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

Re: PATCH: Inserting an item in a listview sometimes generates incorrect messages
by Dimitrie O. Paun March 6, 2003

March 6, 2003

On March 6, 2003 06:04 am, Adam Gundy wrote: [wine-devel Cc:ed to keep people in the loop] > Is a listview guaranteed to be single threaded? I assume it is because it > is all message driven - so some flag in the LISTVIEW_INFO structure could > be set. I'd say that is even more opaque. I think this would be better. It's localized in two points (where we set it, and where we use it), and it's easier to follow. The rest of the code should be unaware of this problem, so hiding it is the right thing to do IMO. The reason it's easier to follow is that if I don't see it, I don't care about it, which is what we want, and 99% of people looking at the code will fall in this category. If you do happen to stumble upon it, you can search for the infoPtr->blah thing that we're setting to see where and how this thing is used. You can't easily search with the current flag solution. > Actually, there is the rather unhelpful comment "But we are not supposed to > modify nHotItem!* in LISTVIEW_ShiftIndices() - does 'real' Windows maintain > the focus after an insertion? Perhaps that's why they don't send the > messages - you know how they work: write the code then document its broken > behaviour after the fact ;-) Yes, I know. This is why I think it should just result from some other behavior naturally, I don't think MS went through the trouble of passing a boolean around all over the place just to not send this notification. One more reason to not uglify our interfaces, as it will just get fixed automatically when we properly support the HotItem. -- Dimi.

3 2

Re: Janitorial dlls/advapi32/registry.c W->A cleanup
by Rolf Kalbermatter March 6, 2003

March 6, 2003

> Dmitry Timoshkov wrote: > >> "Tony Lambregts" <tony_lambregts(a)telusplanet.net> wrote: >> Yeah I actually had it that way at one point due to my familairity >> with drive.c and well... I ran into some compiler errors (in >> RegSaveKeyW) that ended up making the problem worse. Perhaps it was >> something I simple I missed. I can take another look at it if it will >> save some CPU cycles. > >when I change the code to the following > LONG WINAPI RegSaveKeyA( HKEY hkey, LPCSTR file, LPSECURITY_ATTRIBUTES sa ) { WCHAR fileW[MAX_PATH]; - LONG ret, len; + LONG ret; - len = WideCharToMultiByte(CP_ACP, 0, fileW, -1, file, 0, NULL, NULL); - WideCharToMultiByte(CP_ACP, 0, fileW, -1, file, len, NULL, NULL); + MultiByteToWideChar(CP_ACP, 0, file, -1, fileW, MAX_PATH); - ret = RegSaveKeyW( hkey, fileW.Buffer, sa ); + ret = RegSaveKeyW( hkey, fileW, sa ); return ret; } >I get the following warning > >warning: passing arg 5 of `WideCharToMultiByte' discards qualifiers from >pointer target type Well you should get a lot more warnings and even errors IMO. Look above as to what I think the code should really look like. Rolf Kalbermatter

1 0

Re: New conformance test for user32.dll
by Ferenc Wagner March 6, 2003

March 6, 2003

Tony Lambregts <tony_lambregts(a)telusplanet.net> writes: > The patch is still in the wrong format. New files need to > be diff'ed against /dev/null. Thanks for the clarification! Winehq says new files can be included as separate attachments. Well, this is exactly one kind of problems/comments I expected for first contribution. Also, I found no mention of license info on the web page. I Cc-ed the devel list so that someone can fix the docs. Thanks for all: Feri.

6 6

Re: (2nd) Janitorial dlls/advapi32/registry.c W->A cleanup
by Vincent Béron March 6, 2003

March 6, 2003

Tony Lambregts a écrit: > With Dmitry's suggested change and Vincent's advice (doh!) Second advice: don't forget the patch :)

1 0

Winelib article in this month's C/C++ User's Journal
by Matthew Bloch March 6, 2003

March 6, 2003

http://www.cuj.com/current/feature.html?topic=current An immodest plug for my introductory Winelib article that the CUJ editor Joe Casad was soliciting through thist list back in December. It's now in the print journal and on their web site at the (temporary, by the looks of things) link above. Thanks to the people who offered criticism before publication; hopefully in the three months since I wrote it the information isn't too out of date! cheers, -- Matthew Bloch Bytemark Hosting tel. +44 (0) 8707 455026 http://www.bytemark-hosting.co.uk/ Dedicated Linux hosts from 15ukp ($26) per month

4 3

Re: Add messaging for SB_CTL bars to scroll.c
by Dimitrie O. Paun March 6, 2003

March 6, 2003

On March 5, 2003 09:22 pm, Andrew Johnston wrote: > +/* Determine if the info is valid */ > +#define SCROLL_INFO_INVALID(info) \ > + ((info->fMask & ~(SIF_ALL | SIF_DISABLENOSCROLL)) || \ > + ((info->cbSize != sizeof(*info)) && \ > + (info->cbSize != sizeof(*info) - sizeof(info->nTrackPos)))) Please use inline functions for this sort of thing. Also, it's typically better to ask the positive question: inline static BOOL SCROLL_IsScrollBarInfoValid(SCROLLBAR_INFO *infoPtr) { /* ... */ } -- Dimi.

2 1

Re: PATCH: Inserting an item in a listview sometimes generates incorrect messages
by Dimitrie O. Paun March 6, 2003

March 6, 2003

On March 5, 2003 09:28 am, Adam Gundy wrote: > When shifting the list's indices, we must not send notify messages > if we have to refocus. Item modification handlers are not supposed > to be called here. I'm not too hot about passing that boolean around... I'd rather have a special way to avoid the notification in the case of shifting items, rather than clutter the interface that's used all over the place to pass a strange TRUE that's rather opaque. I've tried hard to not extend the interface beyond what's documented for the messages. It makes the code simpler to follow. -- Dimi.

1 0

Delphi with Wine
by S.Sistemas Informação March 6, 2003

March 6, 2003

Hi, My environment is three layers. On second layer (Applications Server), I'm not getting to run the TSimpleObjectBroker and TSocketConnection Delphi components in my application with wine. The client BD and BDE application works perfectly. Somebody can help me? Thanks. Viação Ouro Branco S/A Sistemas de Informação e Tecnologia Grimaldo Ghioti ssi(a)viacaoourobranco.com.br Fone(43)3373-2630 Fax(43)3373-2615 Brazil ______________________________________ Inflex - installed on mailserver for domain @viacaoourobranco.com.br Queries to: postmaster(a)viacaoourobranco.com.br

3 2

Re: New conformance test for user32.dll
by Duane Clark March 5, 2003

March 5, 2003

Tony Lambregts wrote: > > Ok. then > > Change Log: Clarify patch requirements. > > Files changed: documentation/patches.sgml > > Index: patches.sgml > =================================================================== > RCS file: /home/wine/wine/documentation/patches.sgml,v > retrieving revision 1.6 > diff -u -r1.6 patches.sgml > --- patches.sgml 18 Feb 2003 23:23:35 -0000 1.6 > +++ patches.sgml 5 Mar 2003 21:01:34 -0000 > @@ -48,8 +48,8 @@ > </para> > <para> > For additions: mention that you have some new files and > - include them as either separate attachments or by appending > - the <command>diff -u /dev/null /my/new/file</command> output of > them ^^^^ Which demonstrates one of the complications with inlined plain text, because generally extra effort is required to prevent undesired word wrap.

1 1

Re: winedbg b0rken on FreeBSD?
by Eric Pouech March 5, 2003

March 5, 2003

> Is there anything I can do? hard to tell from here... from the things you could do is try to do is: 1/ winedbg diablo 2/ then when winedbg seems to hang, try to attach gdb to it (to the winedbg process, not the diablo's as you did). it may tell us where it fails another option is to run with --debugmsg +server,+relay but you will get lots of traces... but it may show where things stop and the third one is to run gdb thru winedbg (run winedbg -- --gdb diablo, and see what gives) A+ -- Eric Pouech

1 0

Re: Add wineconsole FIXME
by Eric Pouech March 5, 2003

March 5, 2003

> wineconsole has problems with SetConsoleTitleA() when running with > CURSES backend. > > Document that by adding a FIXME. I think we should add it to wineconsole documentation too... once it gets written :-( > Any solution to that problem? not really. do you know what your app does after getting the window handle ? in some cases, it's used to either: - check if the app is running - or send messages (minimize/maximize, show/hide, quit) I don't see very good solutions here... curses backend is intended to be run without USER loaded... so that would break this rule... even, if we create a dummy window in wineconsole, without requiring USER to be loaded (using bare server requests), I'm not sure it'll cover all aspects of the issues (esp. what the app does once it gets the handle) BTW, XP introduced a new API to implement in a sane way this window lookup 'GetConsoleWindow' (for the console attached to the current process) (in the same type of issues, GetWindowThreadProcessId() will return bogus results) A+ -- Eric Pouech

1 0

Resources and more with Darwin
by Pierre d'Herbemont March 5, 2003

March 5, 2003

Hi all! I am trying to build wine onto Max OS X/Darwin. I am getting trouble with windres and the *.res files. I would like to know if it would be possible to have winelib running without those resources, I mean, could I build a program and link it fine without those resources? In a second time I would like to know if you plan to use wrc as the only (without windres) resource manager. I have read in the list something like that, could you give me the confirmation? Also I would like to know in which measure is the Elf file format implicated in wine (in opposition to darwin's mach-o). Thanks, Pierre

5 7

Re: Clean-up of scroll.c (1/4)
by Alexandre Julliard March 5, 2003

March 5, 2003

"Andrew M. Johnston" <johnstonam(a)logica.com> writes: > - return (info->fMask & SIF_ALL) != 0; > + > + return (info->fMask & SIF_ALL); [...] > - if (lpMin) *lpMin = infoPtr->minVal; > - if (lpMax) *lpMax = infoPtr->maxVal; > - return TRUE; > + TRACE("hwnd=%p nBar=%d lpMin=%p lpMax=%p\n", hwnd, nBar, lpMin, lpMax); > + > + if (lpMin) *lpMin = infoPtr ? infoPtr->minVal : 0; > + if (lpMax) *lpMax = infoPtr ? infoPtr->maxVal : 0; > + > + return (BOOL)infoPtr; Please don't do this kind of gratuitous changes. Boolean functions should return TRUE or FALSE, unless you have evidence that Windows does it differently. Yes, I know any non-zero value is true in C, but you can't rely on every Windows programmer to get it right. -- Alexandre Julliard julliard(a)winehq.com

3 6

Re: SystemParametersInfo patch
by Duane Clark March 5, 2003

March 5, 2003

Jon Bright wrote: > ... > (as an aside are patches preferred as MIME attachments or straight in > the mail here?) > I will just throw in a couple more comments to add to those already made. The main problem with the way you attached this is that it has modified a lot of the "special" characters, which you normally don't see because your mail reader has displays them ok. So you should do a "view source" of what you have attached. For example, it starts out: diff -u -r1.49 sysparams.c --- windows/sysparams.c 19 Feb 2003 22:04:46 -0000 1.49 +++ windows/sysparams.c 5 Mar 2003 10:39:39 -0000 @@ -1662,17 +1662,26 @@ } =20 case SPI_GETMOUSEHOVERWIDTH: /* 98 _WIN32_WINNT >=3D 0x400 || _W= IN32_WINDOW > 0x400 */ Notice the "=20", "=3D", and "=<cr>". Not only has the mailer changed some of the characters, but it has inserted line feeds. For many of us, all this is invisible. I use Mozilla, and it recreates the file correctly as long as I Save As... text. But I don't think we should assume that everyone can do that.

2 1

Spy++ messages
by Mike Hearn March 5, 2003

March 5, 2003

Hi, I tried debugging why toggling check items in the NSIS installer doesn't work (bug 1298), and thought it'd be a good idea to try running it under Spy++ on Windows, but for some reason it only seems to know about the TVM_* messages, but not TVN_* or TVS_*... is this correct, does Spy++ only show you a subset of the messages used? thanks -mike

2 1

Re: Screenshots?
by Brian Vincent March 5, 2003

March 5, 2003

> Lately I've been finding myself refer people to the preview screenshots > page hosted at users.theshell.com, and was wondering when they'd be > moved over to winehq.org? Yeah, that server tends to suck. Not sure what's been causing them problems but "high availability" doesn't seem to be a priority. My goal is to move them over when Jeremy completes changes to the web site. There's a lot of problems doing that before then, mostly having to do with integrating nicely with Jeremy's stuff. For instance: 1) All the graphics need to be optimized - they're pretty bad right now. And if I have to do that then I might as well make sure they're the right size. (I picked a rather arbitrary size, that probably needs to be adjusted.) 2) The list needs to be cut down. I think it's about twice as large as a it really needs to be. 3) And of course the HTML (XML? PHP?) needs to be written for them. None of those are things I want to do twice. -brian

1 0

Re: SystemParametersInfo patch
by Andreas Mohr March 5, 2003

March 5, 2003

On Wed, Mar 05, 2003 at 11:48:54AM +0100, Jon Bright wrote: > Hi, > > In the process of trying to get a component installed in Borland > C++Builder 5, I ran across an AV in SystemParametersInfo as a result of > a bug in the component. It seems Windows checks for NULL being passed > and returns an error instead of AVing. The attached patch makes Wine do > the same. Ah, good. Hmm, are you sure GetLastError() says zilch, nada? SystemParametersInfoA does have some SetLastErrors at other places... > (as an aside are patches preferred as MIME attachments or straight in > the mail here?) Straight in the mail is the usual way. However, after having raised some eye-brows, they will get applied anyway ;) -- Help prevent Information Technology Fascism! - before it's too late... http://www.againsttcpa.com

2 1

Re: SystemParametersInfo patch
by Uwe Bonnes March 5, 2003

March 5, 2003

>>>>> "Jon" == Jon Bright <jon(a)siliconcircus.com> writes: ... Jon> (as an aside are patches preferred as MIME attachments or straight Jon> in the mail here?) Straight text is preferred. And also a test case for the appropriate wine/dll/../test directory. Bye -- 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

Microsoft Installer compatability
by Mike Hearn March 5, 2003

March 5, 2003

I should spam the list less :) http://msi2xml.sourceforge.net/ Easy way to write a Wine version of MSI? Hardly important at the moment, but might be in future. -- Mike Hearn <m.hearn(a)signal.qinetiq.com> QinetiQ - Malvern Technology Center

2 1

Screenshots?
by Mike Hearn March 5, 2003

March 5, 2003

Hi, Lately I've been finding myself refer people to the preview screenshots page hosted at users.theshell.com, and was wondering when they'd be moved over to winehq.org? In particular, for some reason that server seems pretty unreliable from where I am, roughly half the time I try to access it, it appears to be offline. Does anybody know the status of the new website, or moving the screenshots to the current one? -- Mike Hearn <m.hearn(a)signal.qinetiq.com> QinetiQ - Malvern Technology Center

1 0

New conformance test for user32.dll
by Ferenc Wagner March 5, 2003

March 5, 2003

Hello, I put together a new test file for dlls/user/tests (find attached). It tests undocumented behaviour (at least I couldn't find a word), but one of my apps depend on it. It's a first try, please have a look at it (I tried to follow the Developers' Guide) and tell me what I should do. I compiled a previous version (some code rearrangement happened since then) with MSVC, it ran fine on XP, ME and Win95, ie. all tests passed. In short: a LISTBOX window can have the LBS_NOSEL style option. For me it seems it's only about appearance and nothing more. But Wine implements it in a more meaningful way, unfortunately... I'd like to be sure about what's correct before patching the implementation. Looking forward to your comments: Feri. Insert listbox.c into CTESTS in Makefile{,.in} to run:

3 3

New winmm/waveIn tests fail
by Sylvain Petreolle March 5, 2003

March 5, 2003

[syl(a)snoop winmm]$ make test make[1]: Entre dans le répertoire `/home/wine/dlls/winmm/tests' ../../../tools/runtest -q -P wine -M winmm.dll -T ../../.. -p winmm_test.exe.so wave.c && touch wave.ok wave.c:395: Test failed: waveInOpen: opening the device at 2MHz should fail 0: rc=4 make[1]: *** [wave.ok] Erreur 1 make[1]: Quitte le répertoire `/home/wine/dlls/winmm/tests' make: *** [tests/__test__] Erreur 2 Im using wineoss.drv with alsa-rc7 OSS emulation. ===== Sylvain Petreolle spetreolle(a)users.sourceforge.net Fight Spam ! EuroCauce: http://www.euro.cauce.org/en/index.html ICQ #170597259 "Don't think you are. Know you are." Morpheus, in "Matrix". ___________________________________________________________ Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et en français ! Yahoo! Mail : http://fr.mail.yahoo.com

1 0

Re: ntdll/kernel split 2: SO DLL loading removal
by Dan Kegel March 4, 2003

March 4, 2003

Eric Pouech wrote: > as (shortly) discussed on wine-devel, this removes SO loading for DLL > I think only glide* DLLs would be affected and should require a wrapper ... > - Native ELF libraries. Will not work yet. > + Native ELF libraries. Has been deprecated, ignored. Deprecated means "still works, but not recommended". Please use "obsolete" rather than "deprecated" for things that are removed. - Dan -- Dan Kegel http://www.kegel.com http://counter.li.org/cgi-bin/runscript/display-person.cgi?user=78045

1 0

Handling of forgotten accelerators keys
by Uwe Bonnes March 4, 2003

March 4, 2003

As diskussed on wine-devel. Maybe the TRACE should be a WARN? Changelog: wine/controls/menu.c Don't try any longer when we have decided the accelerator key is invalid -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ---------- Index: wine/controls/menu.c =================================================================== RCS file: /home/wine/wine/controls/menu.c,v retrieving revision 1.162 diff -u -w -r1.162 menu.c --- wine/controls/menu.c 4 Mar 2003 02:14:58 -0000 1.162 +++ wine/controls/menu.c 4 Mar 2003 10:26:38 -0000 @@ -4744,6 +4744,7 @@ TRACE_(accel)(", but won't send WM_{SYS}COMMAND, reason is #%d\n",mesg); if(mesg==0) ERR_(accel)(" unknown reason - please report!"); + return FALSE; } return TRUE; }

1 0

Re: Named Pipe patch
by Alexandre Julliard March 4, 2003

March 4, 2003

Mike McCormack <mike(a)codeweavers.com> writes: > The patch moves away from using unix pipes to implement named pipes, > and instead implements pipes internally. This is necessary to > implement message mode properly (without race conditions). Could you explain this a bit more? Why can't you use a datagram socket for message mode? I think we really want to avoid doing I/O through the server, our I/O is already slow enough as it is... -- Alexandre Julliard julliard(a)winehq.com

3 3