April 2008 - Wine-Devel - list.winehq.org

Windows 7
by tony.wasserka＠freenet.de April 4, 2008

April 4, 2008

I read an interesting article on slashdot today: http://tech.slashdot.org/tech/08/04/04/1437258.shtml > Windows 7 will be a from-the-ground-up packaging of the Windows codebase; partially source, > but not binary compatible with previous versions of Windows." Sounds like they format, i.e. maybe they use sth. other than PE then. But even if they don't, I guess this could be a problem for the existing wine codebase and require some infrastructural changes to it. However, I don't know much about these things though and it's possible that I'm completely wrong, but I though I'd just post this to wine-devel in case anybody is interested. Feel free to discuss, or not to discuss, do whatever you want with this information ;-) Maybe we're lucky and people turn away from windows completely this time and we don't even need to care about Windows 7 xD (though this is unlikely to happen, but dreams never die...) Best regards, Tony Unbegrenzter Speicher, Top-Spamschutz, 120 SMS und eigene E-MailDomain inkl. http://office.freenet.de/dienste/emailoffice/produktuebersicht/power/mail/i…

1 0

xenroll exported functions
by andre furtado April 4, 2008

April 4, 2008

add documentation for exported functions of the xenroll library... IEnroll * WINAPI PIEnrollGetNoCOM(void)*/ returns a pointer to an IEnroll interface. */IEnroll2 * WINAPI PIEnroll2GetNoCOM(void)*/ returns a pointer to an IEnroll2 interface. */IEnroll4 * WINAPI PIEnroll4GetNoCOM(void)*/ returns a pointer to an IEnroll4 interface. */ André Renato Furtado talk2cod(a)hotmail.com _________________________________________________________________ Receba GRÁTIS as mensagens do Messenger no seu celular quando você estiver offline. Conheça o MSN Mobile! http://mobile.live.com/signup/signup2.aspx?lc=pt-br

1 0

Bug in MSVCRT_fseek (with patch)
by Kirill K. Smirnov April 4, 2008

April 4, 2008

Hi, > A successful call to fseek() shall clear the end-of-file indicator > for the stream and undo any effects of ungetc() and ungetwc() on > the same stream. After an fseek() call, the next operation on an > update stream may be either input or output. > } >+ /* Clear end of file */ >+ file->_flag &= ~MSVCRT__IOEOF; > return (_lseek(file->_file,offset,whence) == -1)?-1:0; What if you try to fseek to the end of file (SEEK_END)? Should IOEOF flag be cleared in this case? And may _lseek() function be better place for this? Current _lseeki64() implementation contains FIXME concerning the same issue: MSVCRT_fdesc[fd].wxflag &= ~(WX_ATEOF|WX_READEOF); /* FIXME: What if we seek _to_ EOF - is EOF set? */ -- Kirill

2 1

Re: [PATCH 4/4] configure.ac: Add support for MSVC.
by Alexandre Julliard April 4, 2008

April 4, 2008

Robert Shearman <rob(a)codeweavers.com> writes: > MSVC needs a lot more in the way of rule changes and/or helper scripts > for it to work properly, but this is a good start. I'm not sure I see the point. You can't really run configure or make on an MSVC setup anyway. It seems to me that something like msvcmaker is a better solution. -- Alexandre Julliard julliard(a)winehq.org

2 1

Bug in MSVCRT_fseek (with patch)
by Erik de Castro Lopo April 4, 2008

April 4, 2008

Hi all, I think I have found a bug in MSVCRT_fseek. If I fread() to the end of a file, then do fseek (file, 0, SEEK_SET), the feof() function still returns true. According to the ISO standard A successful call to fseek() shall clear the end-of-file indicator for the stream and undo any effects of ungetc() and ungetwc() on the same stream. After an fseek() call, the next operation on an update stream may be either input or output. The following patch to dlls/msvcrt/file.c fixes the issue. I also have a small standalone test program below that fails before the patch and passes after. Cheers, Erik The patch: -------------8<-------------8<-------------8<-------------8<------------- --- a/dlls/msvcrt/file.c +++ b/dlls/msvcrt/file.c @@ -918,6 +918,8 @@ int CDECL MSVCRT_fseek(MSVCRT_FILE* file, long offset, int whence) if(file->_flag & MSVCRT__IORW) { file->_flag &= ~(MSVCRT__IOREAD|MSVCRT__IOWRT); } + /* Clear end of file */ + file->_flag &= ~MSVCRT__IOEOF; return (_lseek(file->_file,offset,whence) == -1)?-1:0; } -------------8<-------------8<-------------8<-------------8<------------- The test program: -------------8<-------------8<-------------8<-------------8<------------- #include <stdio.h> #include <stdlib.h> #include <stdarg.h> static void error_exit (const char * fmt, ...) { va_list ap ; va_start (ap, fmt) ; vprintf (fmt, ap) ; va_end (ap) ; exit (1) ; } /* error_exit */ int main (void) { const char * filename = "test.dat" ; char buffer [12] ; FILE * file ; /* Create a short data file. */ if ((file = fopen (filename, "w")) == NULL) error_exit ("Error on line %d : fopen\n", __LINE__) ; fputs("abcdefghaiklmnopqrstuvwxyz\n", file) ; fclose (file) ; /* Now read the file. */ if ((file = fopen (filename, "r")) == NULL) error_exit ("Line %d : fopen\n", __LINE__) ; while (fread (buffer, sizeof (buffer), 1, file) > 0) (void) 0 ; if (! feof (file)) error_exit ("Error on line %d : feof should return true.\n", __LINE__) ; fseek (file, 0, SEEK_SET) ; if (feof (file)) error_exit ("Error on line %d : feof should return false.\n", __LINE__) ; fclose (file) ; remove (filename) ; return 0 ; } /* main */ -------------8<-------------8<-------------8<-------------8<------------- -- ----------------------------------------------------------------- Erik de Castro Lopo ----------------------------------------------------------------- "C++ has its place in the history of programming languages. Just as Caligula has his place in the history of the Roman Empire." -- Robert Firth

1 0

Re: [1/6] qedit: Add test framework and test for IMediaDet_put_Filename.
by Alexandre Julliard April 4, 2008

April 4, 2008

Dan Hipschman <dsh(a)linux.ucla.edu> writes: > This sets up some real tests for qedit using an actual AVI file to test > with. The resource is constructed from the AVI file by prepending the > size so we know how big it is in the code. That's ugly, you should use SizeofResource instead. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: Lug Radio Live 2008 wine talk. Slides online, comments welcome.
by pure_evil＠mail.bg April 4, 2008

April 4, 2008

Woah, that's ubercool. The presentation is excellent! However, you could add five collosal IMHO advantages of wine: 1. NO 1000 DISCONTINUED WIN VERSIONS That's what largely made me dump 'doze and go for Linux - got 6 W98 licenses, and at some point, upgraded the hardware to discover that 98 doesn't support the new Mainboards. So I spent two days trying to get all the MBs replaced, managed to get '98 running... And trust me, last thing you wanna do when you're in an empty new office is to extend a phone wire with a UTP wire to use the only connected atm phone socket to get dial-up working and with the collosal 5KBITS(!) per second to google out the only remaining in the PC store's warehouse mainboard to see if that supports '98, so you can call 'em and tell 'em "I'll take 'em". Oh, and you have to bring in your OWN PERSONAL machine to do that. The rest are basically a pile of mainboards which can't run '98 so they can't use the modem and so forth. And then enter my favorite software provider, sees the new machines, claps his hands and says: "Cool, new machines! Now you get the new windows version of the soft, and, it requires XP! And since you are the only ones to use the DOS version, it's discontinued from now on. Oh, and BTW? Your friendly local Samba server? Guess what, we're going MSSQL from now on, you need 1G of RAM on an XP machine... for 40 megs of databases." So with a thousand pains me gets the damn thing working on 98 and sacrifices TWO NEW MACHINES for the SQL server(s) and purchases TWO UNNECESSARY XP LICENSES - cuz two sets of databases, and, well, the express version doesn't support that. And things get better: "Yeah, you got it running under '98. Guess what, we'll be using cutting-edge stuff from now on; Now it's MSShitQL 2005, when something new comes out, you'll be upgrading; And once my soft becomes fully '98 incompatible, THEN you scrap the 98 licenses and by new, XP ones." "- But no money for licenses?" (1 XP = 1 new machine without monitor basically. Oh, and it has to be Professional) "- Bah, go pirate. Everybody's doing it." (What does he mean by on "cutting-edge stuff" BTW?! VISTA?! For an accounting database?!) And every once in a while when there's a bug in his soft, he attributes it to 98. Program closes but press ctrl-alt-del and you get a bunch of instances remaining? 98's fault. Slow? 98's fault. Crash due to an empty piece of database? "'98s fault". And of course, after 98s fault follows the natural "install XP". Enter wine. Select windows version = XP. Work. 2.) No Premium Basic Home Server Client whatever. Enter wine. Work. 3.) Easy reinstall. Delete "drive_c" and the reg files. "wineprefixcreate". Work. 4.) A 1000 windows on one machine. ~3 Megs per "drive_c". Work. 5.) Forget that "viruses under wine" stuff from days past, no, really. - No startup! Hide all you like in the registry you piece of cr... spyware, I won't even know you got there, nor will you ever get started! - No IE5678910! Yeah! Come and get me. - What, no normally inaccessible (even to NOD32 at some point) "System Volume Information"? Where are you gonna hide (or rather, stay exposed frustrating me)? - And where's the MBR of "drive_c"? in ".wine"? All that combined is what makes me want to develop for wine.

1 0

hypertrm exported functions
by andre furtado April 4, 2008

April 4, 2008

add documentation for exported functions of the win32 hypertrm library... BOOL InitInstance(const HINSTANCE hInstance, const LPTSTR lpCmdLine, const int nCmdShow)/* creates the hypertrm session window. args: hInstance - apps instance handle. lpCmdLine - command line. nCmdShow - passed from winmain overhead function arglist. */ int MessageLoop(void) /* window message loop. return a wparam int value, that contains the additional information of the processed message. */ SF_HANDLE sessQuerySysFileHdl(const HSESSION hSession) /* returns the sysfile handle based on hypertrm session. args: hSession -> hypertrm session. */ HTRANSLATE sessQueryTranslateHdl(const HSESSION hSession) /* returns the translate handle associated with specific hypertrm session. translate handle is used for manager character translation. args: hSession -> hypertrm session. */ int sfGetSessionItem(const SF_HANDLE hSF, const UINT uID, unsigned long *pIDSize, void *pvData) /* is used to get data and item size from the session file handle. args: hSF -> session file handle. uID -> item id. [out] pIDSize -> size returned. [out] pvData -> a pointer to an buffer that receives the data. */ int sfPutSessionItem(const SF_HANDLE hSF, const unsigned int uID, const unsigned long uIDSize, const void *pvData) /* is used to add or modify an entry in the session file data associated with the current session file handle. args: hSF -> session file handle. uID -> item id. pIDSize -> size. pvData -> a pointer to an buffer of data. */ talk2cod(a)hotmail.com _________________________________________________________________ Cansado de espaço para só 50 fotos? Conheça o Spaces, o site de relacionamentos com até 6,000 fotos! http://www.amigosdomessenger.com.br

3 3

Question about LoadImageW and gifs (affects all new Autodesk installers)
by Louis. Lenders April 4, 2008

April 4, 2008

Hi, does anyone know if LoadImageW is supposed to work with *.gif files? I'm asking because Autodesk installers seem to so something like LoadImageW(0,"c:\\foo.gif",IMAGE_BITMAP, 0, 0, LR_LOADFROMFILE) and then wine just crashes later on, as it tries to get the bitmapinfo header , and ends up with wrong info for size of bitmapinfoheader, width, height etc. Anyone an idea why the installer would do such weird thing? See also http://bugs.winehq.org/show_bug.cgi?id=11784 ___________________________________________________________ Yahoo! For Good helps you make a difference http://uk.promotions.yahoo.com/forgood/

1 0

re: [1/6] qedit: Add test framework and test for IMediaDet_put_Filename.
by Dan Kegel April 4, 2008

April 4, 2008

I think you can't depend on files existing in the source directory; you have to create them yourself (e.g. from a hex array in your source). Otherwise the unified winetest app that we use for semiautomated remote windows testing by users won't work. But if, say, WINETEST_INTERACTIVE is set, you can still use test.avi in the local directory instead of creating one.

2 2