Thread overview
DnD: what is wrong?
Jan 31, 2004
Dmitry Morozhnikov
Feb 01, 2004
chris elliott
Feb 02, 2004
Dmitry Morozhnikov
Feb 01, 2004
Arjan Knepper
Feb 02, 2004
Dmitry Morozhnikov
Feb 02, 2004
chris elliott
January 31, 2004
Hello all again.

What is wrong with DnD in wxWindows? Does anyone have working samples/dnd/ programm (not only compiled, but run and actually dnd without strange debug messages in log window)?

If not -- is it dmc issue or maybe it is problem with my win98?

dmc 8.38
wxWindows 2.4.2
win98se

Big thanks for any help.


February 01, 2004
Dmitry Morozhnikov wrote:

> Hello all again.
> 
> What is wrong with DnD in wxWindows? Does anyone have working samples/dnd/
> programm (not only compiled, but run and actually dnd without strange debug
> messages in log window)?
> 
> If not -- is it dmc issue or maybe it is problem with my win98?
> 
> dmc 8.38
> wxWindows 2.4.2
> win98se
> 
> Big thanks for any help.
> 
> 
Thanks for reporting this ; I have not tried the DnD sample for some months and will look at it again. I know there was an issue with the OLE clipboard and DnD but thought it was fixed (at least in CVS).

I'll try and check it in the next 24 hours
chris

February 01, 2004
I don't know. What "messages" do you see?

I just tried the DnD sample form wxw-2.4.2 (with some small modifications) with 8.38 I don't see any problem.

Arjan.


Dmitry Morozhnikov wrote:
> Hello all again.
> 
> What is wrong with DnD in wxWindows? Does anyone have working samples/dnd/
> programm (not only compiled, but run and actually dnd without strange debug
> messages in log window)?
> 
> If not -- is it dmc issue or maybe it is problem with my win98?
> 
> dmc 8.38
> wxWindows 2.4.2
> win98se
> 
> Big thanks for any help.
> 
> 
February 02, 2004
In article <bvjn4t$ojf$1@digitaldaemon.com>, Arjan Knepper says...
>
>I don't know. What "messages" do you see?
>
>I just tried the DnD sample form wxw-2.4.2 (with some small modifications) with 8.38 I don't see any problem.

It look like with a library compiled with __WXDEBUG__:

08:17:35: Debug: In file ..\..\src\msw\ole\droptgt.cpp at line 332:
'CoLockObjectExternal' failed with error 0x800401f0 (there is russian message
mean CoInitialize was not called).
08:18:02: Debug: In file ..\..\src\msw\ole\dropsrc.cpp at line 222: 'DoDragDrop'
failed with error 0x800401f0 (there is russian message mean CoInitialize was not
called).
08:18:02: Error: Drag & drop operation failed.

After stupid hack (::CoInitialize(0) inserted into OnInit()) it look like:

08:23:07: Debug: In file ..\..\src\msw\ole\droptgt.cpp at line 340:
'RegisterDragDrop' failed with error 0x8007000e (russian message: not enough
memory to finish operation).
08:23:38: Trace: (ole) wxIDataObject::EnumFormatEtc
08:23:38: Trace: (ole) wxIEnumFORMATETC::Next
08:23:38: Trace: (ole) wxIEnumFORMATETC::Next
08:23:38: Trace: (ole) wxIEnumFORMATETC::Reset
08:23:38: Trace: (ole) wxIEnumFORMATETC::Next
08:23:38: Trace: (ole) wxIEnumFORMATETC::Next
08:23:38: Trace: (ole) wxIDataObject::QueryGetData: CF_HDROP unsupported

Sorry, i have no enough knowledge in windows to go any futher..


February 02, 2004
In article <bvifp8$1s1m$1@digitaldaemon.com>, chris elliott says...
>Thanks for reporting this ; I have not tried the DnD sample for some months and will look at it again. I know there was an issue with the OLE clipboard and DnD but thought it was fixed (at least in CVS).

Thanks! I was not seen changes in src/msw/ole in cvs but will look more closely.. Please see my post to Arjan Knepper with actual error messages.


February 02, 2004

Dmitry Morozhnikov wrote:

> In article <bvjn4t$ojf$1@digitaldaemon.com>, Arjan Knepper says...
> 
>>I don't know. What "messages" do you see?
>>
>>I just tried the DnD sample form wxw-2.4.2 (with some small modifications) with 8.38 I don't see any problem.
> 
> 
> It look like with a library compiled with __WXDEBUG__:
> 
> 08:17:35: Debug: In file ..\..\src\msw\ole\droptgt.cpp at line 332:
> 'CoLockObjectExternal' failed with error 0x800401f0 (there is russian message
> mean CoInitialize was not called).

yes, there was a CoInitialize missing bug, which is fixed in both 2.4 and head CVS (src/msw/app.cpp line 78 needs to read

#if defined(__GNUWIN32_OLD__) || defined(__SYMANTEC__) || defined(__SALFORDC__)

chris