It happens all the time in astrophysics. Codes are competitive edges, and the support burden from people asking questions about your code that you did release is also a very real issue.