Was reading one of the feature articles over at Gamasutra Custom Tools: Environment Artists and Game Editors. Was a good read; from my experience I would have to agree with a lot of the article. But I think the author just begins to scratch the surface.
One of the big reasons a level editor takes such a long time to create is the number of things it has to be to so many people. It’s the one of a very few tools you will likely find an artist, designer and programmer all using. Which complicates matters because all 3 groups each look to the tool to solve different problems. Programmers will likely want it to be responsible for baking as much data as it can for the runtime, since it’s generally the last tool in the pipeline for a game, making it naturally well suited for the task.
A designer may look to the tool for it’s data management capabilities; They may need to setup a very complicated quest chain with all sorts of interactions between NPCs, how does the tool provide them with visualizations of interactions and the ability to manage and search a vast database of generated content to find out what they need to know about the world/level.
Artists should be obvious, but in addition to the aesthetic side of things one of the biggest pitfalls you can fall into is their desire to be able to move seamlessly between their DCC tool of choice and your level editor. Which means giving them the ability to make as much of the tool share the same workflow shortcuts they are used to. ex, key binding shortcuts, 3d manipulation gizmos, camera controls…etc.
Tools development is not for the faint of heart :)