Ed Haywood
Final Approach
Reminds me of my first experience with Ada. I was building a predictive parser and so had a very complex data structure to be initialized at compile time. (School project, not how I'd do it in real life.) I bought the compiler to save time as I was working full time and could afford it. It was considered one of the best PC Ada compilers in the early 90s but I don't remember the vendor. Dang thing would fall all the way through 30+ full page (on the screen) elements in the data structure initialization, hit the final closing curly brace and say the equivalent of "I dunno. Broke." I culled the data structure down to two elements and spent several hours debugging the syntax until that worked. Then built up all the elements until it was complete. I will say, once it compiled, there was one small logic error that I caught and corrected and it ran flawlessly. But after years spent in C where if the compiler could somehow make some sense it would compile (maybe not what you wanted, but that's a you problem, programmer) It was incredibly frustrating. Later compilers were, I'm sure, much better.
That is consistent with the design philosophy of Ada. It's a feature, not a bug.
Last edited: