Safety-critical software in the air domain typically conforms to RTCA DO-178C. However, latent failures might arise based on assumptions underpinning the programming language used to write the software, whereas the lack of empirical data may constrain the selection of a promising but untested language. To overcome this difficulty, we propose evaluation criteria drawn from RTCA DO-178C, to help quickly review the potential applicability of programming languages in the air domain. We illustrate the constraints by using them to evaluate the suitability of the Rust programming language.