I don't think we're facing new problems with cloud computing and large systems. We're facing the same problems we've been faced with again and again since the v

I don't think we're facing new problems with cloud computing and large systems. We're facing the same problems we've been faced with again and again since the very beginning of computer science. Over time, the ease and scale of what you can do with computer resources increases and we have to organize and understand the existing resources we have. But we keep reinventing the same solutions at each scale. At one time we had to invent functions to enforce segregation of responsibilities and create abstractions and shorthand. We had to group these together in modules and libraries. We had this clump of programs running on a computer that we had to organize into an operating system. Now an operating system is nearly a program or function and people are regurgitating the Unix philosophy and the end-to-end principle like it's a new thing. In the end, we're going to wind up with a well-architechted series of integrated microservices which present a comprehensible interface to users through a handful of abstractions that have proven useful over the years. Computing is cheap enough that we can now talk about meta-computing, a higher level of abstraction from a computer, which is multiple layers of abstraction on top of eachother. Now we just have to build the next layer. And I think it will basically be a sort of meta-operating system. The same things, but we'll call it "orchestration" and "microservices" instead of a file/process/whatever manager and threads. At the moment, however, we're still offering piecemeal services and products and so we don't have many fully formed concepts of what it is to build a cloud system. So things are still a bit chaotic, but at some point in near future we'll get there.