From: Stefan S. <se...@sy...> - 2004-09-17 23:50:32
|
Grzegorz Jakacki wrote: >>aren't symbols stored by their encoded name, which contains the signature ? > > > If not overloading, then maybe the fact, that OpenC++ cuts corners on type > analysis, which results in encodings of different functions look the same? > E.g. > > void g( char (*)[5] ) {} > > and > > void g( char (*)[6] ) {} > > ? you are right ! That's the same conceptual problem as disambiguating the relational '<' operator from template brackets ! In both cases we lack a symbol lookup table to solve the issue. Let's see what we can do... Regards, Stefan |