On Tue, 12 Dec 1995, d. hall wrote: > > ansi standard is written in stone, much less in each compiler. sun > compilers are notorious for not conforming. true. not only the compiler, also the library. :( > cause if you ever had to deal with OSF/1 from digital, you'd realize just > how ornery it's compiler can be about typecasting. these "warnings" can > turn a simple oversight into a fatal error. > > i've had a (const char *) implicitedly cast into an (unsigned char *) > become fatal upon compilation with a digital unix compiler. what exactly do you mean? I sometimes work on OSF/1 too, but didn't had any major problems with ansi there. You mean the compiler didn't warn you on such a cast? I've tried with native cc compiler on OSF/1 on an alpha, and it reported that error. Which compiler do you use on OSF1 that doesn't report this? I know, the run time behavior of those dec machines, is a bit too forgiving for my tasteB, when developing. You'll catch pointer errors maybe only after porting to other machines. But so far the warning and error messages seemed to be ok. Herbert __ [on public request 12 lines of signature deleted] *snip* ;)
This archive was generated by hypermail 2b30 : 12/07/00 PST