November 2003 - Wine-Devel - list.winehq.org

Re: winmm: move MCI settings from system.ini to the registry
by Eric Pouech Nov. 22, 2003

Nov. 22, 2003

Dimitrie O. Paun wrote: > Eric, > > I think these values (*.drv) sit under > HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\MCI > and not under > HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\MCI32 > as you suggested. On NT, MCI key is for 16 bit drivers and MCI32 key is for 32 bit drivers. The registry key for Win9x is not the same (it's HKML\System\CurrentControlSet\Control\MediaResources\mci). And most of the programs for Win9x still want to advertize their drivers in system.ini and not especially in the registry, so we'll have to provide some backup for that too. And to add the final touch, native drivers are 16bits (even on NT, at least 'til 2k, didn't check on XP), wine's are 32 bits. Wine current code should take of this (ie load a 16 bit driver when a 32 bit was requested, and do all the 16 <=> 32 transformation). So, what I'd suggest is to support all the schemes for configuring drivers, in this order of (decreasing) priority: registry-NT key (16+32), registry-Win9x key (16+32), system.ini (16+32). We also need to be consistent about this for all the drivers (and not only the MCI ones), which mean a deeper look at the code (video, acm...). I also think, since we're about to use the native registry settings, that we need to support the "Disabled" key in driver configuration, otherwise we might get lots of errors (for not correctly working devices). I won't have too much time to look at this in details right now, so if someone wants to jump in, I'll be happy to give him (her ?) a hand. A+ -- Eric Pouech

1 0

RFC: {Global,Local}{Re,}Alloc()
by Dimitrie O. Paun Nov. 22, 2003

Nov. 22, 2003

Hi folks, There are a few places in Wine where we're not using the Heap*() functions, but rather the old {Global,Local}{Re,}Alloc() ones. IIRC we need to do so in controls/edit.c for compatibility with old Win16 code, but why do we do so in places like: * print dialog * ole32 * shell32 * shlwapi Here is a list of places where {Global,Local}ReAlloc() are used. I'd like to replace them with Heap* functions where it is possible, but before I do, I need to know where we need the old calls. ./controls/edit.c: hloc32W_new = LocalReAlloc(es->hloc32W, alloc_size, LMEM_MOVEABLE | LMEM_ZEROINIT); ./controls/edit.c: if ((hNew32W = LocalReAlloc(es->hloc32W, alloc_size, LMEM_MOVEABLE | LMEM_ZEROINIT))) { ./controls/edit.c: hloc32A_new = LocalReAlloc(es->hloc32A, alloc_size, LMEM_MOVEABLE | LMEM_ZEROINIT); ./controls/edit.c: hloc16_new = LOCAL_ReAlloc(hInstance, es->hloc16, alloc_size, LMEM_MOVEABLE | LMEM_ZEROINIT); ./dlls/commdlg/printdlg.c: *hmem = GlobalReAlloc(*hmem, size, GMEM_MOVEABLE); ./dlls/commdlg/printdlg.c: *hmem = GlobalReAlloc(*hmem, size, GMEM_MOVEABLE); ./dlls/commdlg/printdlg.c: lppd->hDevMode = GlobalReAlloc(lppd->hDevMode, ./dlls/commdlg/printdlg.c: lppd->hDevMode = GlobalReAlloc(lppd->hDevMode, ./dlls/commdlg/printdlg16.c: *hmem = GlobalReAlloc16(*hmem, size, GMEM_MOVEABLE); ./dlls/commdlg/printdlg16.c: lppd->hDevMode = GlobalReAlloc16(lppd->hDevMode, ./dlls/kernel/local16.c: hseg = GlobalReAlloc16( hseg, 0x10000, GMEM_FIXED ); ./dlls/kernel/local16.c:HLOCAL16 LOCAL_ReAlloc( HANDLE16 ds, HLOCAL16 handle, WORD size, WORD flags ) ./dlls/kernel/local16.c:HLOCAL16 WINAPI LocalReAlloc16( HLOCAL16 handle, WORD size, UINT16 flags ) ./dlls/kernel/local16.c: return LOCAL_ReAlloc( CURRENT_DS, handle, size, flags ); ./dlls/kernel/local16.c:DWORD WINAPI Local32ReAlloc16( HANDLE heap, DWORD addr, INT16 type, ./dlls/kernel/resource16.c: handle = GlobalReAlloc16( hMemObj, pNameInfo->length << sizeShift, 0 ); ./dlls/kernel/resource16.c: handle = GlobalReAlloc16( hMemObj, pNameInfo->length << sizeShift, 0 ); ./dlls/ole32/hglobalstream.c: This->supportHandle = GlobalReAlloc(This->supportHandle, ./dlls/ole32/ifs.c: Malloc32.SpyedBlocks = (LPVOID*)LocalReAlloc((HLOCAL)Malloc32.SpyedBlocks, NewLength, GMEM_ZEROINIT); ./dlls/ole32/memlockbytes.c: This->supportHandle = GlobalReAlloc(This->supportHandle, ./dlls/ole32/memlockbytes16.c: This->supportHandle = GlobalReAlloc16(This->supportHandle, ./dlls/shell32/shell32_main.c: hargv=GlobalReAlloc(hargv, size, 0); ./dlls/shell32/shellole.c: addr = (LPVOID) LocalReAlloc((HANDLE) pv, cb, GMEM_ZEROINIT | GMEM_MOVEABLE); ./dlls/shlwapi/clist.c: lpTemp = (LPSHLWAPI_CLIST)LocalReAlloc((HLOCAL)pItem, ulSize, ./dlls/shlwapi/clist.c: lpIter = (LPSHLWAPI_CLIST)LocalReAlloc((HLOCAL)*lppList, ./dlls/shlwapi/clist.c: lpList = (LPSHLWAPI_CLIST)LocalReAlloc((HLOCAL)*lppList, ulNewSize, ./dlls/user/dde/misc.c: hData = GlobalReAlloc((HGLOBAL)hData, new_sz + sizeof(DDE_DATAHANDLE_HEAD), ./objects/gdiobj.c: if ((new_handle = (HGDIOBJ)(ULONG_PTR)LOCAL_ReAlloc( GDI_HeapSel, h, size, LMEM_MOVEABLE ))) ./objects/metafile.c: return (HMETAFILE16)GlobalReAlloc16( hMeta, 0, -- Dimi.

3 2

Re: wn20031121_197.xml
by Brian Vincent (C) Nov. 21, 2003

Nov. 21, 2003

> If I am not mistaken, I offered him commit access a while ago. He was > content at the time to have my slow butt commit. The offer still > stands. Yeah.. maybe it's time I take you up on it. Up until recently I didn't have reliable shell access. Now I've got a nice shell account and have reliable access from home. CVS access won't make my life any easier, however. I still have to mail Zack Brown the issues for his Kernel Traffic site. But adding a few steps shouldn't be a big deal. I think Eric always did the commits himself a few years ago. > I try to commit as soon as I see the XML, but when he sends it to > wine-patches at midnight on a Friday, well... it just may not get up > until Monday morning. Or, more likely, 2am on Saturday. I'm usually not drunk enough by midnight to finish a WWN issue ;) CVS access might be a good thing since my winter schedule is going to be changing quite a bit. The late Thursday/early Friday approach might be harder. One nice thing about posting these by midday on Friday is that we probably get a few more people reading them as opposed to publishing them on a Saturday. -brian

1 0

Re: wn20031121_197.xml
by Shachar Shemesh Nov. 21, 2003

Nov. 21, 2003

Vincent Béron wrote: >Le ven 21/11/2003 à 14:44, Shachar Shemesh a écrit : > > >>Wouldn't you say that it's time to give Vincent commit access to the web >>site? >> >> > >That'd be Brian, not me. > >Vincent > > > Luckily, it has recently rained, and so there is plenty of mud for my face. Sorry about that. On a totally different subject - wouldn't you say it was time to give Brian Vincent commit access to the web site? Shachar -- Shachar Shemesh Open Source integration & consulting Home page & resume - http://www.shemesh.biz/

2 1

Re: wine/ graphics/x11drv/xfont.c graphics/x11drv/ ...
by Dimitrie O. Paun Nov. 21, 2003

Nov. 21, 2003

On November 20, 2003 07:17 pm, Alexandre Julliard wrote: > Get rid of a bunch of TSX functions. Yay! Getting rid of the TSX functions seems to be within our reach now. I hope this is the plan, right? :) -- Dimi.

3 6

Re: wn20031121_197.xml
by Shachar Shemesh Nov. 21, 2003

Nov. 21, 2003

Wouldn't you say that it's time to give Vincent commit access to the web site? Brian Vincent wrote: >wn20031121_197.xml > >-brian > > >------------------------------------------------------------------------ > ><?xml version="1.0" ?> ><kc> > ><title>Wine Traffic</title> > > >

1 0

Agent 2.0: escape key problem.
by Rein Klazes Nov. 21, 2003

Nov. 21, 2003

Hi, Next problem with Agent 2.0 beta. The problem is when composing a message (email/usenet) and pressing the escape key. Expected is a dialog to confirm whether you want to save the message, not to save it or to cancel. Instead the dialog flashes and disappears again and nothing happens. Here is cut from the debug (relay,message,msg,key,hook) trace, leaving out most uninteresting lines. | trace:key:X11DRV_KeyEvent KeyPress : keysym=FF1B (Escape), # of chars=1 / 0x1b / '' | trace:key:EVENT_event_to_vkey e->keycode = 9 | trace:key:X11DRV_KeyEvent keycode 0x9 converted to vkey 0x1b | trace:key:X11DRV_KeyEvent bScan = 0x01. | 000b:Call user32.SendInput(00000001,406cf020,0000001c) ret=40d9cf6a | trace:key:queue_kbd_event wParam=001b, lParam=00010001, InputKeyState=81 | 000b:Ret user32.SendInput() retval=00000001 ret=40d9cf6a key press gets queued. | 000b:Call user32.GetMessageA(008f4478,00000000,00000000,00000000) ret=005e1c89 | trace:msg:MSG_peek_message got type 7 msg 100 (WM_KEYDOWN) hwnd 0x10064 wp 1b lp 10001 | trace:hook:HOOK_CallHooks calling hook 0x5fa990 WH_KEYBOARD code 0 wp 1b lp 10001 module L"" | 000b:Call hook proc 0x5fa990 (id=WH_KEYBOARD,code=0,wp=0000001b,lp=00010001) Getmessage picks up the keystroke and calls a hook. |000b:Call user32.MessageBoxA(00010026,41cefa3c "Save changes to message \"No Subject\"?",41cefa8c "Agent",00000023) ret=0066e78c The hook calls up the confirmation dialog. | trace:key:X11DRV_KeyEvent KeyRelease : keysym=0 (No Name), # of chars=0 / 0x00 / '' | trace:key:EVENT_event_to_vkey e->keycode = 9 | trace:key:X11DRV_KeyEvent keycode 0x9 converted to vkey 0x1b | trace:key:X11DRV_KeyEvent bScan = 0x01. | 000b:Call user32.SendInput(00000001,406ce7bc,0000001c) ret=40d9cf6a | trace:key:queue_kbd_event wParam=001b, lParam=c0010001, InputKeyState=1 | 000b:Ret user32.SendInput() retval=00000001 ret=40d9cf6a The keyup message is queued. | trace:msg:MSG_peek_message got type 7 msg 101 (WM_KEYUP) hwnd 0x1006f wp 1b lp c0010001 | trace:key:TranslateMessage (WM_KEYUP, 001B, C0010001) | trace:message:SPY_EnterMessage (0x1006f) L"&Ja" message [0101] WM_KEYUP dispatched wp=0000001b lp=c0010001 | 000b:Call window proc 0x40711b20 (hwnd=0x1006f,msg=WM_KEYUP,wp=0000001b,lp=c0010001) | 000b:Ret window proc 0x40711b20 (hwnd=0x1006f,msg=WM_KEYUP,wp=0000001b,lp=c0010001) retval=00000000 | trace:message:SPY_ExitMessage (0x1006f) L"&Ja" message [0101] WM_KEYUP returned 00000000 and is picked up and processed by the message box without much happening | trace:msg:MSG_peek_message got type 7 msg 100 (WM_KEYDOWN) hwnd 0x1006f wp 1b lp 10001 This is the riddle: the next message found is a escape-key down message. | 000b:Ret x11drv.DestroyWindow() retval=00000001 ret=4076df8b Of course it closes the message box. | 000b:Ret user32.MessageBoxA() retval=00000002 ret=0066e78c Message box finished | 000b:Ret hook proc 0x5fa990 (id=WH_KEYBOARD,code=0,wp=0000001b,lp=00010001) retval=00000001 hook returns | trace:msg:MSG_peek_message dropping msg 100 | trace:msg:MSG_peek_message got type 6 msg 0 (WM_NULL) hwnd 0x10026 wp 0 lp 0 | 000b:Ret user32.GetMessageA() retval=00000001 ret=005e1c89 and the Getmessage is finished. Suggestions? An unedited version of the log above is on http://www.xs4all.nl/~rklazes/temp/agkey.log.bz2 Rein. -- Rein Klazes rklazes(a)xs4all.nl

1 0

Re: removed problem with OpenGL extensions under X11
by Lionel Ulmer Nov. 21, 2003

Nov. 21, 2003

On Thu, Nov 20, 2003 at 11:50:13PM +0100, Daniel Skorka wrote: > This patch fixes a problem with bad OpenGL libraries wich uncorrectly > report the Extensions they support. Those libs would incorrectly report > not supporting glXGetProcAddresARB, on which wine relies for OpenGL > extension support. The problem was noticed with libGL from XFree86 4.2 > under Linux x86. Well, this poses the question : what should we do regarding broken libraries Wine depends upon ? Add 'incorrect' code in Wine (although in this case, it's not THAT incorrect :-) ) or report the bug to the library maintainers and wait for them to fix the issue ? Note that according to some DRI maintainer 'I had forgotten that the old libGL was that broken' :-) Lionel -- Lionel Ulmer - http://www.bbrox.org/

2 3

building wine docs
by Ivan Leo Murray-Smith Nov. 21, 2003

Nov. 21, 2003

Hi list, how can I build the wine docs in different little files (index.html, wine-stats.html and so on, like the docs for download on the web site) instead of one big file? make doc generates one big file, but this isn't what I want. Ivan.

3 2

[OT] HOWTO use unnamed KRun objects ?
by Subhobroto Sinha Nov. 21, 2003

Nov. 21, 2003

I thought a simple 'KRun(KURL("SomeBinaryFilename"))' should work, but it doesn't ? (I get ld telling me that there was an undefined reference to KRun::KRun(const KURL&...) !) Please tell me/redirect me(just make sure that the reference is working!) to an example where I can use KRun in such a manner, because KDE3API really does not come with sample code to help the uninitiated. I really just want to open files with their bound apps through my program, for ex: A KRun("Page.html") should open the 'Page.html' in Konqueror Thanks for letting me bug you ;) Regards Subhobroto Sinha __________________________________ Do you Yahoo!? Protect your identity with Yahoo! Mail AddressGuard http://antispam.yahoo.com/whatsnewfree

1 0