QTransform vs QMatrix

[wysota]

The reason for having two classes is simple. QMatrix was introduced in Qt 3.0 as QWorldMatrix and is a direct port to Qt4 and QTransform was introduced in Qt 4.3 (I think) and simply extends QMatrix with new possibilities. AFAIK since 4.3 QMatrix is (or at least could be if not for complexity reasons) implemented using QTransform. As you can't remove a class from a newer version because of compatibility reasons, another class had to be added. That's a normal approach in Qt. For instance we have a few classes with "V2" (and even "V3") in their names that inherit their previous versions and extend them with new capabilities.

[Michiel]

Yes, I thought that would be the reason. I don't like backward compatibility. ;-)

[maverick_pol]

Ok,

SO why not to extend the existing class, or add some other functionality, but to create a new class?

Ok..Just being curious.

Thanks

[wysota]

These are different matrices. For instance the detatchment (det()) returns different values for both matrices, therefore you can't just extend QMatrix to become QTransform. Already written code would stop working.