Hi,

> Another argument against replacing Qt's error handling with exceptions is the 
> following:
> In situations where size matters (on embedded platforms - what did you 
> think...) you can turn off RTTI (and implicitly exception support) and 
> produce *much* smaller binaries.

That's OK, but you need to ensure that non-exception based error
handling actually exists.

A number of the template collection methods return void instead of
reporting failure - such as resize(), push_back() et al.  I'm guessing
the API was copied from STL, but those explicitly use exceptions.

Maybe Qt5 could return a bool to denote success/failure?  If one ignores
the return value, the behaviour would be the same as in Qt4.

The documentation should probably be updated to explicitly say that Qt
does not catch any exceptions thrown by new.  Actually, the current
class docs don't seem to say anything about handling failure.

Steve
_______________________________________________
Qt5-feedback mailing list
[email protected]
http://lists.qt.nokia.com/mailman/listinfo/qt5-feedback

Reply via email to