wine-devel June 2006

wine-devel@winehq.org

174 participants
352 discussions

Re: wine autorun utility
by William Knop 30 Jun '06

30 Jun '06

> On 6/30/06, William Knop <william.knop(a)gmail.com> wrote: > >> Parsing a windows inf hardly belongs anywhere but wine. > > Actually, Troy makes that point rather well in an earlier mail: > >> This is not true. The existing action-on-CD-insertion programs >> provided by the >> desktop environment try to detect the contents of the CD to see >> what they >> should do, so they will be looking for the autorun.inf file. >> Additionally the >> autorun.inf file format is designed to include specifications of >> different >> commands for multiple environments, so if autorun.inf files are to be >> respected at all it makes sense that they should also be able to >> start a >> native Linux executable or shell script (discovered from an >> [autorun.linux.i386] section, for example). >> >> There is nothing in this that requires or enhances the Win32 API >> facilities >> that Wine seeks to provide. Only once the native Windows >> executable has been >> identified as the only (or best) target for autorun would Wine become >> involved, when the program in the desktop environment invoked Wine >> to run the >> executable. Sorry, I missed this one (mailer digest mode). Hmm... You're saying the autorun.inf format is os-independent? If so, I was unaware, and I agree that the functionality belongs elsewhere. Will

1 0

Re: advapi32: Implement RegisterServiceCtrlHandleExW.
by Andrey Turkin 30 Jun '06

30 Jun '06

Robert Shearman wrote: > --- ... > @@ -603,10 +608,18 @@ static BOOL service_handle_control(HANDL > > TRACE("received control %ld\n", dwControl); > > - if (service_accepts_control(service, dwControl) && service->handler) > + if (service_accepts_control(service, dwControl)) > { > - service->handler(dwControl); > - ret = ERROR_SUCCESS; > + if (service->extended && service->handler.handler) > + { > + service->handler.handler(dwControl); > + ret = ERROR_SUCCESS; > + } > + else if (service->handler.handler_ex) > + { > + service->handler.handler_ex(dwControl, 0, NULL, service->context); > + ret = ERROR_SUCCESS; > + } > } Just wondering why you use Ex handler when service->extended==FALSE and vice versa.

2 1

Re: DDRAW: Releasing attached surfaces
by Stefan Dösinger 30 Jun '06

30 Jun '06

Hi, > That's weird, now it works without the "fix" and even Tiberian Sun > starts where previously it hung. > I must have wrongly assumed that the bug was related to surface > implementation code... > But i'm still curious what might be the cause. Will check that out later. > Sorry for bothering then... ;) It was most likely the fix for Tiberian Sun that fixed dune. The problem with ts was that the primary surfaces(front and back) were missing some flags. Maybe dune didn't like the surfaces it got and tried to get rid of them and hit a buggy code path in its code. Stefan

1 0

Re: wine autorun utility
by William Knop 30 Jun '06

30 Jun '06

> Gee sounds like a "great" idea. We all waiting too see some patches... > > It sure would be cool to have: > - Multiuser Wine > - Wine stable enough to run as service (err hmm whatever the hell > that means... > ah you mean daemon ?) > - Run something more complicated then 'printf("hello world\n");' > without X > - Talk to WMs to show icons and ask questions. > > Chris if you think that autostart is such a great idea - you are > very welcome to > start sending patches in. And if they are reasonable enough they > might get in. > But if you want to rant about that Linux doesn't have some > absolutely required > "feature" that windows has - this not the right place. > > Vitaliy Um hold on a second. Clearly many developers have different ideas about what's reasonable. It makes sense to obtain a semblance of unanimity and mutual understanding before taking action. You shouldn't go around quashing discussions like that. Will

4 7

Re: wine autorun utility
by Tim Schmidt 30 Jun '06

30 Jun '06

On 6/30/06, William Knop <william.knop(a)gmail.com> wrote: > Parsing a windows inf hardly belongs anywhere but wine. Actually, Troy makes that point rather well in an earlier mail: > This is not true. The existing action-on-CD-insertion programs provided by the > desktop environment try to detect the contents of the CD to see what they > should do, so they will be looking for the autorun.inf file. Additionally the > autorun.inf file format is designed to include specifications of different > commands for multiple environments, so if autorun.inf files are to be > respected at all it makes sense that they should also be able to start a > native Linux executable or shell script (discovered from an > [autorun.linux.i386] section, for example). > > There is nothing in this that requires or enhances the Win32 API facilities > that Wine seeks to provide. Only once the native Windows executable has been > identified as the only (or best) target for autorun would Wine become > involved, when the program in the desktop environment invoked Wine to run the >executable. --tim

1 0

Re: wine autorun utility
by Chris 30 Jun '06

30 Jun '06

On Thursday 29 June 2006 07:49, you wrote: > what is the gain of autorun, while i dont have wine running? would not > the common user expect, once she knows, that wine support autorun.inf, > that she can insert a cd and away it goes? while the explorer.exe is > not running permanently - e.g. via xinitrc - then i still would have to > deal with starting my setup.exe by hand. i have my doubts, that wine is > the correct place to implement this - but something like hotplugd or > whatevernowadaysdoesthetrickwithcdroms should handle this? One could think of Wine as a service, and start it with X if they wanted, so they could then run Windows programs whenever they wanted with potentially faster startup times, and with Windows "features" like autorun, and with a persistant Windows state (eg. instead of being shut down when all Windows programs close, it would stay around until Wine itself is explicitly stopped). Or, like me currently, they could run a Windows desktop/taskbar replacement in a Wine virtual desktop with managed windows off on a seperate X desktop and pretend to have Windows running along side X. I'd think if Wine were trying to "emulate" Windows as much as possible, it'd be backwards to not have autorun capabilities. As I said before, autorun.inf needs to be parsed anyway since it can define an icon to use for the drive, so if you're going to automatically parse it to load an icon, why not also ask (one time only) if the user wants to turn autorun on or off to run the defined program?

4 4

Re: [AppDb] update the comments in some files
by Chris Morgan 30 Jun '06

30 Jun '06

Why not replace the require_once() with requires instead of putting it on the todo? We don't really need the $aClean TODO item since thats high on the priority list at the moment and we'll just end up removing it in a day or two right? The move of show_note() into its class is on my todo list for class cleanups today or tomorrow as well. Basically I think we can do without the todo notes for now since most of it is in the process of being fixed. The rest of the comments look good though. Chris On Thursday 29 June 2006 9:43 am, Jonathan Ernst wrote: > You might need to apply the previous patches first. > > Changelog: > - update the comments in some files

2 1

wine autorun utility
by Segin Noname 30 Jun '06

30 Jun '06

(FYI: I have no internet connection at the moment, so I say 'hi' to all Wine developers!) Wine is nortorious for NOT supporting Autorun. Sure, it's stupid to load the autorun program right after the disk is inserted, but who wants to have to parse the autorun.inf by hand? So I wrote this little program. You can scan a CD for viruses first, then run this to emulate the Windows autorun feature. All you tell it is the drive letter of the CD, and it parses autorun.inf, and runs the listed program. It also check to make sure the file is a valid autorun.inf. It is not complete, and it is pure standard C. No Win32-specific functions are used. It uses several functions that aren't in the msvcrt, so I have included the portable snprintf library (version 2.1) and some portablized code from glibc. The tarball includes a prebuilt binary for your convience, but i strongly suggest you recompile the program from source (emerge xmingw on gentoo, and you can use my Makefile unmodified!)

11 17

Re: urlmon: Implement URLDownloadToCacheFile
by Troy Rollo 29 Jun '06

29 Jun '06

On Friday 30 June 2006 07:36, James Hawkins wrote: > + lstrcatW(cache_path, backslash); > + lstrcatW(cache_path, content); > + > + if (!GetTempFileNameW(cache_path, prefix, 0, cache_path)) > + return E_FAIL; > + > + if (lstrlenW(cache_path) > dwBufLength) > + return E_INVALIDARG; > + > + lstrcpyW(szFileName, cache_path); > + > + return URLDownloadToFileW(lpUnkCaller, szURL, szFileName, 0, pBSC); This is not the correct behaviour - doing this will result in the cache directories filling up with files that are not mentioned in the cache index. You should be calling CreateUrlCacheEntryW to get the file name for the new file. -- Troy Rollo - wine(a)troy.rollo.name

2 1

[AppDB] Viewing application famlies screen shots is broken.
by Tony Lambregts 29 Jun '06

29 Jun '06

Viewing application famlies screen shots is broken. I don't think is from todays patches. You can see the problem in the following link. http://appdb.winehq.org/screenshots.php?appId=631 The random screenshot shows up fine on the application page though. http://appdb.winehq.org/appview.php?appId=631 -- Tony Lambregts

2 1

Re: appdb/ ./appbrowse.php ./appsubmit.php ./appvi ...
by Tony Lambregts 29 Jun '06

29 Jun '06

WineHQ wrote: > ChangeSet ID: 26125 > CVSROOT: /opt/cvs-commit > Module name: appdb > Changes by: wineowner(a)winehq.org 2006/06/29 11:07:20 > > Modified files: > . : appbrowse.php appsubmit.php appview.php > admin : adminAppDataQueue.php adminAppQueue.php > adminMaintainers.php > include : application.php bugs.php comment.php > monitor.php note.php screenshot.php > sidebar_login.php url.php util.php version.php > vote.php > > Log message: > Chris Morgan <cmorgan(a)alum.wpi.edu> > Clean up application and version classes. Move class related functions into the class as static member functions > > Patch: http://cvs.winehq.org/patch.py?id=26125 > > Old revision New revision Changes Path > 1.13 1.14 +1 -1 appdb/appbrowse.php > 1.43 1.44 +2 -2 appdb/appsubmit.php > 1.82 1.83 +1 -1 appdb/appview.php > 1.22 1.23 +6 -6 appdb/admin/adminAppDataQueue.php > 1.57 1.58 +4 -4 appdb/admin/adminAppQueue.php > 1.19 1.20 +3 -3 appdb/admin/adminMaintainers.php > 1.50 1.51 +15 -53 appdb/include/application.php > 1.10 1.11 +7 -5 appdb/include/bugs.php > 1.21 1.22 +2 -2 appdb/include/comment.php > 1.5 1.6 +10 -4 appdb/include/monitor.php > 1.10 1.11 +4 -3 appdb/include/note.php > 1.39 1.40 +7 -5 appdb/include/screenshot.php > 1.13 1.14 +2 -2 appdb/include/sidebar_login.php > 1.6 1.7 +7 -5 appdb/include/url.php > 1.61 1.62 +18 -2 appdb/include/util.php > 1.58 1.59 +47 -38 appdb/include/version.php > 1.17 1.18 +2 -2 appdb/include/vote.php > This patch breaks the monitor notification emails. -- Tony Lambregs

2 4

Re: [AppDb] prefix all GPC variables
by Chris Morgan 29 Jun '06

29 Jun '06

Patch looks good and although it is large like you said it isn't something we can do much about at this point. It unfortunately doesn't apply cleanly against cvs head. If you could submit all of the smaller cleanup patches ahead of big patches like this that would be the best. Its more likely that larger patches like this will require changes than simple renames or refactoring and getting those out of the way means neither of us have to track which patches have gone in vs. need to be rediffed against cvs head. I'll usually make a single change in a single checked out appdb directory. That way I'm always diffing against cvs head and I can avoid having patches that depend on each other. I'm concerned now if you've made further changes to the directory where this patch was generated that some fixes will be missed or unwanted changes will sneak in. Lets get all of the rest of the isolated cleanups that you've submitted into the appdb and then lets give this patch another shot. Chris On Thursday 29 June 2006 4:59 am, Jonathan Ernst wrote: > Before you ask, there was no possibility to split this one up because > everything is interdependant. > > Changelog: > - rename all Get/Post/Cookies variables using type prefixes so that we can > later use the autofiltering function

1 0

Re: [AppDb] factor exit calls after redirect
by Chris Morgan 29 Jun '06

29 Jun '06

Patch doesn't apply against CVS head. Chris On Thursday 29 June 2006 7:05 am, Jonathan Ernst wrote: > You might need to apply the previous patches first. > > Changelog: > - factor exit calls after redirect

1 0

Re: [Tools/Wine.inf] Set Internet Explorer version
by Steven Edwards 29 Jun '06

29 Jun '06

Hi, On 6/28/06, Maarten Lankhorst <M.B.Lankhorst(a)gmail.com> wrote: > > --- tools/wine.inf.ori 2006-06-28 18:32:28.000000000 +0200 > +++ tools/wine.inf 2006-06-28 18:35:22.000000000 +0200 > @@ -2467,3 +2467,6 @@ > HKLM,%Control%\ProductOptions,"ProductType",2,"WinNT" > HKLM,%Control%\Windows,"CSDVersion",0x10003,0x400 > HKLM,%Control%\Session Manager\Environment,"OS",2,"Windows_NT" > +HKLM,"Software\Microsoft\Internet Explorer","Version",,"6.0.2900.2180" > +HKLM,"Software\Microsoft\Internet Explorer","W2kVersion",,"6.0.2900.2180" > +HKLM,"Software\Microsoft\Internet Exlporer","Build",,"62900" Does installing IE6 in win98 mode still work after this? I think its going to have problems. We really should report IE 5 so the IE 5.5 and 6 installers dont get confused. -- Steven Edwards "There is one thing stronger than all the armies in the world, and that is an idea whose time has come." - Victor Hugo

4 4

Re: [AppDb] util_show_error_page
by Chris Morgan 29 Jun '06

29 Jun '06

We'll want to perform this move of error_page() and the renaming of calls to it in a single patch. Chris On Wednesday 28 June 2006 1:21 pm, Jonathan Ernst wrote: > Changelog: > - new util_show_error_page that will replace errorpage

2 3

Re: msiexec: msiexec also supports options using a hyphen
by Alexandre Julliard 29 Jun '06

29 Jun '06

"James Hawkins" <truiken(a)gmail.com> writes: > Changelog: > * msiexec also supports options using a hyphen. Surely there's a better way to do this than duplicating all the string compares... -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: DDraw: Return DDERR_NOPALETTEATTCHED if there is no palette
by Stefan Dösinger 29 Jun '06

29 Jun '06

Am Donnerstag 29 Juni 2006 11:40 schrieb Stefan Dösinger: > According to the msdn this should happen, and native d3drm.dll depends on > this > > ChangeLog: > Stefan Dösinger: Return DDERR_NOPALETTEATTCHED if there is no palette Do not apply this patch, it is not completely correct. It breaks palettized textures in the dx7 sdk. I'll write a test case for this

1 0

winegcc atof() bug?
by Benjamin Fabricius 29 Jun '06

29 Jun '06

Hi, Im experiencing weird things when using atof() in certain places of my software and compiling with winegcc (wine20050725). It works fine when calling it out of the main function but as soon as i migrated some routines to another c file things became really strange. atoi works fine but as soon as i use atof with any "x.x" string i keep on getting some negative constant value as it seems. thought this might be HUGE_VAL of strtod but atof doesnt return any errors.. Anybody experience similar problems. unfortunately i am dependent on this wine version.. so changing wine versions won't help since current builds interfere with other parts of the software running in wine. Heres a really simple example code of what im trying to do: --- snip --- void print_set_parameters(char **setparams, int startofparams, int endofparams) { int param_num; double param_value; /* this works fine setparams has got the right * string values */ printf("test %d\n", startofparams); printf("test %s\n", setparams[3]); printf("test %s\n", setparams[4]); for(i = startofparams; i < endofparams; i++) { // works fine! param_num = atoi(setparams[i]); printf("string is %s, param number %d\n", setparams[i],param_num); i++; // heres where the trouble starts param_value = atof(setparams[i]); // tried this as an alternative same problem // param_value = strtod(".1", 0); // no errors reported // perror(__FUNCTION__); printf("param value %f\n", param_value); } } --- snip --- Any help would be great!! Cheers, ..Ben -- ------------------- Benjamin Fabricius German Student Lawo AG

1 0

msdn calling everything .net now?
by Dan Kegel 29 Jun '06

29 Jun '06

Lately, when I search for win32 API functions like CoGetObject, the top few hits I get at microsoft.com are the .net ones. I guess it's not a problem, but it feels weird.

2 1

automatic winetest builds stopped
by Ivan Leo Puoti 29 Jun '06

29 Jun '06

Does someone know why the tests at test.winehq.com/data aren't being built any more? Ivan.

3 3

Battlefront blending problem is fixed.
by Jesse Allen 29 Jun '06

29 Jun '06

Hi Henri, You latest round of wined3d patches fixed the blending problem in SW Battlefront including blocky text. Now I can properly aim at targets. :) However, some screen items seem to be a little more transparent than they should be. It doesn't fix the blocky text in BF1942. BF1942 is a d3d8 game and Battlefront is a d3d9 game. Thanks, Jesse

2 1

Freetype - Compile error when making a regression test
by Augusto Arcoverde da Rocha 28 Jun '06

28 Jun '06

Hi all, I'm here again with a little question about my dispersed efforts over Wine development. Now I'm trying to make some regression tests to identify when Lotus Notes started to show blank squares rather than icons for attachments. And by this way I get some errors compiling (see bellow) Wine versions 0.9.1 and 0.9.4, but with the version 0.9.9 this errors don't occurs. My guess is some version problem with freetype, but I didn't find any hints about this on the list history, though I have found some posts asking about the same problem. I tried to compile Wine on a Debian r3.1 with gcc versions 3.3 and 4.1, and with freetype 2.2.1, but didn't get any success. Any help is appreciated. Regards, --- 8< --- freetype.c:106: error: 'FT_Get_Module' undeclared here (not in a function) freetype.c:106: warning: type defaults to 'int' in declaration of 'pFT_Get_Module' freetype.c: In function 'WineEngGetFontData': freetype.c:3698: error: 'TT_Face' undeclared (first use in this function) freetype.c:3698: error: (Each undeclared identifier is reported only once freetype.c:3698: error: for each function it appears in.) freetype.c:3698: error: syntax error before 'tt_face' freetype.c:3699: error: 'SFNT_Interface' undeclared (first use in this function) freetype.c:3699: error: 'sfnt' undeclared (first use in this function) freetype.c:3703: error: syntax error before ')' token freetype.c:3703: error: 'tt_face' undeclared (first use in this function) freetype.c:3708: error: syntax error before ')' token freetype.c: In function 'is_hinting_enabled': freetype.c:3776: error: called object 'pFT_Get_Module' is not a function freetype.c:3777: warning: implicit declaration of function 'FT_DRIVER_HAS_HINTER' --- 8< ---

1 1

Wide Strings Query
by Andrew Talbot 28 Jun '06

28 Jun '06

I have written a patch to fix some of the write-strings warnings in regedit's listview.c file, in its OnGetDispInfo() fuction, but I need to know that my understanding is right, before I submit it. Am I correct in thinking that I can replace the _T() macros with TCHAR * variables, as exemplified with reg_szT, below? @@ -252,6 +252,19 @@ static void OnGetDispInfo(NMLVDISPINFO* plvdi) { static TCHAR buffer[200]; + static TCHAR reg_szT[] = {'R','E','G','_','S','Z',0}, plvdi->item.pszText = NULL; plvdi->item.cchTextMax = 0; @@ -263,43 +276,43 @@ case 1: switch (((LINE_INFO*)plvdi->item.lParam)->dwValType) { case REG_SZ: - plvdi->item.pszText = _T("REG_SZ"); + plvdi->item.pszText = reg_szT; break; Thanks, -- Andy.

1 0

re: msi: Download the MSI package if it is a remote URL
by Dan Kegel 28 Jun '06

28 Jun '06

Your patch ends in an empty sig, i.e. two dashes, which makes patch complain "patch ends in middle of a line" after finishing. I tried your patch to see if it fixed a similar problem in Ability Office (see http://ability.com/v4/download.php ) but sadly that still fails with fixme:urlmon:URLDownloadToCacheFileW ((nil) L"http://www.abilitydownload.com/download4/ability4_15/Ability.MSI" 0x7fd49f40 520 0 0x7fb9fae8) -- Wine for Windows ISVs: http://kegel.com/wine/isv

2 2

Re: WineD3D: Add one extra line to the dib section
by Saulius Krasuckas 28 Jun '06

28 Jun '06

* On Fri, 23 Jun 2006, Stefan [iso-8859-1] D�singer wrote: > Some DirectDraw apps access the surface memory via DWORDS and do not > take the necessary care at the end of the surface. The old DDraw code > allocated 4 extra bytes for the surface and used the unofficial > DIB_CreateDibSection function to create the DIB section. The normal > CreateDIBSection function doesn't use the requested size. To increase > the surface size to catch bad accesses one extra line is added to the > dib section. It's nice to hear you are working on this... > This patch addresses the regression in Diablo 1 and some odd crashes in > the DX7 sdk demos ...but the patch doesn't fix the Diablo 1 game on my box. I even made distclean here to no avail.

1 0

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

wine-devel June 2006