Re: [PATCH 2/2] urlmon: If ProtocolCF_CreateInstance doesn't support aggregation retry without it.

[Dmitry Timoshkov]

Jacek Caban  wrote:
> Those changes look mostly good, but we should probably not forward QI 
> calls in BindProtocol_QueryInterface if the protocol doesn't support 
> aggregation.
>
> Also I don't mind tests from patch 1, but it would be probably easier to 
> extend test_CreateBinding(). It would also make a testing the above 
> comment trivial.

Thanks for the review. Perhaps the patches could be accepted in current
state, and I'll have a look at improving the tests according to your
comments?