wine-devel December 2003

wine-devel@winehq.org

109 participants
267 discussions

Re: [Resend] SHLWAPI fixes.
by Dmitry Timoshkov 19 Dec '03

19 Dec '03

"Flameeyes" <dgp85(a)users.sourceforge.net> wrote: > The attached patches fixes some functions of shlwapi I needed to run an > application I wrote with BCB that was using MS XML Parser. > I'm using it and works well AFAIK. > @@ -2346,6 +2376,9 @@ > } > StrCpyW(pszPointer, pszPath); > StrCpyW(pszUrl, pszNewUrl); > + > + LPWSTR slash; > + while((slash = StrChrW(pszUrl, '\\'))) *slash = '/'; > return S_OK; > } Your patch will not compile with older C compilers. You need to put variables on the top of the function body. Please resend. -- Dmitry.

1 0

DC ownerdrawn background mode
by Fabian Cenedese 18 Dec '03

18 Dec '03

Hi I have a control derived from CListCtrl. It's completely ownerdrawn. The problem comes when some item is selected. The whole column is drawn selected but the text is drawn with the actual back color instead of transparent. As this is white I get white text on white background... very readable :) I found that the difference between Windows and wine is the background mode. In Windows it's transparent whereas in wine it's opaque. I never change the background mode in my app, I guess I just lived from the fact that it's default transparent. Where is this set? Is this a wrong default parameter in wine or do I explicitly have to set the background mode in my app? I can do this, then it shows ok. But that's another change to my application I have to do because of wine. Maybe other apps may have similar problems. bye Fabi // from my code: void CMyCtrl::DrawItem(LPDRAWITEMSTRUCT lpDrawItemStruct) { // Prepare and test device context CDC* pDC = CDC::FromHandle(lpDrawItemStruct->hDC); // this is 1 for windows but 2 for wine // pDC->GetBkMode(); // wine fix pDC->SetBkMode(TRANSPARENT);

2 1

Fwd:SourceForge.net Sitewide update: Dec 9th, 2003
by Ivan Leo Murray-Smith 18 Dec '03

18 Dec '03

>Projects will need to ''opt-in'' in order to accept >donations. Dimi, are you going to opt in?

4 3

Fix for debugger breakage
by Mike Hearn 18 Dec '03

18 Dec '03

Hi, What is the status of this problem? I tried setting WINELOADER to point to wine-pthread, and while that allowed winedbg to load symbols it spewed loads of errors while doing so. Is there a proper fix for this? BTW I think the recent URL Cache patch broke the IE setup program, I get a crash there calling InterlockedExchange with a null pointer dereference. thanks -mike

3 6

winedbg question
by Chris Morgan 18 Dec '03

18 Dec '03

Should winedbg be able to load symbols up for non-wine libraries? After the export WINELOADER idea I've got wine symbols but I'm still trying to debug a bizzare crash in winejack where I get: ~/c$ wine sndrec32.exe Could not stat /mnt/fd0 (No such file or directory), ignoring drive A: Segmentation fault and when running under winedbg: *** Invalid address 0x40017f40 (_end+0x2ff03f40) Loaded debug information from ELF '/usr/local/lib/wine/ midimap.drv.so' (0x41090000) In 32-bit mode. Wine-dbg>cont Stopped on breakpoint 1 at 0x0100d32e (sndrec32.exe.EntryPoint in sndrec32.exe) Wine-dbg>cont Invalid address for breakpoint 1, disabling it Invalid address for breakpoint 2, disabling it Process of pid=0000000a has terminated Wine-dbg> >From single stepping I've found the function where the segfault appears to occur but its inside of jack_activate() in libjack.so.0 but symbols for this library don't appear to be loaded: *** Invalid address 0x40017f40 (_end+0x2ff03f40) No debug information in ELF '/usr/local/lib/libjack.so.0' (0x40f56000) No debug information in ELF '/lib/tls/librt.so.1' (0x40f62000) nm libjack.so.0 lists a bunch of symbols and I'm able to set a breakpoint on jack_activate() using gdb on a native linux app so it looks like debug information is built into libjack.so.0. Thanks, Chris

4 5

Problem with "exec shield" workaround
by Troy Rollo 18 Dec '03

18 Dec '03

The "exec shield" workaround (loader/main.c, libs/wine/port.c) has a side-effect of leaving a several thousand byte area of memory allocated just after the reserved pe_area. Under Windows, it is possible to fairly reliably map files or allocate memory explicitly using a fixed address of 0x10000000. The left-over memory from the loader causes programs that rely on this to fail. It seems that this workaround should either be a user setting or should mmap the memory area desired so that it can be removed without leaving behind a small fragment after that area.

1 0

default config file creation
by Chris Morgan 17 Dec '03

17 Dec '03

I was chatting with Rudolf Kastl(che on #winehq) and he was mentioning the things that he was doing for the rpm package. One of the patches he applies to wine prior to packaging it is a patch that creates local user config files if they don't exist. This helps end users as they don't have to worry about the package maintainer providing valid config files. I thought this sounded like a pretty neat idea to improve usability. The patch currently copies files from /etc/* which means they would need to be installed there during the process. I was wondering if it would make more sense during the compile process to take the default configuration files, config, system.reg, user.reg and userdef.reg and convert these to binary form and compile them into the wine binary itself. We could detect the lack of these files at startup and install them for the user. Another feature could be a command line switch that would rebuild a users config, like: 'wine --rebuild-config' that would reset the users config back to the defaults. Sound like something reasonable? Should we instead be putting the default files somewhere on the users machine and copying them over like the patch does? Chris

2 3

Re: default config file creation
by Ivan Leo Murray-Smith 17 Dec '03

17 Dec '03

> Trading off configurability with usability seems fine to > me. We aren't as one can tweak things with winecfg if he wants/needs to > The goal is to solve both > the lack of config files for binary installs, fixed by the > patches that Che has in his rpm package and Marcus has for > Suse, and for users that run 'make install' but never > create config files. All this can be solved by shipping a config file with with wine, get wine to adapt it to the users system, install it, install the registry and create the fake windows partition. This would work for all binaries on all distros, even on freebsd, and would work when installing from source. Only that someone has to implement it. And volunteers?

1 0

Re: default config file creation
by Ivan Leo Murray-Smith 17 Dec '03

17 Dec '03

>The less modification to the install process the better. >If there was a single location the files could be >installed to that would work for all distros I think this >is the route to head in. You can install the files anywhere, as long as the location isn't already used for something else, /etc/wine could do. Also, compiling the files means nobody could change the default config without recompiling wine, that sound short of weird. Ivan.

2 1

Re: default config file creation
by Ivan Leo Murray-Smith 17 Dec '03

17 Dec '03

> Read documentation/PACKAGING, I had it added to CVS with most ideas already > inside on Jun 27, 1999. I've read this. I also had a look at the recording of your speech about wine packaging at wineconf 2002. I've followed the indications of documentation/PACKAGING at lines 385-404, mainly for usability. > Feel free to copy the attached SuSE default make_fakewindows setup script I > use and the included minimalistic wine patch to loader/main.c. It all seems quite SuSE specific, anyway I haven't yet seen a script that works as well as winesetuptk, and there are complaints from time to time about helper scripts shiped with RPMs. Also a script doesn't give a user the freedom of choosing if he wants to use a real or a fake windows partition, and doesn't allow the user to chose where to have the fake windows partition, simply because it can't interact with the user. Nobody seems to be currently shipping winelauncher,and having a scipt to start wine isn't a great idea for a beta/stable build and ideally we want a long term solution. Bill Medland wrote: >I don't like the idea; I much prefer the idea of copying them from somewhere >in /etc, as redhat does. The default config file in /etc can then be >modified so that it matches the distribution (redhat have their cd at >/mnt/cdrom) and it can be modified by the sysadmin There is no need to add text files to the wine binary, but the default config should be modified on the fly. Some computers don't have a cd or a floppy, some have 2 cd drives, some don't use distro-standard paths, so drives should be detected on the fly, already we have enough logs from users full of /mnt/floppy not found and /mnt/cdrom not found, and it's not a great feature for 0.9 or 1.0. Winesetuptk does thit very very very well, I'm no engineer but I suppose that something written in tcltk can be easily ported to C. A binary package (But also a normal make install) should install configb and registry in /etc/wine or /usr/share/doc/wine-version/samples or something like that. The user should then start wine, that would look for $HOME/.wine, create it if necesarry, figure out the removable media paths, create a config file on the fly, write it to $HOME/.wine, run 'wine regedit path/to/winedefault.reg' and start normally.This does not require a script, will create a properly working config file, and the default configuration can be tweaked/restored with winecfg, once it's finished. Ivan.

1 0

← Newer
1
...
10
11
12
13
14
15
16
...
27
Older →

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 December 2003