Actually, it is a problem if you're not aware of this SR limitation, the issue is not about the string representation.
It makes it impossible to override with SRE method emitted with csc which contains a parameter whose type is a MD array. Even though the types are semantically equivalent, the CLR won't recognize the override. On Friday, January 28, 2011, Johannes Rudolph <[email protected]> wrote: > It is not a "problem" in that it will generate semantically equivalent IL > too. Be aware that Cecil in general does not stick to the conventions > employed by System.Reflection but is more aligned with ILAsm.When importing a > multidimensional array type into Cecil, you need to manually set the array > dimensions to make them show up (if the string representation is what you > care about). > > Having said that, it is common agreement that the use of array types with > lowerbounds > 0 should be discouraged. > Regards,Johannes > > > On Wed, Jan 26, 2011 at 11:53 AM, impworks <[email protected]> wrote: > > I see. So the problem exists in System.Reflection.Emit too and it's > Microsoft's fault. > Sorry for causing a distress. > > -- > -- > mono-cecil > > > > > -- > -- > mono-cecil -- -- mono-cecil
