While building up an engine as a hobby while following along with your Shader VMKs, I've been utilizing wstring through the code to support what I thought was better Unicode cross-compatibility. Turns out the more I read the more I believe I need to utilize std::string throughout where it should be assumed content generated and consumed is UTF-8, the preferred Unicode encoding at this time.
UTF-8 w/o BOM seems to be the most stable and sensible step forward, but it's actual implementation in C++ and cross-compatibility with Windows, Mac, Linux etc. is proving a challenge for me to get a good grasp on and implement.
Do you know if your current engine implementation as written would allow UTF-8 international characters for filenames, etc? My heavily modified version does not properly accept these characters and build across the 3 platforms currently. I think the best step would be clone my current source so far and go through the tough process of converting everything to std::string and only widen as necessary to interact with the specific pre-built calls, but it's going to take awhile.
I've been studying a few links referenced here:http://utf8everywhere.org/http://utfcpp.sourceforge.net/http://www.nubaria.com/en/blog/?p=289
Before I go through this re-write, I know your input and thoughts on the topic would be greatly appreciated whether through a new VMK or forum discussion if your time permits.
Thanks in advance,