October 2007 - Wine-Devel - list.winehq.org

Re: shlwapi: Update exported API entries according to the info published by Geoff Chappell
by Alexandre Julliard Oct. 25, 2007

Oct. 25, 2007

Dmitry Timoshkov <dmitry(a)codeweavers.com> writes: > @@ -1,18 +1,18 @@ > -1 stdcall -noname ParseURLA(str ptr) > -2 stdcall -noname ParseURLW(wstr ptr) > +1 stdcall ParseURLA(str ptr) > +2 stdcall ParseURLW(wstr ptr) That doesn't look right, why are you removing -noname on these and many other functions? AFAICT they are not exported by name in most shlwapi versions. -- Alexandre Julliard julliard(a)winehq.org

2 4

Re: shlwapi: Update exported API entries according to the info published by Geoff Chappell
by Robert Shearman Oct. 25, 2007

Oct. 25, 2007

Dmitry Timoshkov wrote: > +1 stdcall ParseURLA(str ptr) > +2 stdcall ParseURLW(wstr ptr) > +12 stdcall SHCreateMemStream(ptr long) > +168 stdcall ConnectToConnectionPoint(ptr ptr long ptr ptr ptr) > +169 stdcall IUnknown_AtomicRelease(long) > +172 stdcall IUnknown_GetWindow(ptr ptr) > +174 stdcall IUnknown_SetSite(ptr ptr) > +176 stdcall IUnknown_QueryService(ptr ptr ptr ptr) > +184 stdcall IStream_Read(ptr ptr long) SHIStream_Read > +185 stdcall SHMessageBoxCheckA(ptr str str long long str) > +191 stdcall SHMessageBoxCheckW(ptr wstr wstr long long wstr) > > +199 stdcall IUnknown_Set(ptr ptr) > +203 stdcall SHStripMneumonicA(str) > +204 stdcall SHIsChildOrSelf(long long) > +212 stdcall IStream_Write(ptr ptr long) SHIStream_Write > +213 stdcall IStream_Reset(ptr) > +214 stdcall IStream_Size(ptr ptr) > +215 stdcall SHAnsiToUnicode(str ptr long) > +217 stdcall SHUnicodeToAnsi(wstr ptr ptr) > +219 stdcall QISearch(long long long long) > +225 stdcall SHStripMneumonicW(wstr) > +256 stdcall IUnknown_GetSite(ptr ptr ptr) > +276 stdcall WhichPlatform() > +280 stdcall SHRegGetIntW(ptr wstr long) > +345 stdcall SHAnsiToAnsi(str ptr long) > +346 stdcall SHUnicodeToUnicode(wstr ptr long) > +353 stub SHFormatDateTimeA > +354 stub SHFormatDateTimeW > +388 varargs ShellMessageBoxW(long long wstr wstr long) ShellMessageBoxWrapW > +418 stdcall MLFreeLibrary(long) > +432 stub SHSendMessageBroadcastA > +433 stub SHSendMessageBroadcastW > +437 stdcall IsOS(long) > +462 stub UrlFixupW > +467 stub SHRunIndirectRegClientCommand > +512 stub IStream_ReadPidl > +513 stub IStream_WritePidl > These are all exported by ordinal only up to Windows 2003 SP1 (w/ IE 6.0). +196 stub -noname SHVerbExistsNA #stdcall @(ptr ptr ptr ptr) dnsapi.DnsRecordSetCompare You need to fix the name of the function in the comment here. -- Rob Shearman

2 1

Re: winex11.drv: Support HWND_TOPMOST/HWND_BOTTOM windows using NETWM states.
by Alexandre Julliard Oct. 25, 2007

Oct. 25, 2007

Robert Shearman <rob(a)codeweavers.com> writes: > */ > -static void update_wm_states( Display *display, struct x11drv_win_data *data, BOOL force ) > +static void update_wm_states( Display *display, struct x11drv_win_data *data, > + BOOL force, BOOL set_zorder, HWND insert_after ) > { > static const unsigned int state_atoms[NB_WM_STATES] = > { > XATOM__NET_WM_STATE_FULLSCREEN, > XATOM__NET_WM_STATE_SKIP_PAGER, > - XATOM__NET_WM_STATE_SKIP_TASKBAR > + XATOM__NET_WM_STATE_SKIP_TASKBAR, > + XATOM__NET_WM_STATE_ABOVE, > + XATOM__NET_WM_STATE_BELOW, It's not that easy, there's a lot more to HWND_TOPMOST than simply setting window manager hints, it needs to enforce proper ordering in the server, and set/clear the WS_EX_TOPMOST style as needed. I have that on my todo list... -- Alexandre Julliard julliard(a)winehq.org

1 0

start:start.c Display messages in the console encoding.
by Kirill K. Smirnov Oct. 25, 2007

Oct. 25, 2007

> static void output(const char *message) > { >- DWORD count; >- WriteFile(GetStdHandle(STD_OUTPUT_HANDLE), message, strlen(message), >&count, NULL); >+ DWORD count = 0; >+ WCHAR *mesW = NULL; >+ char *mes = NULL; >+ int wlen = 0; >+ int len = 0; >+ >+ wlen = MultiByteToWideChar( CP_ACP, 0, message, -1, NULL, 0 ); >+ mesW = HeapAlloc( GetProcessHeap(), 0, wlen * sizeof(WCHAR) ); >+ mes = HeapAlloc( GetProcessHeap(), 0, wlen * sizeof(char) ); >+ >+ MultiByteToWideChar( CP_ACP, 0, message, -1, mesW, wlen ); >+ len = WideCharToMultiByte( GetConsoleOutputCP(), 0, mesW, wlen, NULL, 0, >NULL, NULL ); >+ WideCharToMultiByte( GetConsoleOutputCP(), 0, mesW, wlen, mes, len, NULL, >NULL ); >+ >+ WriteFile(GetStdHandle(STD_OUTPUT_HANDLE), mes, len, &count, NULL); >+ if( !count && len ) >+ SetLastError( ERROR_FUNCTION_FAILED ); >+ >+ HeapFree(GetProcessHeap(), 0, mesW); >+ HeapFree(GetProcessHeap(), 0, mes); > } WriteFile does not output unicode characters to console screen buffer, so if ConsoleOutputCP is smth like 65001, this will fail. The page http://msdn2.microsoft.com/en-us/library/ms687401.aspx contains remedy by Microsoft how to deal with problem: 1) If device handle is console, then output wide chars using WriteConsoleW 2) If device handle is not console, then convert wide chars to ConsoleOutputCP and output using WriteFileA. This way is already implemented in cmd (See WCMD_output_asis_len function). And it may be much better to convert start utility to unicode. -- Kirill P.S. all this manipulations is supposed to be done by msvcrt printf function family, but wine lacks this functionality :-(

2 1

any clues as to what's wrong with patches (system palette entries)?
by Alexander Dorofeyev Oct. 24, 2007

Oct. 24, 2007

Hello. I've sent a patch that changes X11DRV_GetSystemPaletteEntries behavior and a test something like a week ago. I guess they fell into that so called not obviously correct category (or maybe not correct at all one). I'll appreciate any comments or suggestions on possible improvements. Thanks. > On Windows, GetSystemPaletteEntries() call in non-palettized graphics > modes returns back reserved colors from the default palette. This patch > modifies winex11.drv's X11DRV_GetSystemPaletteEntries() function to > imitate this behavior instead of just returning 0. This partially solves > bug #201 (helps igowin.exe transparency problem, untested with another app).

1 0

Git Page on website
by Alistair Leslie-Hughes Oct. 24, 2007

Oct. 24, 2007

Hi, On the git web page, the commands to run to perform "Cleaning up your existing tree" and "Staying Up-To-Date" have been merge to one line. Best Regards Alistair Leslie-Hughes

1 0

AppDB: login
by speedator＠gmx.de Oct. 24, 2007

Oct. 24, 2007

Hi, would be great to have a longer period to stay logged in. A setting under preference for the period seems a good solution, Btw: What is the status of appdb/bugzilla(wiki?)-account-fusion? Bye, Julian

2 1

PulseAudio as a sound output?
by King InuYasha Oct. 24, 2007

Oct. 24, 2007

I would like to suggest that eventually Wine would support PulseAudio as a sound output natively. I am already aware of Wine supporting ESD, which PulseAudio can use, but supporting PulseAudio natively I think would be much better.

13 20

Anything wrong with my patch from 16.10. about w32x86 and win40 creation?
by Pat Oct. 24, 2007

Oct. 24, 2007

I usually use the archives for viewing patches, sometimes the patches are archived separately to the email. Markus' patch is here: http://www.winehq.org/pipermail/wine-patches/2007-October/045288.html

3 3

Re: iphlpapi: Intel Mac do not implement /proc filesystem so return Unimplemented for the functions trying to read from there.
by Juan Lang Oct. 24, 2007

Oct. 24, 2007

Hi Aric, @@ -104,6 +104,9 @@ DWORD getInterfaceStatsByName(const char { FILE *fp; +#ifdef __APPLE__ + return ERROR_NOT_SUPPORTED; +#endif This is ugly. Why not just change the case when fp is NULL? You've already got: fp = fopen("/proc/net/dev", "r"); if (fp) { ... } else ERR ("unimplemented!\n"); You could change this to set the return value, rather than using all these #ifdefs. Same thing throughout the patch. --Juan

2 1