Re: Scrollbars, an application bug or a wine regression ? (#2314)
"George Marshall" <g.marshall(a)nospam.com> wrote:
Does reverting of the following patch help?
Yes it does make the application work as it did before the June Release :-)
Do you have a small application I could test with? -- Dmitry.
Dmitry Timoshkov wrote:
Do you have a small application I could test with?
Unfortunatly not :-( , the application you see in the screenshots (bug #2314) is run by a cobol runtime which is not freely distributable, it needs a license key. But I could test things for you if you want. Regards. PS: You could try with the application submitted in bug #2302
"George Marshall" <g.marshall(a)nospam.com> wrote:
Do you have a small application I could test with?
Unfortunatly not :-( , the application you see in the screenshots (bug #2314) is run by a cobol runtime which is not freely distributable, it needs a license key.
But I could test things for you if you want.
Perhaps you could at least try to diff relay traces with and without an offending patch applied and send the results to me.
PS: You could try with the application submitted in bug #2302
The app attached to the bug shows identical behaviour with 20040615 and 20040505. -- Dmitry.
Dmitry Timoshkov wrote:
Perhaps you could at least try to diff relay traces with and without an offending patch applied and send the results to me.
Sure, I'l do that as soon as I can.
PS: You could try with the application submitted in bug #2302
The app attached to the bug shows identical behaviour with 20040615 and 20040505.
Then bug #2314 and #2302 are two different issues, and not the same thing as Nicolai Kuntze wrote in this trhead. PS: for further reference you can directly write to g dot gabriele AT teamSPAM dot com (replace 'SPAM' with 'system')
Sorry, I really thought the errors are linked. Can someone have a look at my case (2302). If someone tells me where to start I can try to debug it myself ... Am 05.07.2004 um 15:29 schrieb George Marshall:
Dmitry Timoshkov wrote:
Perhaps you could at least try to diff relay traces with and without an offending patch applied and send the results to me.
Sure, I'l do that as soon as I can.
PS: You could try with the application submitted in bug #2302 The app attached to the bug shows identical behaviour with 20040615 and 20040505.
Then bug #2314 and #2302 are two different issues, and not the same thing as Nicolai Kuntze wrote in this trhead.
PS: for further reference you can directly write to g dot gabriele AT teamSPAM dot com (replace 'SPAM' with 'system')
---- Diese Nachricht wurde auf Viren und andere gefaehrliche Inhalte sowie Spam untersucht.
"Nicolai Kuntze" <g-8(a)gmx.net> wrote:
Sorry, I really thought the errors are linked. Can someone have a look at my case (2302). If someone tells me where to start I can try to debug it myself ...
Did you try to run your app with current CVS? It has some scroll regressions fixed. -- Dmitry.
Yes, of course. I would not ask if I hadn't. This is about 5 hours ago. But my problem still exists. I do not think that there is some fundamental problem .... but unfortunatly I can not debug it myself. If I know exactly what you need from the developer of the program I can see what I can do because the people want to sell us the system :-) Am 14.07.2004 um 11:02 schrieb Dmitry Timoshkov:
"Nicolai Kuntze" <g-8(a)gmx.net> wrote:
Sorry, I really thought the errors are linked. Can someone have a look at my case (2302). If someone tells me where to start I can try to debug it myself ...
Did you try to run your app with current CVS? It has some scroll regressions fixed.
-- Dmitry.
---- Diese Nachricht wurde auf Viren und andere gefaehrliche Inhalte sowie Spam untersucht.
"Nicolai Kuntze" <g-8(a)gmx.net> wrote:
Yes, of course. I would not ask if I hadn't. This is about 5 hours ago. But my problem still exists. I do not think that there is some fundamental problem .... but unfortunatly I can not debug it myself. If I know exactly what you need from the developer of the program I can see what I can do because the people want to sell us the system :-)
All I need is a small test case which shows the problem. -- Dmitry.
I have the program online with some help how to get it started and how to get the error/problem. I hope you have seen this. Does this help you? I try to get from the developer some test case but I can not promise anything. Nicolai Am 15.07.2004 um 15:15 schrieb Dmitry Timoshkov:
"Nicolai Kuntze" <g-8(a)gmx.net> wrote:
Yes, of course. I would not ask if I hadn't. This is about 5 hours ago. But my problem still exists. I do not think that there is some fundamental problem .... but unfortunatly I can not debug it myself. If I know exactly what you need from the developer of the program I can see what I can do because the people want to sell us the system :-)
All I need is a small test case which shows the problem.
-- Dmitry.
---- Diese Nachricht wurde auf Viren und andere gefaehrliche Inhalte sowie Spam untersucht.
"Nicolai Kuntze" <g-8(a)gmx.net> wrote:
I have the program online with some help how to get it started and how to get the error/problem. I hope you have seen this. Does this help you? I try to get from the developer some test case but I can not promise anything.
If I can't get an .exe or a code snippet (with a description how it's intended to behave) showing the problem I can't do anything to understand and eventually to solve the problem. -- Dmitry.
Am 16.07.2004 um 15:25 schrieb Dmitry Timoshkov:
"Nicolai Kuntze" <g-8(a)gmx.net> wrote:
I have the program online with some help how to get it started and how to get the error/problem. I hope you have seen this. Does this help you? I try to get from the developer some test case but I can not promise anything.
If I can't get an .exe or a code snippet (with a description how it's intended to behave) showing the problem I can't do anything to understand and eventually to solve the problem.
-- Dmitry.
Today I got some modified code snipplet: case WM_PAINT: PAINTSTRUCT MalInfo; BeginPaint( hwnd, &MalInfo); EndPaint( hwnd, &MalInfo); HDC hdc= GetDC( hwnd); BitBlt( hdc, 0, 20, doublebuffersize.right-doublebuffersize.left, doublebuffersize.bottom-doublebuffersize.top, doublebufferHdc, 0, 0, SRCCOPY); ReleaseDC( (HWND)ViewGetHWND(), hdc); I hope this helps. Yours, Nicolai ---- Diese Nachricht wurde auf Viren und andere gefaehrliche Inhalte sowie Spam untersucht.
Am 19.07.2004 um 17:10 schrieb Nicolai Kuntze:
Am 16.07.2004 um 15:25 schrieb Dmitry Timoshkov:
"Nicolai Kuntze" <g-8(a)gmx.net> wrote:
I have the program online with some help how to get it started and how to get the error/problem. I hope you have seen this. Does this help you? I try to get from the developer some test case but I can not promise anything.
If I can't get an .exe or a code snippet (with a description how it's intended to behave) showing the problem I can't do anything to understand and eventually to solve the problem.
-- Dmitry.
Today I got some modified code snipplet:
case WM_PAINT: PAINTSTRUCT MalInfo; BeginPaint( hwnd, &MalInfo); EndPaint( hwnd, &MalInfo); HDC hdc= GetDC( hwnd); BitBlt( hdc, 0, 20, doublebuffersize.right-doublebuffersize.left, doublebuffersize.bottom-doublebuffersize.top, doublebufferHdc, 0, 0, SRCCOPY); ReleaseDC( (HWND)ViewGetHWND(), hdc);
I hope this helps.
Yours,
Nicolai
Hi, I have done some debugging. I use the native comctrl.dll. So the error must be in the gui part of wine. The calls of the software are all OK, but the scrollbars were not shown. Is it possible that wine does not recognize that the bitmap is greater than the window? Where can I look for it? NIcolai ---- Diese Nachricht wurde auf Viren und andere gefaehrliche Inhalte sowie Spam untersucht.
"Nicolai Kuntze" <g-8(a)gmx.net> wrote:
Today I got some modified code snipplet:
case WM_PAINT: PAINTSTRUCT MalInfo; BeginPaint( hwnd, &MalInfo); EndPaint( hwnd, &MalInfo); HDC hdc= GetDC( hwnd); BitBlt( hdc, 0, 20, doublebuffersize.right-doublebuffersize.left, doublebuffersize.bottom-doublebuffersize.top, doublebufferHdc, 0, 0, SRCCOPY); ReleaseDC( (HWND)ViewGetHWND(), hdc);
I hope this helps.
No, that doesn't help.
I have done some debugging. I use the native comctrl.dll. So the error must be in the gui part of wine. The calls of the software are all OK, but the scrollbars were not shown. Is it possible that wine does not recognize that the bitmap is greater than the window? Where can I look for it?
Are you sure that your app really regressed and had not shown the same behaviour before my scrollbar fixes? -- Dmitry.
Am 24.07.2004 um 16:34 schrieb Dmitry Timoshkov:
"Nicolai Kuntze" <g-8(a)gmx.net> wrote:
Today I got some modified code snipplet:
case WM_PAINT: PAINTSTRUCT MalInfo; BeginPaint( hwnd, &MalInfo); EndPaint( hwnd, &MalInfo); HDC hdc= GetDC( hwnd); BitBlt( hdc, 0, 20, doublebuffersize.right-doublebuffersize.left, doublebuffersize.bottom-doublebuffersize.top, doublebufferHdc, 0, 0, SRCCOPY); ReleaseDC( (HWND)ViewGetHWND(), hdc);
I hope this helps.
No, that doesn't help.
I thougth this ...
I have done some debugging. I use the native comctrl.dll. So the error must be in the gui part of wine. The calls of the software are all OK, but the scrollbars were not shown. Is it possible that wine does not recognize that the bitmap is greater than the window? Where can I look for it?
Are you sure that your app really regressed and had not shown the same behaviour before my scrollbar fixes?
No I am not sure but I think this is no regession. I think there is an error in the detection in the heigth of the bitmap. But I don't khow where to start to search the error. can you give me a hint? The downloadeable archive (http://hubrichn.dyndns.org/bmed.tar.gz) is still there. There is also a readme discribing my problem. (I already said this...) Can you please have a look at this. As far as I can see all calls of BitBlt and CreateWindowEx and so on are correct implemented in bmed. ---- Diese Nachricht wurde auf Viren und andere gefaehrliche Inhalte sowie Spam untersucht.
"Nicolai Kuntze" <g-8(a)gmx.net> wrote:
Are you sure that your app really regressed and had not shown the same behaviour before my scrollbar fixes?
No I am not sure but I think this is no regession. I think there is an error in the detection in the heigth of the bitmap. But I don't khow where to start to search the error. can you give me a hint? The downloadeable archive (http://hubrichn.dyndns.org/bmed.tar.gz) is still there. There is also a readme discribing my problem. (I already said this...) Can you please have a look at this. As far as I can see all calls of BitBlt and CreateWindowEx and so on are correct implemented in bmed.
I have that app already and had a look at it. It looks like a problem with listview control, but besides I'm not an expert in that area, I have lots of other things assigned to me to do. -- Dmitry.
Dmitry Timoshkov wrote:
Perhaps you could at least try to diff relay traces with and without an offending patch applied and send the results to me.
Iv'e tried to run wine exporting WINEDEBUG=trace+relay with both the unpatched 20040615 version and the patched 20040615 version. 196M Jul 6 10:59 /tmp/wine-20040615-patched.trace-relay.log 158M Jul 6 11:32 /tmp/wine-20040615-unpatched.trace-relay.log The diff between the two file is (big) 185M Jul 6 11:34 /tmp/wine-20040615.diff I'm attaching the diff (bzipped(2)): 2.4M Jul 6 11:34 /tmp/wine-20040615.diff.bz2 PS: for further reference you can directly write to g dot gabriele AT teamSPAM dot com (replace 'SPAM' with 'system')
"George Marshall" <g.marshall(a)nospam.com> wrote:
Iv'e tried to run wine exporting WINEDEBUG=trace+relay with both the unpatched 20040615 version and the patched 20040615 version.
196M Jul 6 10:59 /tmp/wine-20040615-patched.trace-relay.log 158M Jul 6 11:32 /tmp/wine-20040615-unpatched.trace-relay.log
The diff between the two file is (big) 185M Jul 6 11:34 /tmp/wine-20040615.diff
I'm attaching the diff (bzipped(2)): 2.4M Jul 6 11:34 /tmp/wine-20040615.diff.bz2
There was no need to send a huge attachment to the mailing list. Please next time place log files somewhere on web, and post an url to them. In order to make files smaller you have to do some real work: find out all places where SetScrollRange is called, compare them in both logs, and find out what has changed. Then post not just a diff, but full relay trace for compared SetScrollRange calls - old and new. Since SetScrollRange is the only API which has changed it should be easy enough to find a regression. -- Dmitry.
participants (4)
-
Dmitry Timoshkov -
Dmitry Timoshkov -
George Marshall -
Nicolai Kuntze