Deliberate practice is important in many areas of learning, including that of learning to program computers. However, beliefs about the nature of personal traits, known as "mindsets," can have a profound impact on such practice. Previous research has shown that those with a "fixed mindset" believe their traits cannot change; they tend to reduce their level of practice when they encounter difficulty. In contrast, those with the "growth mindset" believe their traits are flexible; they tend to maintain regular practice despite the level of difficulty. However, focusing on mindset as a single construct focused on intelligence may not be appropriate in the field of computer programming. Exploring this notion, a self-belief survey was distributed to undergraduate software engineering students. It revealed that beliefs about "intelligence" and "programming aptitude" formed two distinct constructs. Furthermore, the mindset for programming aptitude had greater utility in predicting software development practice, and a follow-up survey showed that it became more fixed throughout instruction. Thus, educators should consider the role of programming-specific beliefs in the design and evaluation of introductory courses in software engineering. In particular, they need to situate and contextualize the growth messages that motivate students who experience early setbacks.