5 ms·
I’m someone who had a front row seat to the emergence of Docker, and some might say competed with them (I’d disagree on that point). I don’t plan on commenting
by mitchellh 4y ago
I’m someone who had a front row seat to the emergence of Docker, and some might say competed with them (I’d disagree on that point). I don’t plan on commenting on their company, business model, or recent decisions. The only thing I want to comment on is the claim Docker was evolutionary, not revolutionary.
I disagree, I believe Docker /was/ revolutionary. And I feel like I see heavy technologists make this sort of dismissal based on technical points too soon. From a technical perspective, it was arguably evolutionary — a lot of people were poking at LXC and containerization a long time before Docker came around — but from a product perspective it was surely revolutionary.
I used to joke, in my own experience building a business in the DevOps space, that you’d spend 2 years building a globally distributed highly scalable complex piece of software, and no one would pay for it. Then you slap a GUI on it, and suddenly someone is willing to pay a million dollars for it. Now, that’s mostly tongue in cheek, but there is a kernel of truth to it.
The kernel of truth is that the technology itself isn’t valuable; it’s the /humanization/ of a technology, how it interfaces with the people who use it every day.
So what Docker did that was revolutionary was take a bunch of disparate pieces, glue them together, and put an incredible user experience on top of it so that that technology was now instantly available in minutes to just about anyone who cared.
At some point in the article, the author says it’s maybe something about a “workflow.” I’m… highly biased to say yes, absolutely. One of my core philosophies (that became the 1st point of the Tao of the company I helped start) is “workflows, not technologies.” When I talk about it, I mean it in a slightly different way, but it’s highly related: the workflow is super valuable for adoption, the technology is to a certain extent, but less so.
Technology enthusiasts (hey, I’m one of you!) usually hate to hear this. We all want to think you build the best thing or a revolutionary thing and then it just wins. That’s sometimes, but rarely, the case. You need that aspect, and you ALSO need timing to be right, the interface to be right, the explanation to be right, etc. Docker got this all right.
(Now, turning the above success into a business is a whole different can of worms, and like I said in the first paragraph, I don’t plan on commenting.)
For the author: I don’t mean any offense by this. I mostly agree with the other points of your post. The “FROM” being revolutionary I was nodding quite vigorously. Being able to “docker run ubuntu” was super magical, etc. I mostly wanted to point this because I see MANY technologists dismiss the excitement of technologies purely on the basis of technology over, and over, and over again, and the sad thing is its just one part of a much bigger package.
- jcrawfordor 4y ago(author) I'm not sure that we really disagree, but I wrote this sort of late and I also think I wasn't entirely clear. The point I was trying to make is that the "container runtime" part of Docker is a lot less important than the tooling they put around it, and they made Docker Hub a very core part of that broader ecosystem.
- femto113 4y agoI think that in many ways creating a shared, stable namespace for images was actually a bigger contribution than any of the technology. The ability to type somtething like 'FROM python:3' at the top of your Dockerfile and have that automagically mean what you expect was definitely revolutionary in terms of productivity. Behind the scenes I don't know it really matters that much whether that references an image hosted in a repository by Docker the company, or a file in AWS S3, or a tarball from the Python Software Foundation. And that namespace is exactly what they're stabbing in the heart.
- crabmusket 4y agoNamespaces are so important to ecosystems. See the issues with NPM packages, discussions about Cargo organisation names, etc. I was a huge fan of Deno early on because it used "the web" as its package namespace. Every time I see a new tool come out which bakes an assumed default into a "bare name" I die a little.
- Dalewyn 4y ago>The kernel of truth is that the technology itself isn’t valuable; it’s the /humanization/ of a technology, how it interfaces with the people who use it every day. Apple in a nutshell.
- EntrePrescott 4y agoApple even takes it even one step further: they understood (Just like haute couture, cosmetics and luxury watch brands before them) that it's neither the product nor the technology nor even the interface itself that's what's really valuable (in the sense of: monetizeable), but the user EXPERIENCE in the sense of how it makes the user FEEL. Which is Why Apple excels in brand marketing. If you want a cash cow, you don't want a technology-focused project or a mere company, but something between a luxury lifestyle brand and a cult.