Wow, Docker in 100 lines! It runs as root? Oh. It is written in bash? Oh. It needs a ton of manual setup? Oh. It doesn't actually implement the package format which is most of the point of Docker? Oh. So is it that easy to reimplement Docker? Despite the obvious snarky intent, it appears not.
No, that's a dumb joke. Actually I don't really even use Docker personally.
HN's standard middlebrow dismissal of Docker is to claim that it's nothing except LXC, which really misses the point entirely. But at least it lets people pretend they are smarter, which is what really counts.
They are rightfully dismissive of Docker because it's just the current cycle of trendy abstractions. It's a barely passable solution to a bigger problem.
"We're running many services on a single machine. But this is complicated and difficult to update and maintain."
"We took our machine, ran a virtualization platform on it, and split each service into its own VM. But this comes at the cost of increased resource usage."
"Instead of separate VMs we created a container format to decrease the overhead while retaining many of the benefits of virtualization. But this is still resource heavy, and insecure as the containers will rarely see updates."
"So we created 'lightweight' containers which are very thin wrapper around the base OS so that containers can take advantage of updated shared libraries to mitigate the security problems, and further decrease the overhead."
"We're running many services on a single machine..."
The cycle will eventually come around and we'll be in a better place having learned that was was really needed was improvements to the base OS, package management, more robust MAC policy, and name-spacing rather than containers.
The problem is that the buzz will often entirely miss the point also. If the reader understands the underlying CS concept it is better to focus on the benefits of the product as a particular implementation of a concept, rather than attribute all the benefits to the actual product itself.
