the reason i fell into this trap in the first place because i wanted to understand what "C formalized in HOL" was on about https://www.cl.cam.ac.uk/techreports/UCAM-CL-TR-453.pdf
C also combines a number of interesting features on the theoretical front, making it additionally interesting as a subject of study.
this is not something i wanna hear from someone claiming to have formalized it
For example, C’s expressions both are side-effecting and have very under-specified evaluation orders. If these semantic features were the main area of interest in studying a language, then it would clearly be easier to construct a simple calculus that included these features and little else.
this is ridiculous. obviously these semantic features are not ideal when attempting to write code that runs e.g. in ring 0. yet people do it (and this is meaningfully outside the C standard). the UB becomes defined thanks to our friends who write the compiler. is it worth attempting to standardize ring 0 properties?
However, we prefer to attack as much of C as possible all at once. As Milner and Tofte point out in the commentary on the definition of SML [MT90], this study of languages in their entirety has its own grounds for interest, and we further feel that our study of C gives us a possible
application in the area of software verification.
they didn't even mention a single concrete implementation until the appendices