September 2003 - Wine-Devel

Installshield GetDC/ReleaseDC regression?
by Robert Reif Sept. 5, 2003

Sept. 5, 2003

Installshield for dx8.1 sdk with win98 set in config is giving this error: err:dc:DCE_FreeWindowDCE[0x10023] GetDC() without ReleaseDC! A trace is attached.

1 0

Re: commctrl.h cleanup
by Alexandre Julliard Sept. 5, 2003

Sept. 5, 2003

"Dimitrie O. Paun" <dpaun(a)rogers.com> writes: > I know, it would be nice to have an intelligent utility to > cleanup the headers, but by the looks of it, it has to be > rather smart. This is probably one of the more difficult > cleanups, and I did it all by hand, which gave me a chance > to cleanup/reorganize the includes a bit. Looks like you have been reorganizing a bit too much, config.h and wine/port.h should always be included before anything else. > All other (except > for winbase.h) are *a lot* simpler. People can do them by > hand quite easily. I'm not so sure. For instance prsht.h still includes windef.h, wingdi.h and winuser.h, and since it's included from commctrl.h in fact your patch didn't actually remove these. I think there will still be a lot of changes to make, and I'd really prefer to avoid having to fix the same C files again and again. -- Alexandre Julliard julliard(a)winehq.com

3 7

Re: [dcom 1.2] Implement dispatch variant marshalling
by Mike Hearn Sept. 4, 2003

Sept. 4, 2003

On Thu, 2003-09-04 at 20:53, Dimitrie O. Paun wrote: > I'd prefer we don't have there -- they really clutter the code, > provide a bad example to follow, and provide little, if any, > benefit. What you can do is put some strategic TRACEs that would > help someone who knows the code to identify the problem. > But I wouldn't worry about memory allocation failures... Another possibility is to have a wine convenience macro, similar to the glib g_return_if_fail() macro, that works for instance: wine_return_if_fail_hr(CoDoSomethingCOMRelated()) which expands to { HRESULT hr = CoDoSomethingCOMRelated() if (hr != S_OK) { ERR("CoDoSomethingCOMRelated() failed, hr=%ld\n", hr); return hr; } or whatever..... could be convenient, but perhaps too much magic. I'll drop the ERRs then. thanks -mike

2 1

Re: [dcom 1.2] Implement dispatch variant marshalling
by Dimitrie O. Paun Sept. 4, 2003

Sept. 4, 2003

On 4 Sep 2003, Mike Hearn wrote: > I removed the use of asserts to check the results, and replaced them > with ERRs. > + working_mem = GlobalAlloc(0, size); > + if (!working_mem) { > + ERR("failed to get a global alloc\n"); > + return oldpos; > + } I don't think this is such a good idea. If the API can fail, it should just fail. ERR should be used only to signal internal inconsistencies. Doing an ERR on very memory allocation / function call that can fail is not only useless, but harmful, as it bloats and clutters the code. The above should just be: working_mem = GlobalAlloc(0, size); if (!working_mem) return oldpos; -- Dimi.

1 0

TreeView & WM_NOTIFY
by Igor Grahek Sept. 4, 2003

Sept. 4, 2003

Hi :) My problem is very similar to bug 1227 (http://bugs.winehq.org/show_bug.cgi?id=1227) Treeview in application is sending WM_NOTIFY on double click (NM_DBLCLK) to the wrong Window handle. GetParent() function returns handle to the root window and root window is not forwarding/broadcasting that message and message is lost. Everything works fine under Windows or with --dll comctl32=n Any suggestions? Igor

2 2

Re: winres.h cleanup
by jonwil＠tpg.com.au Sept. 4, 2003

Sept. 4, 2003

I dont see the reason for this change... The real version of winres.h from microsoft includes winresrc.h

3 2

Bugzilla down?
by Dustin Navea Sept. 4, 2003

Sept. 4, 2003

Hi guys, I just tried to access Bugzilla, and it isnt responding, checked with a ping and a traceroute but both gave me request timed out, although that could be because of a firewall, I just figured someone might want to know (jeremy).. ===== Dustin Navea Minor Contributor, http://www.winehq.com Bugzilla Janitor, http://bugs.winehq.com Network Admin, irc://irc.blynk.net (down) __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com

2 1

Re: stdarg.h in winbase.h
by Alexandre Julliard Sept. 4, 2003

Sept. 4, 2003

"Dimitrie O. Paun" <dpaun(a)rogers.com> writes: > With this patch, we can get rid of the stdarg.h > include from winbase.h if it weren't for the > FormatMessage() functions: > > DWORD WINAPI FormatMessageA(DWORD,LPCVOID,DWORD,DWORD,LPSTR,DWORD,va_list*); > DWORD WINAPI FormatMessageW(DWORD,LPCVOID,DWORD,DWORD,LPWSTR,DWORD,va_list*); > > Alexandre, how do we handle this one? We have to include stdarg.h everywhere we include winbase.h, about 650 files at last count. That's why we need some kind of script... -- Alexandre Julliard julliard(a)winehq.com

4 6

Re: wineproc: a new wine utility
by Dimitrie O. Paun Sept. 4, 2003

Sept. 4, 2003

On September 3, 2003 03:53 pm, Eric Pouech wrote: > any comment welcomed!! This is very nice (from the comments and the code)! I will compile it soon, but I was hoping Alexandre bits me to it and commits it, so I can be lazy and just to a 'cvs up' :) > (thanks to Dimi for his quick fixies on the listview control BTW) You are most welcome, thank you for spotting the problems! -- Dimi.

1 0

Re: [gethostbyname] Set appropriate error code
by Rein Klazes Sept. 3, 2003

Sept. 3, 2003

On Wed, 3 Sep 2003 18:20:56 +0200, you wrote: > Changelog: > - Set appropriate error code for gethostbyname and WSAAsyncGetHostByName > - Restore NULL name/aq->host_name to NULL at the end of each function respectively > > Stephan > > aq->host_name = buf; > if( gethostname( buf, 100) == -1) { > - fail = WSAENOBUFS; /* appropriate ? */ > + fail = (errno == EINVAL) ? WSAEFAULT : wsaErrno(); Explanation why I arrived at WSAENOBUFS as appropriate value.. (Remember it is used as the error value for the windows code calling gethostbyname(), not gethostname();) - glibc documentation as I read it indicates that in this case the only way gethostname() may fail is an insufficient bufferlength. Clearly invalid name pointer or negative length do not apply. - WSAEFAULT means (by Microsoft): "The system detected an invalid pointer address in attempting to use a pointer argument in a call" At the vary least this will confuse somebody that does gethostbyname(NULL), thinking it is the null pointer that is causing this error. This is the one error code that I did not want to return, that is why I cannot use the wsaErrno functions. But perhaps you have found another cause where gethostname() will fail? > + if (name==buf) name=NULL; > return retval; name is an argument, passed by value. Destroyed as the function returns. No need to restore. Rein. -- Rein Klazes rklazes(a)xs4all.nl

1 0