- From: Xavier Franc <xfranc@online.fr>
- Date: Thu, 18 Sep 2003 00:37:51 +0200
- To: public-qt-comments@w3.org
certainly overloading based on the type signature would be best. [And in addition with a guaranty that always the most precise signature is chosen, like for example in Java.] It is not so complex to implement per se, but it requires complete type inference. So overloading based on the type signature is closely related with static type checking. It could perhaps be part of the static typing feature? Michael Rys wrote: > Because it adds complexity to the language and to the implementations > and it may contradict a future overloading based on the type signature > that is preferred by many members but which was postponed to Vnext. > > We designed the built-in overloaded functions that either approach > should work, but allowing users to overload would lock us in. >
Received on Wednesday, 17 September 2003 18:36:59 UTC