knudsen (12/29/82)
I agree with Darwyn: a consistent interpretation of C semantics demands that foo(x++) get the updated value, whatever the actual truth is for real implementations. Yes, ++ comes from the PDP-11 and now the 6809 (and I guess the 68K?), and the irony is, if a compiler forces the example to deliver the old value, then the post-increment power of the '11 is wasted. I also agree that LINT ought to kvetch about such usages. Maybe I'm glad that dumb languages like BASIC don't let you get into trouble with nested assignment stmts/exprs, etc. Not afraid to write in C, but should I B? --mike k
preece (01/05/83)
#R:ihnss:-117800:uicsl:6400001:000:241 uicsl!preece Jan 4 15:02:00 1983 Actually, real Dartmouth Basic (as of 1970, when i last used it) did in fact allow nested assignments. It also had multi-line DEFs, local variables, and the assumption that all DEFs are allowed recursion. And, of course, dynamic strings...