June 2005 - Wine-Devel - list.winehq.org

Re: WININET: declare some missing functions
by Jonathan Wilson June 9, 2005

June 9, 2005

> +BOOLAPI FtpCommandA(HINTERNET, BOOL, DWORD, LPCSTR, DWORD_PTR, HINTERNET *); > +BOOLAPI FtpCommandW(HINTERNET, BOOL, DWORD, LPCWSTR, DWORD_PTR, HINTERNET *); Was it intentional that FtpCommand doesnt have a WINELIB_NAME_AW #define?

1 0

Google Summer of code - rc files translator proposition
by Piotr Caban June 9, 2005

June 9, 2005

Hello. I'm interested in Google Summer of code. My cousin, Jacek, has wrote here about it. I'm going to write a translator to rc files. I was thinking about how should it look like. In my opinion it has to have its own rc parser (I can integrate the wrc parser). I think this tool should have: - Listed every string that has to be translated - Preview of dialogs so that user can change its parameters characteristic for the language version (like size of controls) - RC files upgrading options - Nice UI with some useful options that makes translating easier Because rc files have very different syntax than po files and we need some specific options, I don't think using tools like poEdit is good choice. It won't make translations comfortable for users. This tool has to be as easy as possible, so users that have no coding experience can use it. Translating cannot expect user to know anything about syntax of rc files. Thanks, Piotrek

1 0

msvcmaker has stopped working
by Saulius Krasuckas June 9, 2005

June 9, 2005

Any quick ideas? $ tools/winapi/msvcmaker dlls/ntdll/ntdll.dsp (file 119 of 213)Can't open file './misc/ntdll_misc.dsp.tmp' at tools/winapi/util.pm line 114. $ mkdir misc $ tools/winapi/msvcmaker wine.dswUse of uninitialized value in pattern match (m//) at tools/winapi/msvcmaker line 1047. Use of uninitialized value in pattern match (m//) at tools/winapi/msvcmaker line 1047. Use of uninitialized value in pattern match (m//) at tools/winapi/msvcmaker line 1047. Use of uninitialized value in pattern match (m//) at tools/winapi/msvcmaker line 1057. Use of uninitialized value in pattern match (m//) at tools/winapi/msvcmaker line 1057. Use of uninitialized value in pattern match (m//) at tools/winapi/msvcmaker line 1057. Use of uninitialized value in concatenation (.) or string at tools/winapi/msvcmaker line 987. Use of uninitialized value in concatenation (.) or string at tools/winapi/msvcmaker line 994. $ rmdir misc rmdir: `misc': Directory not empty $ ll misc/ total 4 -rw-rw-r-- 1 s2 s2 2333 Jun 9 11:01 ntdll_misc.dsp $ rm -rf misc/ $ And yes, it works OK with "--no-wine": $ tools/winapi/msvcmaker --no-wine $ A bug?

2 1

Re: config check for videodev.h
by Juan Lang June 9, 2005

June 9, 2005

Replying to you both.. Adrian wrote: > The following patch fixes it for me, but I have little knowledge of > autoconf, so I'm not sure if I've done the right thing? That looks correct to me.. Dustin wrote: > I don't think that would be appropriate, as systems that have > sys/time.h may not always have sys/types.h .. Actually, during the configure check, HAVE_SYS_TYPES will not be defined. He's really only checking for the presence of <linux/videodev.h>, so on systems that don't have <sys/types.h>, <linux/videodev.h> won't exist either (it's always present on Linux.) So if the configure check fails for either reason, HAVE_LINUX_VIDEODEV won't be defined--and that's what he wants. --Juan __________________________________ Discover Yahoo! Have fun online with music videos, cool games, IM and more. Check it out! http://discover.yahoo.com/online.html

2 1

Re: ntdll: Implement and test RtlAnsiCharToUnicodeChar
by Alexandre Julliard June 9, 2005

June 9, 2005

James Hawkins <truiken(a)gmail.com> writes: > +/************************************************************************** > + * RtlAnsiCharToUnicodeChar (NTDLL.@) > + * > + * Converts an ansi character to a unicode character. > + * > + * PARAMS > + * ansi [I/O] Pointer to the ansi string. > + * > + * RETURNS > + * Success: Unicode representation of the first character in the ansi string. > + * Failure: NULL. > + * > + * NOTES > + * Upon successful completion, the char pointer ansi points to is > + * incremented by one. It should be incremented by two for multi-byte characters. -- Alexandre Julliard julliard(a)winehq.org

1 0

Problems in bitblt (in x11drv)
by David Albrecht June 9, 2005

June 9, 2005

Sorry about the resend, but I used an old copy of the source inadvertently. I'm new to wine and relatively new to windows programming being mostly a unix hack. Just for yucks I thought I'd get a relatively simple app emu48 to work. Turns out emu48 is not so simple (multi-threaded, etc.). First task towards functionality is to get the boot up display to look right and a BitBlt didn't seem to be functioning correctly. Much investigation later it turns out that in the file dib.c: /* if the source bitmap is 8bpp or less, we're supposed to use the * DC's palette for color conversion (not the DIB color table) */ is NOT true at least not for my case where the DC containing the bitmap has one palette and the colormap for the bitmap is not the same. If I disable the hPalette tests as I did below and always use the DIB colormap it boots up with the correct display otherwise the color mapping doesn't match the DIB bits and I get garbage. if (dib.dsBm.bmBitsPixel <= 8) { HPALETTE hPalette = GetCurrentObject( physDevSrc->hdc, OBJ_PAL ); if (/*!hPalette || (hPalette == GetStockObject(DEFAULT_PALETTE))*/1) { /* HACK: no palette has been set in the source DC, * use the DIB colormap instead - this is necessary in some * cases since we need to do depth conversion in some places * where real Windows can just copy data straight over */ colorMap = physBitmap->colorMap; nColorMap = physBitmap->nColorMap; } else { colorMap = X11DRV_DIB_BuildColorMap( physDevSrc, (WORD)-1, dib.dsBm.bmBitsPixel, (BITMAPINFO*)&dib.dsBmih, &nColorMap ); if (colorMap) aColorMap = TRUE; } } Not knowing the full ramifications of this code I thought I'd pass this along so that someone more familiar with it can reassess it. The DC and bitmap are created thusly (nLcdDoubled == 2): #define B 0x00000000 #define W 0x00FFFFFF #define I 0xFFFFFFFF static struct { BITMAPINFOHEADER Lcd_bmih; DWORD dwColor[64]; } bmiLcd = { {0x28,0/*x*/,0/*y*/,1,8,BI_RGB,0,0,0,64,0}, { W,B,B,B,B,B,B,B,B,B,B,B,B,B,B,B, B,B,B,B,B,B,B,B,B,B,B,B,B,B,B,B, I,I,I,I,I,I,I,I,I,I,I,I,I,I,I,I, I,I,I,I,I,I,I,I,I,I,I,I,I,I,I,I } }; VOID CreateLcdBitmap(VOID) { // create LCD bitmap _ASSERT(nLcdDoubled == 1 || nLcdDoubled == 2 || nLcdDoubled == 4); bmiLcd.Lcd_bmih.biWidth = LCD1_ROW * nLcdDoubled; bmiLcd.Lcd_bmih.biHeight = -64 * nLcdDoubled; hLcdDC = CreateCompatibleDC(hWindowDC); _ASSERT(hLcdDC != NULL); hLcdBitmap = CreateDIBSection(hLcdDC, (BITMAPINFO*)&bmiLcd,DIB_RGB_COLORS, (LPVOID*)&pbyLcd, NULL, 0); _ASSERT(hLcdBitmap != NULL); hOldLcdBitmap = SelectObject(hLcdDC, hLcdBitmap); _ASSERT(hPalette != NULL); SelectPalette(hLcdDC, hPalette, FALSE); // set palette for LCD DC RealizePalette(hLcdDC); // realize palette UpdateContrast(Chipset.contrast); } In the new source update contrast doesn't do a SetDIBColorTable it just plays with the pattern used to write to the dib. static DWORD Pattern[16]; VOID UpdateContrast(BYTE byContrast) { DWORD c = byContrast; DWORD b = byContrast + 0x20; if (bmiLcd.dwColor[b] == 0xFFFFFFFF) b = 0; _ASSERT(nLcdDoubled == 1 || nLcdDoubled == 2 || nLcdDoubled == 4); if (nLcdDoubled == 1) { WORD i,j; for (i=0; i<16; ++i) { Pattern[i] = 0; for (j=8; j>0; j>>=1) { Pattern[i] = (Pattern[i] << 8) | ((i&j) ? c : b); } } return; } c = (c<<8) | c; b = (b<<8) | b; if (nLcdDoubled == 2) { Pattern[0] = (b<<16)|b; Pattern[1] = (b<<16)|c; Pattern[2] = (c<<16)|b; Pattern[3] = (c<<16)|c; return; } c = (c<<16) | c; b = (b<<16) | b; if (nLcdDoubled == 4) { Pattern[0] = b; Pattern[1] = c; } return; } The BitBlt: static BYTE Buf[36]; VOID UpdateMainDisplay(VOID) { UINT x, y, nLines; DWORD d = Chipset.start1; BYTE *p = pbyLcd; #if defined DEBUG_DISPLAY { TCHAR buffer[256]; wsprintf(buffer,_T("%.5lx: Update Main Display\n"),Chipset.pc); OutputDebugString(buffer); } #endif _ASSERT(nLcdDoubled == 1 || nLcdDoubled == 2 || nLcdDoubled == 4); if (!Chipset.dispon) { nLines = 64; ZeroMemory(pbyLcd, LCD1_ROW * nLcdDoubled * nLines * nLcdDoubled); } else { nLines = LINES(Chipset.lcounter); // main display lines if (nLcdDoubled == 4) { for (y = 0; y < nLines; ++y) { Npeek(Buf,d,36); for (x=0; x<36; x++) { *(((DWORD*)p)++)=Pattern[Buf[x]&1]; *(((DWORD*)p)++)=Pattern[(Buf[x]>>1) & 1]; *(((DWORD*)p)++)=Pattern[(Buf[x]>>2) & 1]; *(((DWORD*)p)++)=Pattern[(Buf[x]>>3) & 1]; } CopyMemory(p, p-LCD3_ROW, LCD3_ROW); p+=LCD3_ROW; CopyMemory(p, p-LCD3_ROW*2, LCD3_ROW*2); p+=LCD3_ROW*2; d+=Chipset.width; } } if (nLcdDoubled == 2) { for (y = 0; y < nLines; ++y) { Npeek(Buf,d,36); for (x=0; x<36; x++) { *(((DWORD*)p)++)=Pattern[Buf[x]&3]; *(((DWORD*)p)++)=Pattern[Buf[x]>>2]; } CopyMemory(p, p-LCD2_ROW, LCD2_ROW); p+=LCD2_ROW; d+=Chipset.width; } } if (nLcdDoubled == 1) { for (y = 0; y < nLines; ++y) { Npeek(Buf,d,36); for (x=0; x<36; x++) *(((DWORD*)p)++)=Pattern[Buf[x]]; d+=Chipset.width; } } } EnterCriticalSection(&csGDILock); // solving NT GDI problems { BitBlt(hWindowDC, nLcdX, nLcdY, 131*nLcdDoubled, nLines*nLcdDoubled, hLcdDC, Chipset.boffset*nLcdDoubled, 0, SRCCOPY); GdiFlush(); } LeaveCriticalSection(&csGDILock); return; } As emu48 still doesn't work I'll probably have more things I find in likely much different areas later. Dave

2 2

Re: Add a couple of CloseHandle calls
by Alexandre Julliard June 9, 2005

June 9, 2005

Dmitry Timoshkov <dmitry(a)baikal.ru> writes: > (is this file supposed to have 2002 as a copyright year?) Yes, that's when I wrote it. It only took me 3 years to get around to making the code pass the tests ;-) -- Alexandre Julliard julliard(a)winehq.org

1 0

Alpha Status ?
by Tom Wickline June 9, 2005

June 9, 2005

Hello All, Can we change Wine from Alpha to Beta Status? Or is this scheduled for the 0.9 release ? Tom

1 0

Re: wcmd: Remove user32 dependencies
by Marcelo Duarte June 9, 2005

June 9, 2005

James Hawkins escreveu: >Hi, > >This patch is in responese to Bug #1982 'Can't SSH into a computer and >use "wine cmd" or wineconsole' > >http://bugs.winehq.org/show_bug.cgi?id=1982 > >The problem as identified by Marcus is that wcmd is importing user32 >which brings in the x11 driver causing the ssh'ed wcmd to fail: > >(~)> wine cmd >x11drv: Can't open display: > > The command "wine wcmd" or "wcmd" produces to me the same results and I do not see the errors described in bug1892. Only the keyboard input is bad, but the message above is not displayed to me. The command "wineconsole wcmd" functions. My terminal is 80x50 and it display on top 40x50. >I couldn't figure out any other way to get rid of LoadString. It's >pretty essential since we're loading multilingual help strings from >resource files. > >Changelog >* Dynamically call LoadString from user32. >* Remove all direct user32 dependencies. >* Don't import user32. > > > I dont think that this is necessary. This is not the cause of bug 1892. Maybe export DISPLAY

1 0

Re: Remove user32 dependencies
by Dmitry Timoshkov June 9, 2005

June 9, 2005

"James Hawkins" <truiken(a)gmail.com> wrote: > Changelog > * Dynamically call LoadString from user32. > * Remove all direct user32 dependencies. > * Don't import user32. > --- programs/wcmd/Makefile.in 9 Mar 2004 04:54:07 -0000 1.22 > +++ programs/wcmd/Makefile.in 9 Jun 2005 00:20:31 -0000 > @@ -4,7 +4,7 @@ SRCDIR = @srcdir@ > VPATH = @srcdir@ > MODULE = wcmd.exe > APPMODE = -mconsole > -IMPORTS = shell32 user32 kernel32 > +IMPORTS = shell32 kernel32 wcmd still indirectly depends on user32 due to the fact that shell32 imports user32, and that's unavoidable. -- Dmitry.

1 0