Hi, thanks for your comment. The article is mainly focused on that specific issue on porpuse, to keep it simple and straight to the point.
Yes, is a multi-step process as it is described (which is not ideal) but currently the tools that we have (I’m taking in consideration we want to work only with AWS cli) are not quite there to make the whole process a single step (or I didn’t found one).
That basically put more work on our side to take care of the state of the cloud and handle any failures that can occur in the middle. (You can use Stack´s to help you out with that in AWS). But again, this article was meant to tackle that specific issue.
There are many things I choose to omit on this article (alternative solutions for example). Every CI/CD environment differs depending on the project needs, so I think each team should decide how to solve this issue (if they face it) taking in consideration the pros / cons of each alternative. This article shows one (another alternative could be to introduce another tool that takes care of the whole thing like Serverless, but as you can see, the whole point of discussion shift dramatically to another area).
