Recently, I struggled with an algorithm that was poorly implemented (i.e. the developer was removed from another project and was unable to adequately document what he did) in C #.
I found an alternative (from numerical recipes) that works, but is written in C ++. So I think that probably the safest way to get something would be to wrap C ++ in a DLL.
Whereas, I'm still a little green when it comes to C #, and have never tried making a DLL from scratch, this sounds like a reasonable approach (and if so, has anyone tried this / got any advice) ? Or should I go all pigs and try to port the C ++ routine to C #?
Change . I am not looking for someone to make a decision for me, but if someone has any experience with any of the routes, I would be interested to hear their opinions and any unpleasant pitfalls that they should avoid. For example, how annoying is it to pass a C ++ STL vector in lists of data from C #?
source
share