I guess what I wanted to get at was how in other disciplines, the gap between the ‘tools’ used in early exploration, and the final artifact, is so vast. Still, this exploration is ‘rich’ in its informing the designer to make decisions about the design.
Software isn’t like this. Code in itself is an expression of someone’s ideas. The same goes for a UML model or a sketch on a whiteboard. You aren’t really testing your ideas when you scribble at a whiteboard, or sketch in UML. The only test that takes place is the one against your own experience; ‘I’ve done something like this before, so perhaps it’ll work.’
Part of this is the attitude in software, where the task of creating a software system to be used in a particular context is seen as merely mapping the domain, producing documents, and handing them over to the programmers. Analysis and implementation – design’s not part of it, and hence not exploration or sketching as well.1
1 This was an email reply to Jack Nutting, who shared some thoughts on my original post.