150 likes | 165 Views
Explore the challenges and strategies for tackling defensive and offensive code problems, legacy code issues, and virtualizability problems in software engineering. Discover proposed goals for Secure Execution Subsets (SES) in a clear and concise manner.
E N D
When Two Languages Are Simpler Than One Lessons for SES from Cajita, Original-Caja, and Valija Mark S. Miller
Simultaneous Problems D = Defensive code problemO = Offensive code problemT = Legacy tools problemC = Legacy code problem
Simultaneous Solution? D = Defensive code problemO = Offensive code problemT = Legacy tools problemC = Legacy code problem
Don’t try this at home (or at all) D = Defensive code problemO = Offensive code problemT = Legacy tools problemC = Legacy code problem
Separate Solutions D = Defensive code problemO = Offensive code problemT = Legacy tools problemC = Legacy code problem
Layered Solutions D = Defensive code problemO = Offensive code problemT = Legacy tools problemC = Legacy code problemV = Virtualizability problem
Lessons for SES D = Defensive code problemO = Offensive code problemT = Legacy tools problemC = Legacy code problemV = Virtualizability problem
Proposed SES Goals • SES is smallest secure subset of ~Harmony-strict without loss of functionality. • SES is a good target for a multiply instantiable embedding of ~Harmony-strict.