For example, in a programming language like C++ or Java, one is not required to think about the consequences of their design decisions. It is possible to design an API in such a way that it causes harm to another person or program, for example not using type safety, or being vulnerable to a buffer overflow attack. A vegan programming language would ostensibly prevent such harmful actions by way of its very grammatical and logical structure. It would be next to impossible to write a valid construction in a vegan programming language that causes harm, because you would be required to inform the compiler of your intent. Any such construction would simply be illogical, and cause an error to occur.
The differences between a vegan programming language and a non-vegan programming language could even be extended to include political correctness enforcement features. A compiler for a vegan programming language could be designed to include an artificially intelligent runtime library which attempts to decide if the actions of the program could put people or animals at risk to be harmed, and if so, modify the control flow of the program to prevent possible harm from occurring. This is relevant, because computers, and computer software, are being used as components in machines for the food industry (e.g. meat processing plants) as well as militaries (e.g. missile guidance systems.)
Disclaimer: I don't eat animals.
Tell me more.