just don't make things complicated yourself... Docker has it's uses, but right now it's being used for literally everything. I've been using jails, vms, and containers since the 90s. Docker is niceish, but somehow got lots of attention.
Oh, and you need to be sure your platform won't eat all your revenue.. like oracle :)
Especially on servers, and especially if for some stupid reason licensed software need internet access to confirm it is licensed, coz that's extra crap that needs to be added on proxy or firewalls and will inevitably break when they change something.
If it's "hey, pay us X and we trust you don't cheat us" yeah, it's just extra invoice ,whatever.
(I was already working on those instructions when the pricing change reached our attention, so it became a viable free alternative only by coincidence, but still.)
One other thing I learned: HN crowd is biased towards people who want to learn and understand their tools to a degree. Workplace use of Docker has large component of people who don't know what Docker is for and don't really care, but they need it for some task, so they want a hassle-minimizing option they can use to get their job done, and then forget about.
I bet it's a huge market, not just for Docker, but any other software tool. One that's somewhat alien to the HN audience.
At hindsight, I would say, it costs several man-months of senior engineers' with good connections in various departments.
And this comes on top of productivity losses prior to the solution.
The whole of Docker Desktop isn't replaceable with a drop-in alternative today but parts of it are, and the rest could be very soon if an org with the resources and ability keeps developing things like podman.
So what would Docker as a company do when the equivalent of what happens to Swarm happens to Desktop? Pivot again?
Not meant negatively towards Podman, maintaining such a package in a constantly moving ecosystem is hard. The point is more that you pay Docker to do the extensive testing for you, so that the work-stopping issues are ironed out before they roll out to customers.
And yet there are horror stories of finance not approving purchases that would significantly improve productivity.
While obviously greater than the $5/9/24 you bring up, I've heard of companies not wanting to buy a second monitor for their engineers, whereas I'd wager the productivity gain would pay for itself in only a couple months.
My company received the memo from Docker to upgrade to paid version, and I helped many to move on to Podman. However there were teams for which getting a Docker Desktop license was easier than the Podman move. I am sure there are many more companies that can easily afford the convenience of Docker Desktop for a trivial fee.