Re: [1/2] dimm.idl: Add IActiveIMMMessagePumpOwner.
What's a reason for having it declared like that: -- + HRESULT OnTranslateMessage([in] const struct tagMSG *pMsg); --- Why not use MSG*?
On 2015-06-23 23:51, Nikolay Sivov wrote:
What's a reason for having it declared like that:
-- + HRESULT OnTranslateMessage([in] const struct tagMSG *pMsg); ---
Why not use MSG*?
Hmm I thought the compiler was complaining about not knowing MSG, but I must have imagined that, it's building fine now. I don't think it really makes a difference, but I can resend with that fixed in both files if you think it's cleaner. Thanks, Thomas
On 24.06.2015 0:58, Thomas Faber wrote:
On 2015-06-23 23:51, Nikolay Sivov wrote:
What's a reason for having it declared like that:
-- + HRESULT OnTranslateMessage([in] const struct tagMSG *pMsg); ---
Why not use MSG*?
Hmm I thought the compiler was complaining about not knowing MSG, but I must have imagined that, it's building fine now. I don't think it really makes a difference, but I can resend with that fixed in both files if you think it's cleaner.
Yes, I think there's no reason not to match SDK in that regard.
[ + object, + uuid(b5cf2cfa-8aeb-11d1-9364-0060b067b86e), + pointer_default(unique), + local +]
Why local? SDK idl doesn't have it.
+ HRESULT OnTranslateMessage([in] const struct tagMSG *pMsg); + HRESULT Pause([out] DWORD *pwdCookie); + HRESULT Resume([in] DWORD dwCookie);
Also please don't use prefixed notation, 'msg' and 'cookie' (in Pause and Resume) would be better.
Thanks, Thomas
On 2015-06-24 09:15, Nikolay Sivov wrote:
On 24.06.2015 0:58, Thomas Faber wrote:
On 2015-06-23 23:51, Nikolay Sivov wrote:
What's a reason for having it declared like that:
-- + HRESULT OnTranslateMessage([in] const struct tagMSG *pMsg); ---
Why not use MSG*?
Hmm I thought the compiler was complaining about not knowing MSG, but I must have imagined that, it's building fine now. I don't think it really makes a difference, but I can resend with that fixed in both files if you think it's cleaner.
Yes, I think there's no reason not to match SDK in that regard.
Will do.
[ + object, + uuid(b5cf2cfa-8aeb-11d1-9364-0060b067b86e), + pointer_default(unique), + local +]
Why local? SDK idl doesn't have it.
+ HRESULT OnTranslateMessage([in] const struct tagMSG *pMsg); + HRESULT Pause([out] DWORD *pwdCookie); + HRESULT Resume([in] DWORD dwCookie);
Also please don't use prefixed notation, 'msg' and 'cookie' (in Pause and Resume) would be better.
I was following what the rest of the file does with these two things. Will change it. Thanks. -Thomas
participants (2)
-
Nikolay Sivov -
Thomas Faber