Reading the following: http://gcc.gnu.org/onlinedocs/gcc-4.4.0/gcc/C_002b_002b-Dialect-Options.html#C_002b_002b-Dialect-Options and in particular: -ftemplate-depth-n Set the maximum instantiation depth for template classes to n. A limit on the template instantiation depth is needed to detect endless recursions during template class instantiation. ANSI/ISO C++ conforming programs must not rely on a maximum depth greater than 17. suggested to me that the following code should hit that limit when compiled with option, -ftemplate-depth-2: //-{--cons_bench.cpp-- template<class Head, class Tail> struct cons_hd_tl { typedef Head head; typedef Tail tail; }; template<typename... Elements> struct cons_fold ; template<typename Head, typename... Tail> struct cons_fold<Head,Tail...> : cons_hd_tl<Head,cons_fold<Tail...> > {}; template<> struct cons_fold<> { typedef cons_fold tail; }; template<int I> struct ti {}; typedef cons_fold<ti<1>,ti<2>,ti<3> > cf_3; //-}--cons_bench.cpp-- I tthought it should fail because cons_fold with five arguments would call cons_fold recursively 4 or 5 times, which is more than the 2 specified on the command line. However, it compiles OK. What does -ftemplate-depth-XX really mean? TIA. -regards, Larry
