[Development] [Mingw-w64-public] Fwd: Choosing a new MinGW for Qt 5

[kai.koehne at nokia.com]

> -----Original Message-----
> From: ext Loaden [mailto:loaden at gmail.com]
> Sent: Tuesday, September 11, 2012 4:00 PM
> To: mingw-w64-public at lists.sourceforge.net
> Subject: Re: [Mingw-w64-public] Fwd: [Development] Choosing a new MinGW
> for Qt 5
> 
> Does anyone think about we can use cross compilation on Linux.
> e.g. on Ubuntu 12.04, the mingw-w64 4.6.3 can works well with Qt.
> It will fast, and can use make -j option.
> But it can't run the QTest (maybe? I am not sure).

There's nothing wrong with cross-compilation. But what we need first and foremost is a reliable, native MinGW environment for developing Qt applications, since the vast majority of Qt developers that develop for Windows also develop _on_ Windows. Now one can argue that Qt itself could still be compiled using a cross-compiler. But if we're not able to compile Qt ourselves with the native toolchain, how can we be sure other complex apps will work?

I haven't found a stock native MinGW 64 bit package yet that can compile Qt 5 + Qt Creator. Peter seems willing to experiment with a custom package, but I personally do not see this for 5.0 . 

I think for 5.0 , we should therefore only put MinGW 32 bit gcc 4.7 as Tier 2 or (preferably) Tier 1 platform. Support for MinGW 64 bit would be on a best effort basis (Tier 3). 

Now for 5.1 (that is, earliest May 2013) we should definitely reevaluate MinGW 64 bit as Tier 1 or Tier 2.
 
What do you all think? I certainly don't want to discourage any effort in getting MinGW-64 bit and Qt 5 to work, but rather want to be realistic in  what we should aim for with 5.0.

Kai

PS: Added development at qt-project.org again to the thread.