I'm now reading source-codes about video capturing.
Mostly I read source-codes with editor running on PC.
But in this instance, I read these on the paper.
Looking at these methods through the lens of user interface, The paper is comfortable than some editors for doing the task.
But if source-codes to read is a source-codes which has thousands of lines, printed papers overflow and absence of search, like incremental search , becomes painful.
So I think that the method to do the task varies by the size of source-codes to read and there is a bound which divides two methods.
By the way, what method are you using for reading a source-code? :-)