4 ms·
I like how it emphasized the "debuggable" aspects of FMs. Most people believe FMs are not resilient due to their coupling with the spec. What if the spec chang
by bordercases 7y ago
I like how it emphasized the "debuggable" aspects of FMs.
Most people believe FMs are not resilient due to their coupling with the spec. What if the spec changes?
Saying that FMs are a debuggable form of specification, i.e. very close to debugging the design itself, turns that criticism on its head and shows that it is trivially the case that you can iterate on or prototype with FMs.
- galaxyLogic 7y agoSo what is a "FM"? Isn't it just any very high-level, preferably declarative programming language?
- bordercases 7y agoI asked for a comparison between TLA+ and Scala/Haskell to a dev that worked on the AWS application of TLA, this is what they responded. https://news.ycombinator.com/item?id=22100536 https://news.ycombinator.com/item?id=22100536 You're right in a limited sense, except that it's not any language, it's particular languages that can count as a Formal Method in the sense of the OP (i.e. can do model checking in some kind of termporal logic).