Jean-Marc Lasgouttes wrote: >>>>>> "Angus" == Angus Leeming <[EMAIL PROTECTED]> >>>>>> writes: > > Angus> As the code stands there's going to be a compilation warning > Angus> about the conversion on 64-bit machines from size_t (8 bytes) > Angus> to int (4 bytes). > > Angus> It looks to me like all these Tabular::setXYZ functions > should Angus> really be taking a size_t argument. The alternative, > of casting Angus> the size_t to an int to explicitly avoid the > warning is ugly Angus> and conceptually wrong... > > Seeing how the tabular is defined as > > typedef std::vector<cellstruct> cell_vector; > /// > typedef std::vector<cell_vector> cell_vvector; > > I guess you are right. Does this mean that all references to 'int' > in the tabular code should be changed to idx_type?
Looks like. All these lazy 32-bit coders... -- Angus
