>Developing the public mental model of containers is integral to Docker’s success and they’re rightly terrified of damaging it.<p>I guess you can all call me a moron or close to it, to this day, I don't know what a container is good for.<p>I'm reading this now: <a href="https://www.docker.com/whatisdocker" rel="nofollow">https://www.docker.com/whatisdocker</a><p>Is this server level, user level, both? Something else? I saw a Hacker Con video and someone was basically containerizing all their apps so that the underlying OS basically did nothing but run the hardware and containers.<p>Does this mean, one day, instead of monitoring my installs on Mac OS X to see what files were put where so when something breaks I can find those files? I could simply install to a container, my OS would not be touched, and I could delete the container and be back to a stable state?<p>Can you even get OS X to run on a container? Would it be a good idea or even feasible to install PhotoShop in a container, or not even possible, as that app tosses stuff all over my OS.<p>Or, is this more like a different way to do things, but it is still AWS and their AMI's or Digital Ocean or any of the others.<p>I feel I am completely falling behind and have no idea why I would need one of these. Hell, if I made an iOS app, I have no idea if there are Apple servers you pay them to use for stuff, or if you deploy your own servers, or you use something like AWS, and how does that scale on demand, do you have to build that into your s=infrastructure, or is there a "auto scale as demand dictates" checkbox?<p>Is AWS, Docker, all the rest, in the end, is this just like in the old days where I would have 42U of rack space, put in a DNS server, put in a few http servers, use this as round robin image load balancers, DB servers, backup DB servers, replication DB servers. And when I needed to grow for heavy demand for a day, that was an issue. I fail to see how you can scale based on demand when a database is involved.<p>Databases scare me, one thing never talked about... We have git for code, what about databases. How does a dev team work that out? If you need a new field, drop some data, alter a table, add an index, etc. How do you get what you have on your local machine in test out to live? How is every little database change tracked and rolled back if need be. How do the DBA guys communicate with the coders to make sure a name change to a table gets updated in the code. Is there git for postgres and others?<p>All this made sense to me 5 yeas ago when the cloud was called the internet and email, ftp, http, etc were all part of the "cloud" or as I call it "the internet". But now, things are confusing, wrapped up in terms like "cloud" that make no sense to me. I have been, as we all have been, using "the cloud" for over a decade, the first time I logged into a slip account and got on some gopher server or similar, that was cloud to me and I believe it still is. POP email is cloud, the internet is cloud. It is now just convoluted to the developer so end users understand something that really only developers need to understand. Am I making the same mistake with containers, and they are nothing special and have been around for ages like the cloud and it is just a buzzword now? Apple has sandboxing and something called containers in their OS, is this similar in principle?<p>Auto makers don't burden their end users with engine shop talk, nor dumb it down for them into simple terms, but we seem to with tech. This should be it's own post but I just kinda got on a roll, sorry.