- Wine-Devel - list.winehq.org

Re: EnableWindow
by Alexandre Julliard May 29, 2004

May 29, 2004

Ulrich Czekalla <ulrich.czekalla(a)utoronto.ca> writes: > ChangeLog: > Ulrich Czekalla <ulrich(a)codeweavers.com> > EnableWindow should not remove the focus of child windows Thanks for providing a test, but could you please merge it with the existing win.c and/or msg.c tests? We can't create a new file for every API being tested. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: Unsupported feature in SHFileOperationW
by Phillip.Ezolt＠hp.com May 28, 2004

May 28, 2004

Hi all, Here is patch for the SHFileOperationW problem. (Which I originally sent out in January. ) It isn't perfect, and it fixes the problem for my application. (which happens to be the same as Lionel's) According to Lionel the fix is better than what is there today, but it falls short of windows behavior in the following manner (for the record): > But it's not 100 % correct as, for example, it does not work if one needs to > create more than one level of directories (which is working fine in Windows, > I just checked it). Moreover, there should be a message box (if not disabled > like in RollerCoasterTycoon) asking if the users wants or not to create the > directory (I checked this on Windows too). --Phil Hewlett-Packard: High Performance Technical Computing/Visualization --------------------------------------------------------------------------- Phillip.Ezolt(a)hp.com Performance/Development ---------- Forwarded message ---------- Date: Fri, 30 Jan 2004 12:54:48 -0500 (EST) From: Phillip Ezolt <phillip.ezolt(a)hp.com> To: wine-patches(a)winehq.org Subject: SHFileOperationW: Create subdir on Copy Hi All, Here's a small fix. Comments/questions are welcomed. Summary ------- This patch allows SHFileOperationW to automatically create subdirectories if they don't exists. (The Microsoft documentation states that this is correct.) This patch fixes bug #1954. I have verified that it works with Today's CVS. (January 30,2004) I ran into this problem with the Rollercoaster Tycoon demo installer. It can successfully copy files to this newly created directory, but coping into any (currently non-existant) directories below it fail. According to MS: (http://msdn.microsoft.com/library/default.asp?url=/library/en-us/shellcc/pl…) " Copy and Move operations can specify destination directories that do not exist and the system will attempt to create them. " (See footnote #1 for more info) For example, this succeeds: trace:shell:SHNotifyCopyFileW (L"E:\\WZS21db.tmp\\English\\rct.exe" L"C:\\Program Files\\Hasbro Interactive\\RollerCoaster Tycoon Demo\\rct.exe" ) But this fails: trace:shell:SHNotifyCopyFileW (L"E:\\WZS21db.tmp\\Saved Games\\001" L"C:\\Program Files\\Hasbro Interactive\\RollerCoaster Tycoon Demo\\Saved Games\\001" ) The directory "Saved Games" should be created before "001" is copied into it. Instead, the call fails. --Phil Hewlett-Packard: High Performance Technical Computing/Visualization --------------------------------------------------------------------------- Phillip.Ezolt(a)hp.com Performance/Development

1 0

Re: wine/ windows/win.c server/user.c
by Uwe Bonnes May 28, 2004

May 28, 2004

>>>>> "Alexandre" == Alexandre Julliard <julliard(a)winehq.org> writes: Alexandre> ChangeSet ID: 12496 CVSROOT: /opt/cvs-commit Module name: Alexandre> wine Changes by: julliard(a)wine.codeweavers.com 2004/05/28 Alexandre> 14:35:37 Alexandre> Modified files: windows : win.c server : user.c Alexandre> Log message: Only allocate even-numbered window handles, MS Alexandre> Project depends on that (found by Dmitry Timoshkov). Alexandre> Patch: http://cvs.winehq.org/patch.py?id=12496 When Opera sued MS for having non compatible Webpages and won, we should consider the same for this stupid thing :-) -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

1 0

Re: Add documentation on the address space layout in Wine
by Mike Hearn May 28, 2004

May 28, 2004

On Fri, 28 May 2004 09:10:22 +0200, Pouech Eric DMI AEI CAEN wrote: > Hi Mike > I've started to revisit the architecture.sgml file, and started to write > also about this matters (and a few others). For coherency reasons, if > you don't mind, I'll reintegrate your bits in the architecture.sgml file > (what you wrote covers also process image loading...). Sure, go ahead and do what you want with it. Do you know when you'll have something to submit? Also what else have you written about? I'm keen on getting as much of this low level technical detail down in writing so developers new to Wine stand a chance of understanding the scarier bits of the code. I was thinking of tackling DCOM next. thanks -mike

2 1

Re: Unsupported feature in SHFileOperationW
by fenix＠club-internet.fr May 28, 2004

May 28, 2004

LOL I have seen (and sent a patch) a problem on FindExecutable who seems equivalent: trying to create a Link to an executable while executable not already present. Installers are funny :) Regards, Raphael ----Message d'origine---- >De: Lionel Ulmer <lionel.ulmer(a)free.fr> >A: wine-devel(a)winehq.org >Sujet: Unsupported feature in SHFileOperationW >Date: Fri, 28 May 2004 18:02:50 +0200 > >Hi all, > >A game demo installer uses SHFileOperationA to copy files over... And it >expects also the destination directories to be created when the copy is >done. > >Basically, it gives a list of fully qualified source files and their >corresponding fully qualified destination files and cann the >SHFileOperationA function with the 'FO_COPY' function and >'FOF_MULTIDESTFILES FOF_NOCONFIRMATION FOF_NOCONFIRMMKDIR' flags. > >This seems to not be handled properly in the current Wine tree as it fails >as soon as it needs to create the directory on the destination files. I >tested this on Windows and it works just fine (with nice MessageBoxes if one >ommits the 'NOCONFIRMMKDIR' flag :-) ). > >Is anyone (Martin, Ge ?) currently working on Shell32 who has an updated >tree with fixes in the Shell file operations ? > >If not, I will try to fix it on my own (be aware though that as I find the >style of coding used in this file particularely hard to decipher, I may do a >'NIH' on it and rewrite most of it to fit my tastes :-) ). > > Lionel > >-- > Lionel Ulmer - http://www.bbrox.org/ > >

1 0

msvcmaker - fix building makefiles on Windows
by Steven Edwards May 28, 2004

May 28, 2004

Hello, I am not 100% sure if this is correct as I dont have a Windows box handy and dont know perl but this has been a known bug where msvcmaker did not properly build Makefiles when using Perl on Windows. A anonymous slashdot reader sent me this patch to fix this oversite. He saif to either attribute the changelog entry to Fred Patrik or to me. If someone has a Windows box handy could you test msvcmaker with this patch and see if it generates the makefiles properly now? Thanks Steven Changelog: Fix generation of MS_VC Makfiles on Windows __________________________________ Do you Yahoo!? Friends. Fun. Try the all-new Yahoo! Messenger. http://messenger.yahoo.com/

1 0

Unsupported feature in SHFileOperationW
by Lionel Ulmer May 28, 2004

May 28, 2004

Hi all, A game demo installer uses SHFileOperationA to copy files over... And it expects also the destination directories to be created when the copy is done. Basically, it gives a list of fully qualified source files and their corresponding fully qualified destination files and cann the SHFileOperationA function with the 'FO_COPY' function and 'FOF_MULTIDESTFILES FOF_NOCONFIRMATION FOF_NOCONFIRMMKDIR' flags. This seems to not be handled properly in the current Wine tree as it fails as soon as it needs to create the directory on the destination files. I tested this on Windows and it works just fine (with nice MessageBoxes if one ommits the 'NOCONFIRMMKDIR' flag :-) ). Is anyone (Martin, Ge ?) currently working on Shell32 who has an updated tree with fixes in the Shell file operations ? If not, I will try to fix it on my own (be aware though that as I find the style of coding used in this file particularely hard to decipher, I may do a 'NIH' on it and rewrite most of it to fit my tastes :-) ). Lionel -- Lionel Ulmer - http://www.bbrox.org/

1 0

focused code review
by Glen Kaukola May 28, 2004

May 28, 2004

Hi, So one project suggested on the website is to do a focused code review. In particular it talks about getting rid of inefficient uses of strcat. But that leaves me wondering what exactly constitutes an inefficient use of strcat? If anyone could provide an example I'd appreciate it. Thanks. Glen

2 1

Re: Default to 96dpi for the no config file case
by Rolf Kalbermatter May 28, 2004

May 28, 2004

>If you mean via DDC then I guess this should work in most places, but for >whatever reason it does not work here :( In particular, xdpyinfo says: > >screen #0: > dimensions: 1024x768 pixels (321x241 millimeters) > resolution: 81x81 dots per inch > >So it appears to know the screen size (I measured with a ruler and >321x241 is right), but it gets the DPI too small anyway. At 81 dpi the >text is hardly readable. But: 1024/321*25.4 = 81.026 768/241*25.4 = 80.943 Seems to me X has it very right. That Windows usually doesn't care about real resolutions and instead fudges with it (consider the Large Font setting in the Display Control Panel) is probably another issue. Or the metrics this function requires are not dots per inch but something else. Rolf Kalbermatter

1 0

Re: Default to 96dpi for the no config file case
by Rein Klazes May 28, 2004

May 28, 2004

On Thu, 27 May 2004 22:18:20 +0100, you wrote: > wineprefixcreate doesn't create a config file. This means we use a (wrong) > formula to decide what the DPI is supposed to be. Why do you mean (wrong) ? Do you not get the dpi values used by X? Or is it that is uses the X values but those are wrong? > It's wrong because > nobody tells the system how big their monitor is these days so it always > picks a DPI that is too small. Here X gets it from the monitor at start-up (via the logic of the video card). I did not realize that my situation is not normal. Somehow, I think setting a proper dpi should ideally be done in X. But in practice ... I agree with you that 95 dpi is about the minimum that is being used today. But I would keep this the minimum: if the dpi is set bigger then it is probably intentional. Either monitor magic or from a careful system administration. Rein. -- Rein Klazes rklazes(a)xs4all.nl

4 4