> In software we work with things that are just so _abstract_, it's nice to just be able to understand a system and it's flaws at a glance. It's nice to be able to see/hear/smell/touch what is wrong, what is right - in software, the most common way we get data is through text. Sometimes we even see a 2D image that (usually badly) represents some complex concept.
I think programmers are very good at building intuition about systems without help, partly because we rarely have good help from documentation. When we want to share our understanding of a system, we go to a whiteboard and have a very rich conversation about a simple diagram. Sometimes we draw different diagrams for different contexts: the performance context, the business logic context, the security context, etc. Fully detailed diagrams are overwhelming. Once we have good understanding, we intuitively edit the big picture down to the relevant parts. I don't know how to reproduce this in static documentation.
I think programmers are very good at building intuition about systems without help, partly because we rarely have good help from documentation. When we want to share our understanding of a system, we go to a whiteboard and have a very rich conversation about a simple diagram. Sometimes we draw different diagrams for different contexts: the performance context, the business logic context, the security context, etc. Fully detailed diagrams are overwhelming. Once we have good understanding, we intuitively edit the big picture down to the relevant parts. I don't know how to reproduce this in static documentation.