http://bugs.winehq.org/show_bug.cgi?id=16502
Summary: InstallAnywhere (java) just displays en empty canvas Product: Wine Version: 1.1.10 Platform: PC OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: directx-ddraw AssignedTo: wine-bugs@winehq.org ReportedBy: aelschuring@hotmail.com CC: stefan@codeweavers.com
This is related to bug #11603. InstallAnywhere is Java-based installation software used in Football Manager 08. After extracting, the installer is started with the following command line (shortened for brevity):
cwd: C:\windows\temp\I1229269847\Windows cmd: ".\resource\jre\bin\javaw.exe" -Xms16777216 -Xmx50331648 -classpath "..." com.zerog.lax.LAX "./setup.lax" "C:/windows/temp/lax6771.tmp"
The installer window appears, but is all grey. The screen should display an image and a language selection menu. Clicking in the area where the menu should be, a box is shown, but again it is all grey.
There are multiple separate issues with this installer (one is that confirmation dialogs are spawned infinitely). This bug will make it very hard to bisect the other bugs, but I'll see what I can do.
Bisection results follow. I'm a little surprised to see a d3d commit, I had expected to find a ddraw or a gdi patch. Is this plausible or did I mess up?
34b37fe2411d83f1a6f3e11a7a1e4d28bd9b36ba is first bad commit commit 34b37fe2411d83f1a6f3e11a7a1e4d28bd9b36ba Author: Stefan Dösinger stefan@codeweavers.com Date: Tue Aug 5 14:24:10 2008 -0500
d3d: Remove IWineD3DDevice::SetHwnd.
http://bugs.winehq.org/show_bug.cgi?id=16502
--- Comment #1 from H. Verbeet hverbeet@gmail.com 2008-12-14 11:00:53 --- (In reply to comment #0)
Bisection results follow. I'm a little surprised to see a d3d commit, I had expected to find a ddraw or a gdi patch. Is this plausible or did I mess up?
It's plausible. Wine's ddraw is implemented on top of wined3d. Note that the patch modifies both wined3d and ddraw. IIRC, you can use "-Dsun.java2d.noddraw=true" to prevent Java from using ddraw, that should help with bisecting.
http://bugs.winehq.org/show_bug.cgi?id=16502
--- Comment #2 from Arno Schuring aelschuring@hotmail.com 2008-12-14 13:24:52 --- Thanks for the pointer. I can confirm that the same problem still exists in current git (hadn't tested beyond 1.1.10), and that the problem disappears if I start the installer manually with -Dsun.java2d.noddraw=true added on the command line. I'll try to find a small, freely distributable Java app that has the same problem.
As an aside, I also tried simply reverting above commit on top of the current master tip, but that (predictably) results in merge conflicts.
http://bugs.winehq.org/show_bug.cgi?id=16502
--- Comment #3 from Arno Schuring aelschuring@hotmail.com 2008-12-14 15:04:32 --- well I found one app that I can reproduce it with, it's at http://actualdownload.com/Risk---Classic-Risk-Board-Game-Clone_review.html .
However, recent JRE versions do not exhibit this problem: using an existing JRE 6 (build 1.6.0_03-b05) or a freshly downloaded Java5 RE (build 1.5.0_17-b04) I could not reproduce this issue. Luckily, the problem still appears in the latest JRE 1.4 (build 1.4.2_19-b04), which can still be found at http://java.sun.com/j2se/1.4.2/download.html
Start the game with $ wine ../jre14/bin/javaw.exe -jar Risk.jar to get only a gray window $ wine ../jre14/bin/javaw.exe -Dsun.java2d.noddraw=true -jar Risk.jar to get the game as intended
The original VM identifies itself as: Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_03-b07). If required (and legally sound), I can supply this specific VM as well.
Final note: there is still a maybe-related redraw issue present in recent JRE versions: if you wait for a few seconds on the main screen (in the app mentioned above), a popup message will tell you that there is a newer version available. That dialog gets painted only partially, and sometimes disappears entirely after dragging it around.
http://bugs.winehq.org/show_bug.cgi?id=16502
Vincent Povirk madewokherd@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |madewokherd@gmail.com
http://bugs.winehq.org/show_bug.cgi?id=16502
--- Comment #4 from Vincent Povirk madewokherd@gmail.com 2008-12-14 16:33:55 --- Created an attachment (id=17937) --> (http://bugs.winehq.org/attachment.cgi?id=17937) patch to ignore clipper windows
There were some changes to ddraw's clipping that I had some part in instigating (but that Stefan ultimately implemented). If a program sets a clipper with a window on the primary surface, our ddraw will generally draw to that window instead of the guess we would otherwise use.
Here is a patch to ignore clipper windows entirely. Does it have any effect?
http://bugs.winehq.org/show_bug.cgi?id=16502
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |patch
http://bugs.winehq.org/show_bug.cgi?id=16502
--- Comment #5 from Arno Schuring aelschuring@hotmail.com 2008-12-17 05:13:35 --- sorry, no change with proposed patch after incremental build. I will try again with a full build, but I don't expect that will change anything.
http://bugs.winehq.org/show_bug.cgi?id=16502
--- Comment #6 from Andrew Nguyen arethusa26@gmail.com 2008-12-21 04:17:04 --- This seems to be a duplicate of bug 15893.
http://bugs.winehq.org/show_bug.cgi?id=16502
--- Comment #7 from Arno Schuring aelschuring@hotmail.com 2008-12-21 12:58:35 --- Yes it appears to be a duplicate of #15983, considering comment #5 and #7 indicate that the problem goes away when using a more recent JRE. But then again, my bisection indicates this problem was introduced between 1.0.2 and 1.0.3, and one reporter on that bug mentions that Wine 1.0.0 doesn't work either (to be certain, I just tested wine-1.0.1 (there is no tag wine-1.0.0?) and the installer works correctly). Finally, I don't have FM2009 so I can't verify (the URL in the report leads to a 404).
If you are able to test FM2009, try to run the Java installer manually with -Dsun.java2d.noddraw=true added on the command line. If that works, my report is indeed a duplicate. Bisection information should then be added to bug #15983 as well.
(and re #5: performing a full build made no difference)
http://bugs.winehq.org/show_bug.cgi?id=16502
--- Comment #8 from Arno Schuring aelschuring@hotmail.com 2008-12-21 13:00:45 --- (In reply to comment #7)
Yes it appears to be a duplicate of #15983
Typo: should be bug #15893 of course
http://bugs.winehq.org/show_bug.cgi?id=16502
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED Resolution| |DUPLICATE
--- Comment #9 from Austin English austinenglish@gmail.com 2008-12-22 12:17:56 --- Dupe
*** This bug has been marked as a duplicate of bug 15893 ***
http://bugs.winehq.org/show_bug.cgi?id=16502
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #10 from Austin English austinenglish@gmail.com 2008-12-22 12:18:02 --- Closing.
http://bugs.winehq.org/show_bug.cgi?id=16502
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Component|directx-ddraw |directx-d3d
-
wine-bugs＠winehq.org