Another Doom Engine Question

In what order where the various parts of the Doom engine written? You've got am_map, am_data, Doomdef, Doomdata...all kinds of stuff. Another thing: How did Carmack and Co. figure out what part of the engine to write first?

You'd have to ask the id folks themselves. however, I do know that the doom 3 has it's roots in the quake 3 engine.

all of their games tend to start with a new shell and previous components... then they re-write as they progress.
I bet that even doom->quake had a lot of 'borrowing'.

Heh, yeah, like in the AI department :)
Quake monsters are as stupid as Doom monsters.

According to Aardappel: If you're making a game engine, you start by figuring out how you want to represent data internally (overall structure, level format, etc). Then you write the code to process that data. After that, you write the code to render the stuff on-screen. Additional parts of the engine are added in turn.

Generalized, you start with the core of an engine and work your way out to the perimeter.

The core was coded by Carmack we can assume, while Romero and Taylor probably did some of the more peripheral things. We know Romero did some complementary apps, for instance (like setup.exe) and he also coded things like lifts and crushers and other such details. Not sure what Taylor may have worked on, though.