3 ms·
I think you'll have a hard time persuading existing corporations and developers that this is worth doing; formal verification seems to be currently constrained
by cbhl 13y ago
I think you'll have a hard time persuading existing corporations and developers that this is worth doing; formal verification seems to be currently constrained to academia and safety-critical uses (say, Software Engineers that have been accredited by a local regulator and are working on shutdown routines for a nuclear power plant).
But perhaps if you built a formally proven shim around key pieces of software, you might be able to get someone else to adopt them and/or disrupt (usurp) an industry incumbent.
- lallysingh 13y agoI think that's a matter of tool quality. If they're high quality tools/libraries/documentation, then it could just become another secondary skillset like writing good tests and using version control. For motivation, talk to the folks worried about costs of substantial failures. This is really analogous to insurance. An unpleasant, but tolerable amount of overhead that gives you a maximal bound on how bad a failure can be.