- Wine-Devel - list.winehq.org

Mandrake RPMs
by Mike Hearn June 29, 2004

June 29, 2004

Hi, Ivans comment that the Mandrake RPMs didn't work on Red Hat/Fedora surprised me, so after encountering a stuck Mandrake user for whom the RPMs weren't working either on IRC I decided to check them out. It turns out the reason they don't work on Fedora is because they don't work at all - wine.inf is in /usr/share/doc/wine-20040615 which is wrong, it needs to be in /usr/share/wine. As this file is in the wrong place Wine refuses to start on a new users account. Once I manually copied the file to the right place Wine worked just fine. Could somebody please fix the Mandrake RPMs, and maintainers of other packages please ensure you have details like this right. thanks -mike

1 0

Re: fix typo in wwn
by Brian Vincent June 29, 2004

June 29, 2004

> -A few other people were in favor of saving it to support a > +A few other people were in favour of saving it to support a "The Americans are identical to the British in all respects except, of course, language." - Oscar Wilde Brian

3 2

How to handle odbc/registry
by Bill Medland June 29, 2004

June 29, 2004

The builtin odbc32.dll basically simply proxies the underlying unix ODBC provider (e.g. unixODBC or iODBC). I have just discovered that somewhere in the compilation/link phase under Windows some standard ODBC function calls somehow actually get translated into direct registry calls. For example the SQLGGetInstalledDrivers call actually gets translated into an OpenKey and EnumerateKey on the HKLM/Software/ODBC/ODBCINST.INI/ODBC Drivers So should we have code within the registry code that recognises access to the ODBC tree when odbc32 is builtin and redirects it to the corresponding calls to the unix ODBC manager? -- Bill Medland mailto:billmedland(a)mercuryspeed.com http://webhome.idirect.com/~kbmed

2 1

Re: objects/region.c: REGION_FrameRgn
by Alexandre Julliard June 29, 2004

June 29, 2004

Filip Navara <xnavara(a)volny.cz> writes: > Indeed Dimtry was right about the cause of the balloon tooltips > drawing issue. Here is the code for FrameRgn ported from ReactOS that > actaully does the right thing. I don't think you should modify the source region at all, even temporarily; it could be in use somewhere else. -- Alexandre Julliard julliard(a)winehq.org

1 0

AT_SYSINFO under RedHat 8
by Bill Medland June 28, 2004

June 28, 2004

preloader.c doesn't make due to the most recent change. Presumably AT_SYSINFO is supposed to be in the Elf.h? Bill -- Bill Medland mailto:billmedland(a)mercuryspeed.com http://webhome.idirect.com/~kbmed

1 0

Crash when resizing window in Desktop mode
by Uwe Bonnes June 28, 2004

June 28, 2004

Hallo, resizing the window of ISE (Xilinx FPGA programming suite), the mouse remains captured inside the wine desktop and following backtrace can be requested: Backtrace: =>1 0x404fdc04 DebugBreak+0x4 [debugger.c:274] in kernel32 (0x406ef1fc) 2 0x41f11067 error_handler+0x87(display=0x78089550, error_evt=0x406ef248) [x11drv_main.c:186] in x11drv (0x406ef210) 3 0x4114fc6b (0x406ef2c0) 4 0x4114e086 (0x406ef300) 5 0x41148f15 (0x406ef340) 6 0x41148ff5 (0x406ef360) 7 0x41130269 (0x406ef380) 8 0x41ee08a1 BITBLT_InternalStretchBlt+0x6e1(physDevDst=0x403d8f60, xDst=0x53, yDst=0x64, widthDst=0x286, heightDst=0x4, physDevSrc=0x0, xSrc=0x0, ySrc=0x0, widthSrc=0x0, heightSrc=0x0, rop=0x5a0049) [bitblt.c:1429] in x11drv (0x406ef424) 9 0x41ee0d93 X11DRV_PatBlt+0x43(physDev=0x403d8f60, left=0x53, top=0x64, width=0x286, height=0x4, rop=0x5a0049) [bitblt.c:1533] in x11drv (0x406ef470) 10 0x408a34ec PatBlt+0x5c(hdc=0xb38, left=0x53, top=0x64, width=0x286, height=0x4, rop=0x5a0049) [bitblt.c:42] in gdi32 (0x406ef49c) 11 0x41f0d8f7 draw_moving_frame+0x67(hdc=0xb38, rect=0x406ef58c, thickframe=0x1) [winpos.c:1763] in x11drv (0x406ef4cc) 12 0x41f0e7b4 X11DRV_SysCommandSizeMove+0xae4(hwnd=0x20028, wParam=0xf008) [winpos.c:2062] in x11drv (0x406ef5c8) 13 0x407c03e1 NC_HandleSysCommand+0x1b1(hwnd=0x20028, wParam=0xf008, lParam=0x25302d3) [../../windows/nonclient.c:1499] in user32 (0x406ef5f4) 14 0x407b0b5f DEFWND_DefWinProc+0x98f(hwnd=0x20028, msg=0x112, wParam=0xf008, lParam=0x25302d3) [../../windows/defwnd.c:528] in user32 (0x406ef684) 15 0x407b10c3 DefWindowProcA+0xa3(hwnd=0x20028, msg=0x112, wParam=0xf008, lParam=0x25302d3) [../../windows/defwnd.c:916] in user32 (0x406ef6b0) 16 0x407ba2d2 DefFrameProcA+0x42(hwnd=0x20028, hwndMDIClient=0x1002a, message=0x112, wParam=0xf008, lParam=0x25302d3) [../../windows/mdi.c:1272] in user32 (0x406ef6d4) 17 0x6c376a27 (0x406ef7bc) 18 0x6c371cea (0x406ef7dc) 19 0x40e41ccc (0x0000f008) 20 0x00000000 (0x00000000) This is with today's CVS -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

3 3

Re: fix typo in wwn
by Ivan June 28, 2004

June 28, 2004

> And they don't use the metric system. Shame. The matric system is way more used in the UK than in the US, where it's hardly used at all. Ivan.

1 0

Re: Change REFIID to IID* for stack declarations in oaidl_p.c to avoid initializers being discarded
by Mike Hearn June 27, 2004

June 27, 2004

On Sun, 27 Jun 2004 21:13:17 +0200, Marcus Meissner wrote: >> + It is usually generated directly by MIDL, however this file has >> + been tweaked since then to account for slight differences in the way >> + gcc and MSVC++ compile it. In particular, in some functions REFIIDs >> + declared on the stack have been converted to plain IID* in order to eliminate >> + the constness of the REFIID type, ensuring that the zero initializer is not >> + discarded. >> + >> + Therefore, please do not regenerate this file. >> +*/ > > The comment is not understandable. > > Shouldn't it read '"REFIID" should not be used, because the value is not "const"'? I don't understand that wording of the comment either unfortunately :( It's not that it shouldn't be used, it's more that in the original generated code (I think) it was written like this: REFIID riid; (REFIID ) riid = 0; I'm not sure why MIDL did this, but as REFIID is simply #define REFIID const IID* const changing those instances to IID* should "fix" it without any warnings about assigning to a readonly variable. REFIID can still be used elsewhere though. thanks -mike

1 0

Re: Tooltips: Start of Balloon Support
by Dmitry Timoshkov June 27, 2004

June 27, 2004

"Robert Shearman" <R.J.Shearman(a)warwick.ac.uk> wrote: > This patch marks the start of balloon tip support. Note that it also has > the same bug as native due to regions not being handled the same as on > Windows. Do you mean a bug in FrameRgn? If yes, ReactOS might have that bug fixed in their implementation. Could you have a look? -- Dmitry.

2 1

Re: Change REFIID to IID* for stack declarations in oaidl_p.c to avoid initializers being discarded
by Marcus Meissner June 27, 2004

June 27, 2004

On Sun, Jun 27, 2004 at 05:23:27PM +0100, Mike Hearn wrote: > This can cause a subtle marshalling failure in some cases > (inter-thread IDispatch, native MSI, etc) > > > Mike Hearn <mh(a)codeweavers.com> > Change REFIID to IID* for stack declarations in oaidl_p.c to avoid > initializers being discarded > Index: oaidl_p.c > =================================================================== > RCS file: /home/wine/wine/dlls/oleaut32/oaidl_p.c,v > retrieving revision 1.2 > diff -u -p -w -r1.2 oaidl_p.c > --- oaidl_p.c 5 Sep 2003 23:08:33 -0000 1.2 > +++ oaidl_p.c 27 Jun 2004 16:17:04 -0000 > @@ -1,5 +1,14 @@ > -/* this ALWAYS GENERATED file contains the proxy stub code */ > +/* This file contains the proxy/stub code for core COM interfaces. > > + It is usually generated directly by MIDL, however this file has > + been tweaked since then to account for slight differences in the way > + gcc and MSVC++ compile it. In particular, in some functions REFIIDs > + declared on the stack have been converted to plain IID* in order to eliminate > + the constness of the REFIID type, ensuring that the zero initializer is not > + discarded. > + > + Therefore, please do not regenerate this file. > +*/ The comment is not understandable. Shouldn't it read '"REFIID" should not be used, because the value is not "const"'? Ciao, Marcus

1 0