On Wed 13. August 2003 14:58, Maxime Bellengé wrote:
In TREEVIEW_UpdateDispInfo, a test is made to determine if the property we want to change is handled by a callback. If so, the change is dropped. But that test was done incorrectly.
Now the treeview in the statistical page of eMule displays correctly.
Changelog:
- Fix for TREEVIEW_UpdateDispInfo to test with the correct
callbackMask
This patch, which is now commited leads to problems for me. Now the labels in the tree fiew of some explorer test programs remain empty. May be this callback mask has to adjusted at some more points in the code. But I don't know, what it's used exactly for. Anyone any ideas?
The callbackMask is used to determine if a style that a program want to modify is handled by a callback. In this case there is no reason to do something.
NMTVDISPINFOA callback; HWND hwnd = infoPtr->hwnd;
- mask &= wineItem->callbackMask; + mask &= ~wineItem->callbackMask;
if (mask == 0) return;
My 1 line patch changes the test in TREEVIEW_UpdateDispInfo. mask is a flag with the properties you want to change, callbackMask is a flag with the properties handled with a callback. So in my opinion this test was wrong.
The change I made fixes eMule, now the statistical treeview displays correctly.
So I think my fix triggered another bug now rather than it being wrong.
Max
On Fri, 2003-08-15 at 10:44, Martin Fuchs wrote:
On Wed 13. August 2003 14:58, Maxime Bellengé wrote:
In TREEVIEW_UpdateDispInfo, a test is made to determine if the property we want to change is handled by a callback. If so, the change is dropped. But that test was done incorrectly.
Now the treeview in the statistical page of eMule displays correctly.
Changelog:
- Fix for TREEVIEW_UpdateDispInfo to test with the correct
callbackMask
This patch, which is now commited leads to problems for me. Now the labels in the tree fiew of some explorer test programs remain empty. May be this callback mask has to adjusted at some more points in the code. But I don't know, what it's used exactly for. Anyone any ideas?
Martin can you try this patch to see if it helps ?
Maxime
PS: Apply it against the current cvs
On Fri, 2003-08-15 at 11:11, Maxime Bellengé wrote:
The callbackMask is used to determine if a style that a program want to modify is handled by a callback. In this case there is no reason to do something.
NMTVDISPINFOA callback; HWND hwnd = infoPtr->hwnd;
- mask &= wineItem->callbackMask;
mask &= ~wineItem->callbackMask;
if (mask == 0) return;
My 1 line patch changes the test in TREEVIEW_UpdateDispInfo. mask is a flag with the properties you want to change, callbackMask is a flag with the properties handled with a callback. So in my opinion this test was wrong.
The change I made fixes eMule, now the statistical treeview displays correctly.
So I think my fix triggered another bug now rather than it being wrong.
Max
On Fri, 2003-08-15 at 10:44, Martin Fuchs wrote:
On Wed 13. August 2003 14:58, Maxime Bellengé wrote:
In TREEVIEW_UpdateDispInfo, a test is made to determine if the property we want to change is handled by a callback. If so, the change is dropped. But that test was done incorrectly.
Now the treeview in the statistical page of eMule displays correctly.
Changelog:
- Fix for TREEVIEW_UpdateDispInfo to test with the correct
callbackMask
This patch, which is now commited leads to problems for me. Now the labels in the tree fiew of some explorer test programs remain empty. May be this callback mask has to adjusted at some more points in the code. But I don't know, what it's used exactly for. Anyone any ideas?
Hello Maxime,
Martin can you try this patch to see if it helps ?
Maxime
PS: Apply it against the current cvs
Yes, thats better! Now the labels are visible again.
Thanks.