hi all, if we disable the following two lines in DEFDLG_SetFocus function in the defdlg.c (line nos 70 and 71) if (SendMessageW( hwndCtrl, WM_GETDLGCODE, 0, 0 ) & DLGC_HASSETSEL) SendMessageW( hwndCtrl, EM_SETSEL, -1, 0 ); The testcase program given is running correctly, i have asked julliard about this. He said we need extensive testing what is the exact behavior on windows. Does this cause any regressions? Well i dont know whther this is a correct patch ?????? Well before this somebody took a stab at this but having a global variables, this was not accepted as it was a bad idea. Well if there is any other mechanism to know whether the control is a edit box, well we can put some fix so that we save the focus details on the control itself and retrive it form the control. Ideas are welcome Thanks, Vijay
Hi, with what patch? Do not use patch given in bugzilla. Thanks, Vijay On 10/8/05, Richard Cohen <richard(a)daijobu.co.uk> wrote:
Hi Vijay
You wrote:
The testcase program given is running correctly,
The testcase doesn't run correctly here with the patch -- you can no longer TAB out of the edit control.
Richard.
On 10/8/05, Vijay Kiran Kamuju <infyquest(a)gmail.com> wrote:
He said we need extensive testing what is the exact behavior on windows. Does this cause any regressions? Well i dont know whther this is a correct patch ?????? Well before this somebody took a stab at this but having a global variables, this was not accepted as it was a bad idea. Well if there is any other mechanism to know whether the control is a edit box, well we can put some fix so that we save the focus details on the control itself and retrive it form the control. Ideas are welcome
The best thing to do would be to follow Alexandre's suggestion and test the messages that occurr in windows. To see some examples of checking message sequences, take a look at dlls/user/tests/msg.c. -- James Hawkins
participants (3)
-
James Hawkins -
Richard Cohen -
Vijay Kiran Kamuju