January 2009 - Wine-Devel - list.winehq.org

Re: widl: print error on [out, string] without size_is
by Michael Karcher Jan. 1, 2009

Jan. 1, 2009

Am Donnerstag, den 01.01.2009, 11:22 +0100 schrieb Michael Karcher: > The old behaviour was to emit a half-complete C statement without > any error message. This patch is bad. Please ignore. Regards, Michael Karcher

1 0

Re: ole32: Add HICON user marshalling stubs
by Rob Shearman Jan. 1, 2009

Jan. 1, 2009

2009/1/1 Michael Karcher <wine(a)mkarcher.dialup.fu-berlin.de>: > diff --git a/dlls/ole32/ole32.spec b/dlls/ole32/ole32.spec > index 79376ac..f7db2b0 100644 > --- a/dlls/ole32/ole32.spec > +++ b/dlls/ole32/ole32.spec > @@ -122,6 +122,10 @@ > @ stdcall HBITMAP_UserMarshal(ptr ptr ptr) > @ stdcall HBITMAP_UserSize(ptr long ptr) > @ stdcall HBITMAP_UserUnmarshal(ptr ptr ptr) > +@ stdcall HICON_UserFree(ptr ptr) > +@ stdcall HICON_UserMarshal(ptr ptr ptr) > +@ stdcall HICON_UserSize(ptr long ptr) > +@ stdcall HICON_UserUnmarshal(ptr ptr ptr) > @ stub HBRUSH_UserFree > @ stub HBRUSH_UserMarshal > @ stub HBRUSH_UserSize The functions should be added in the correct place so that the list of functions remains sorted. -- Rob Shearman

1 0

Re: [oleaut 1/3] include: add [unique] attributes to IPropertyBag::RemoteRead args
by Rob Shearman Jan. 1, 2009

Jan. 1, 2009

2009/1/1 Michael Karcher <wine(a)mkarcher.dialup.fu-berlin.de>: > diff --git a/include/oaidl.idl b/include/oaidl.idl > index 1cc166a..4ee3b43 100644 > --- a/include/oaidl.idl > +++ b/include/oaidl.idl > @@ -1620,9 +1620,9 @@ interface IPropertyBag : IUnknown > HRESULT RemoteRead( > [in] LPCOLESTR pszPropName, > [out] VARIANT *pVar, > - [in] IErrorLog *pErrorLog, > + [in, unique] IErrorLog *pErrorLog, > [in] DWORD varType, > - [in] IUnknown *pUnkObj); > + [in, unique] IUnknown *pUnkObj); In the PSDK only pErrorLog is a unique pointer. However, I believe you are making this change to avoid getting RPC_X_NULL_REF_POINTER exceptions when a NULL pErrorLog parameter is passed into IPropertyBag_Read, but MIDL ignores the pointer type for interface pointers and never outputs a NULL reference pointer check for them. So widl just needs to be fixed. > > HRESULT Write( > [in] LPCOLESTR pszPropName, -- Rob Shearman

2 2

Re: [oleaut 3/3] oleaut32: Implement IPropertyBag::Read proxying
by Rob Shearman Jan. 1, 2009

Jan. 1, 2009

2009/1/1 Michael Karcher <wine(a)mkarcher.dialup.fu-berlin.de>: > + hr = IPropertyBag_RemoteRead_Proxy(This, pszPropName, &outVariant, pErrorLog, > + V_VT(pVar), pUnk); > + if(SUCCEEDED(hr)) > + hr = VariantCopy(pVar, &outVariant); > + > + return hr; You're leaking the memory in outVariant here, since VariantCopy does a deep copy of all of the data. Is there a reason you don't just pass pVar into IPropertyBag_RemoteRead_Proxy? -- Rob Shearman

2 2

Re: [oleaut 2/3] include: Add [unique] attribute to IPersistPropertyBag::Load arg
by Rob Shearman Jan. 1, 2009

Jan. 1, 2009

2009/1/1 Michael Karcher <wine(a)mkarcher.dialup.fu-berlin.de>: > diff --git a/include/ocidl.idl b/include/ocidl.idl > index c40d155..d4979d7 100644 > --- a/include/ocidl.idl > +++ b/include/ocidl.idl > @@ -927,7 +927,7 @@ interface IPersistPropertyBag : IPersist > > HRESULT Load( > [in] IPropertyBag *pPropBag, > - [in] IErrorLog *pErrorLog); > + [in,unique] IErrorLog *pErrorLog); > > HRESULT Save( > [in] IPropertyBag *pPropBag, Same goes for this change. -- Rob Shearman

1 0

Drumming up qa publicity for sketchup on wine via Digg
by Dan Kegel Jan. 1, 2009

Jan. 1, 2009

Sketchup seems to be working pretty well, so I just posted http://digg.com/linux_unix/Google_Sketchup_7_working_in_Wine_1_1_11 to see if I could attract some real sketchup users to find more bugs. If folks agree that's a good idea, they can support it by digging it so it can show up on the front page. Thanks, Dan

2 2

Contribution to the Toolbar common control
by Alan Feldman Jan. 1, 2009

Jan. 1, 2009

Hello Wine developers, I am writing to this forum because I wanted to contribute some useful development information to your community while at the same time not have to get too involved as I currently have no plans to stay involved going forward. I wanted to say that I am *really* impressed with the codebase that you guys have developed! I stumbled across the Winehq site as I was looking for more information on Microsoft's Toolbar common control. The source code you guys have has made understanding the internals of the Toolbar control really easy and I really appreciate the work you guys have done. (I even took a peek at Microsoft's comctl32.dll disassembly to confirm a couple of the things you guys wrote and it checked out) ;-) So, anyway, in gratitude for what you guys have provided me I wanted to "give something back". I noticed that there are several Toolbar window messages which are listed as "unknown" which at the time of the writing of this e-mail are now definitely "known" and I would like to provide that information to someone to check in to your source code base. Here is my contribution: http://source.winehq.org/source/dlls/comctl32/toolbar.c Function: ToolbarWindowProc() TB_UNKWN45E is really TB_SETHOTITEM2 (WM_USER + 94) TB_UNKWN460 is really TB_SETLISTGAP (WM_USER + 96) TB_UNKWN462 is really TB_GETIMAGELISTCOUNT (WM_USER + 98) TB_UNKWN463 is really TB_GETIDEALSIZE (WM_USER + 99) TB_UNKWN467 (which is not in the Wine source yet) is really TB_GETITEMDROPDOWNRECT (WM_USER + 103) Those Toolbar window messages are currently not in the public Windows headers at this time but they have been *documented* on the MSDN Library web site now. Here are links to the MSDN Library help topics giving evidence to what I wrote up above: http://msdn.microsoft.com/en-us/library/cc835036(VS.85).aspx http://msdn.microsoft.com/en-us/library/bb787441(VS.85).aspx http://msdn.microsoft.com/en-us/library/cc835034(VS.85).aspx http://msdn.microsoft.com/en-us/library/cc835033(VS.85).aspx http://msdn.microsoft.com/en-us/library/cc835035(VS.85).aspx Also, in this source file: http://source.winehq.org/source/include/commctrl.h Someone took a REALLY good guess at the name of a Toolbar notification! TBN_WRAPHOTITEM is definitely the name of notification TBN_FIRST- 24. Good guess! I would say that the comment about this message being undocumented and just a "guess" can now be removed. Even though this message is also not in any public header file, this message is now *documented* and the documentation can be found here: http://msdn.microsoft.com/en-us/library/cc835032(VS.85).aspx Cheers! :-) Alan Feldman

3 5

Commit stats
by Hans Leidekker Jan. 1, 2009

Jan. 1, 2009

$ for year in {2002..2008}; do \ count=$( git log | grep ^Date: | grep $year | wc -l ); \ echo "Number of commits in $year: $count"; \ done Number of commits in 2002: 3094 Number of commits in 2003: 3283 Number of commits in 2004: 3851 Number of commits in 2005: 6006 Number of commits in 2006: 8431 Number of commits in 2007: 9532 Number of commits in 2008: 11292 -Hans

3 2

Shell32 proxies/stubs
by Michael Karcher Jan. 1, 2009

Jan. 1, 2009

Hello developers, this is just to let you know where the 3 standalone patches and three patch series I just posted to wine-devel are going: Having proxies for the COM objects of shell32.dll in actxprxy.dll, just as Windows XP and Vista does. It started with me wanting a guinea pig object that is apartment threaded and has proxies/stubs and present on every Windows installation to test a strange corner case in ole32. Looking around in my XP, I decided to use the MyComputer object that is already used in the ole32 testcase. Then I found out that Windows has them, but Wine misses them and thought to myself: Hey, this is just COM, adding a proxy should be just adding some lines to Makefile.in and widl does the rest. It turned out to be a quite fun project that revealed problems and/or missing features at nearly every component involved. The things I sent up to now fix most of the things below the proxies. What is still unsent is a three-patch-series to actxprxy to really add the proxies, and testcases for the remoting. While the shelllink testcase passes on my system even with remote calls, the shlfolder testsuite does not work correctly yet, revealing even more issues with ole32/widl I'm currently attacking. Regards, Michael Karcher

1 0