Thanks for all opinions. I will apply all changes in coming days to the RFC content.
> Is there a practical, real-world issue this solves? Portability of the autoloading. I had a case in the past where my wrongly cased namespace didn't cause any problem locally, but it failed in the production because of different OS. This is real world-issue. However, this problem touches only namespaces and class names. If the RFC would be only for namespaces and classes would you be in favor of this RFC? It appeared in previous discussions to put all or nothing into such RFC, so I went along the lines, but I'm okay with dropping methods and functions. However, if RFC function autoloading would pass, we would have the same problem with functions. Kind regards, Jorg
