- Wine-Devel - list.winehq.org

Program To Check System Metrics
by William Poetra Yoga H Oct. 23, 2004

Oct. 23, 2004

Hi guys (& gals), if you remember I was working on the caption buttons. I haven't submitted the patch yet (it's still in progress) and I wish to know the relations between system metrics, because I think they are very closely related to drawing on the screen, like borders, sizes, etc. I'm writing an app here, called metrics, but since this is my first Win32 program, I've encountered many problems (and solved only a few :-P). I've read bits of theForge's Win32 tutorial at www.winprog.org (I got the link from Wine), but I just can't figure out how to handle the "Refresh" and "Save" buttons properly. "Refresh" is for refreshing (regetting) the metric values, and save is for saving the values to a file. Please help me with this, and be informed that this is my first Win32 app, so bear with me... File is attached. _______________________________ Do you Yahoo!? Declare Yourself - Register online to vote today! http://vote.yahoo.com

1 0

IE6 install Regression
by Mike McCormack Oct. 23, 2004

Oct. 23, 2004

Hi Eric, The following patch causes the IE6 installer to crash. Thanks to Rob for figuring out which patch caused the problem very quickly :) Mike Log message: Eric Pouech <pouech-eric(a)wanadoo.fr> - msvcrt: the file descriptors are now inherited between parent/child processes - kernel32.CreateProcess: make use of the *reserved2 fields for C-RunTime data inheritance http://www.winehq.org/hypermail/wine-cvs/2004/09/0039.html ... 001f:Call kernel32.GetStartupInfoA(77adfca4) ret=500b34f6 001f:Ret kernel32.GetStartupInfoA() retval=77adfca4 ret=500b34f6 trace:seh:EXC_RtlRaiseException code=c0000005 flags=0 addr=0x500b354f trace:seh:EXC_RtlRaiseException info[0]=00000000 trace:seh:EXC_RtlRaiseException info[1]=780704e4 trace:seh:EXC_RtlRaiseException eax=00570000 ebx=77efd6e8 ecx=00000040 edx=77b004e0 esi=780704e4 edi=500c7ab8 trace:seh:EXC_RtlRaiseException ebp=77adfcec esp=77adfc98 cs=0073 ds=007b es=007b fs=0033 gs=0000 flags=00210 202 trace:seh:EXC_CallHandler calling handler at 0x77c0e240 code=c0000005 flags=0

2 3

Re: Fix for bug 824 - proper handling of REG_MULTI_SZ
by Alexandre Julliard Oct. 23, 2004

Oct. 23, 2004

Randy Galbraith <randy.galbraith(a)cox.net> writes: > This proposed patch addresses the concern raised by bug 824. This > patch changes the handling of REG_MULTI_SZ from REG_SZ to REG_BINARY. Do you actually have an app that depends on this? I agree that in theory we should preserve the data exactly, but this means giving up on an easily editable registry format, so it should be done only if something really needs it. -- Alexandre Julliard julliard(a)winehq.org

3 5

Re: winapi_check and 16 bit -register functions
by Alexandre Julliard Oct. 22, 2004

Oct. 22, 2004

Francois Gouget <fgouget(a)free.fr> writes: > Yesterday's winapi_check path warns about 32 bit -register functions > that are implemented by stdcall functions, but not about 16 bit > -register functions implemented by stdcall functions. The reason is > that for the latter the calling convention is set to 'pascal' or > 'pascal16' in winapi_check. However I suspect these too should be > implemented as cdecl. They could, but that will require some changes to the .spec.c code, right now the calling convention is based on the type specified in the spec file, so pascal register functions have to be WINAPI. -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: [dx9-12] Major change to desgin
by Alexandre Julliard Oct. 22, 2004

Oct. 22, 2004

"Ann and Jason Edmeades" <us(a)the-edmeades.demon.co.uk> writes: > I have changed the design so only the Direct3D<interface><8/9> ones extend > IUnknown, and so there is no reference counting of the IWineD3D > objects. I don't think that's a good idea. If you are using COM interfaces then you need to use them the COM way, this means IWineD3D should derive from IUnknown and freeing it should be done through the Release method. If you don't want to use standard COM then don't declare COM interfaces at all, create your own mechanism. > Instead, when any Direct3D<interface><8/9> is Released to zero, then it > calls a Free method in its WineD3D equivalent interface. I also changed the > IWineD3D interface to contain, as a void *, a pointer to its 'parent' (by > which I mean the Direct3D8/9 interface) plus a function (GetParent) which > returns that pointer. This is bad too, you shouldn't use void*, you should use proper types, at the very least IUnknown*, but preferably define an IWineD3DUser interface or something like that. -- Alexandre Julliard julliard(a)winehq.org

2 4

Wine on OpenBSD
by Sascha Hanse Oct. 22, 2004

Oct. 22, 2004

> >> I just tried to install the recent Wine-tarball (20041019) on my >> OpenBSD 3.6. First I had to remove kthread.c:275: the RFTHREAD flag. >> Then the'make depend && make' worked properly but the make install >> first fails with: >> >> /usr/bin/install -c serialui.dll.so >> /usr/local/lib/wine/serialui.dll.so cd `dirname setupapi/__install__` >> && make install gcc -g -O2 -o install install.c > >Your make tries to build the 'make install' target from install.c, >that won't work of course. I guess we'll have to rename that file, >unless OpenBSD make has an equivalent of the .PHONY target. > >-- >Alexandre Julliard >julliard_at_winehq.org Hi Julliard, thank you for your response. Could you tell me when or how this can be done? -- Anyone who considers arithmetical methods of producing random digits is, of course, in a state of sin. --John von Neumann (1951)

2 2

Re: Add list for monitoring Appdb
by Jeremy Newman Oct. 22, 2004

Oct. 22, 2004

I'm holding off on this patch until I get more comments on why we should or shouldn't create yet another mailing list. My personal opinion is that we don't need a mailing list for this. The change emails can go straight to the app owners themselves, and possibly to the appdb admins. I know I wouldn't care seeing update emails for apps I don't maintain. Just my opinion. I will create this list if there is enough interest in it. On Wed, 2004-10-13 at 10:38 -0700, Tony Lambregts wrote: > I am working on a series of changes that will let changes to the appdb > be emailed to administrators, application owners and other users. > > This will be similar to how bugzilla does it. > > When someone changes an application an email listing the changes gets > sent to the following > > - Appdb administrators. > - Application owners. > - Aplication monitors. > > One of the things that I think is a repository the way wine-bugs(a)winehq.org > does for bugzilla > > I need Jeremy to set up the email account and the mailman account. > > Change log: Add link for monitoring changes to application database > > files changed: lostwages/templates/en/forums.template > > > Tony Lambregts > > Index: forums.template > =================================================================== > RCS file: /home/wine/lostwages/templates/en/forums.template,v > retrieving revision 1.16 > diff -u -r1.16 forums.template > --- forums.template 15 Sep 2004 20:05:31 -0000 1.16 > +++ forums.template 13 Oct 2004 15:46:24 -0000 > @@ -97,6 +97,13 @@ > <a href="http://bugs.winehq.org/">Bug Tracking Database</a>. > > <li> > + [<a href="http://www.winehq.org/mailman/listinfo/wine-appdb">(Un-)Subscribe</a>] > + [<a href="http://www.winehq.org/hypermail/wine-appdb/">Archive </a>] > + <a href="mailto:wine-appdb(a)winehq.org">wine-appdb(a)winehq.org</a> > + A low traffic (2/day) read-only list for monitoring submissions to the > + <a href="http://appdb.winehq.org/">Application Database</a>. > + > + <li> > [<a href="http://www.winehq.org/mailman/listinfo/wine-tests-results">(Un-)Subscribe</a>] > [<a href="http://www.winehq.org/hypermail/wine-tests-results/">Archive</a>] > <a href="mailto:wine-tests-results(a)winehq.org">wine-tests-results(a)winehq.org</a> > > >

3 2

Re: shellpath.c: Avoid copying invalid data on error
by Alexandre Julliard Oct. 22, 2004

Oct. 22, 2004

Juan Lang <juan_lang(a)yahoo.com> writes: > The following is a better patch, I think, in that > - it doesn't add a break statement to the default case > (some compilers warn on this) I seriously doubt that, what compiler does this for you? It's much better practice to always add break to all cases, otherwise when someone adds some code at the end of the switch they will forget to add a break to the previous case. > - WCHAR szBuildPath[MAX_PATH], szTemp[MAX_PATH]; > + WCHAR szBuildPath[MAX_PATH], szTemp[MAX_PATH] = { 0 }; This fills the whole buffer with zeroes, that's not really needed. -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: dlls/shlwapi PathIsValidChar
by Alexandre Julliard Oct. 22, 2004

Oct. 22, 2004

Ge van Geldorp <gvg(a)reactos.com> writes: > Did this patch: http://www.winehq.org/hypermail/wine-patches/2004/09/0446.html > fall through the cracks or is it not acceptable? I think a lookup table would be more elegant (and faster...) -- Alexandre Julliard julliard(a)winehq.org

1 0

unaligned access bug in winspool on sparc
by Eric Frias Oct. 22, 2004

Oct. 22, 2004

I just ran across an evil little bug in the WINSPOOL_GetPrinter_2 function. It looks like this type of bug could be hiding in other API functions too. It causes a segmentation fault because of an unaligned access on Solaris (sparc). This function packs a PRINTER_INFO_2 structure and all of its variable-length fields into one contiguous block of memory. The structure is first, then several strings, and then a DEVMODE structure. The problem is that because the strings can be any length, the DEVMODE structure itself could be placed at an unaligned (odd) address. So on our system, wine runs correctly if we only have printers with an even number of letters in their name in the printcap file, but it crashes if any have an odd number of letters in their name. I put in a quick hack that aligns it to a 4-byte boundary. I think a two-byte boundary would suffice, but I decided to err on the safe side since I wasn't sure. I inserted the following block near line 1822 in info.c: ------ { int bytesOfPadding = (4 - (*pcbNeeded % 4)) % 4; if (space && bytesOfPadding <= left) { ptr += bytesOfPadding; left -= bytesOfPadding; } else space = FALSE; *pcbNeeded += bytesOfPadding; } ------ right before the line: ------ if(WINSPOOL_GetDevModeFromReg(hkeyPrinter, Default_DevModeW, ptr, left, &size, unicode)) { ----- What is the correct way to align this structure? Are there any macros that specify the necessary alignment for each architecture? I'm also wondering if there are any other functions that are likely to contain the same bug. Thanks, Eric

1 0