In the mathematics community it's quite common for professors to write their own notes for the classes they are teaching. The notes are then usually published in both PDF and PS form on the course website. However, it's extremely rare to see anyone publish the LaTeX source code for their notes and slides. As a student this has always bugged me while as an instructor this has to me seemed like something obvious to do, meaning I don't see how it could hurt and it can only help, so why not do it?
Here are a few reasons for publishing source code for notes:
Makes it possible for students to easily create condensed versions of the notes for future review.
They can easily edit the text and add their own comments to it if something is unclear or they can even replace a proof with one of their own, if another proof seems more natural to them. They can also edit them to remove material that's obvious/known to them, thus making the notes quicker for them to read.
If a student in the class has a habit of typing up their own notes, they will be able to copy paste e.g. statements of definitions and theorems from the class notes, which will save them considerable time.
There are flashcard programs out there like e.g. Mnemosyne and Anki that support LaTeX input. Having the source code would make it considerably faster to create cards for e.g. precise definitions and statements of theorems. Especially in fields like algebra, algebraic geometry etc. which have tons of terminology, this could be useful for quick memorization while learning material for the first time.
My questions is then: why aren't most people publishing source code? Is this something they just haven't thought about? On the other hand, as a student I asked professors for the source code quite a few times and my request often got rejected with no explanation.
As an aside I might add that I grew up as a teenager in the early 90s hacking on Linux, so open source has always seemed like the "right" idea to me for any knowledge sharing.