Legacy software systems were often written not just in programming languages typically associated with legacy, such as COBOL, JOVIAL and PL/I, but also in decommissioned 4GLs. Writing compilers and other migration and renovation tools for such languages is an active business that requires substantial e ort but has proven to be a successful strategy for many cases. However, the process of covering such languages is lled with unconventional requirements and limitations: the lack of useful documentation, large scale of codebases, counter-intuitive language engineering principles, buggy reference implementations, etc.

In this paper, we motivate the incremental nature of software language engineering when it concerns legacy languages in particular, and outline a few related challenges.