January 2013 - Wine-Devel - list.winehq.org

Re: windowscodecs: Implement GetThumbnail in the ICO frame decoder.

by Hans Leidekker

On Wed, 2013-01-30 at 18:52 +0800, Dmitry Timoshkov wrote: > > The test added by 6395af1ae7b0cc5f2f1f82796502e2a605bc5a6b says otherwise, GetThumbnail > > is supported for ICO frames. > > It's still better to return WINCODEC_ERR_CODECNOTHUMBNAIL instead of > E_INVALIDARG. Real implementation depends on the codec set, and although > obviously MS implementation returns a real interface, that doesn't mean > that an implementation that returns WINCODEC_ERR_CODECNOTHUMBNAIL is wrong. Well, the app that prompted me to write that test wasn't fooled by returning WINCODEC_ERR_CODECNOTHUMBNAIL (which is what was returned before my patch). And since you're changing the FIXME to a TRACE you are effectively hiding the bug.

Jan. 30, 2013

Re: windowscodecs: Implement GetThumbnail in the ICO frame decoder.

by Hans Leidekker

On Wed, 2013-01-30 at 18:02 +0800, Dmitry Timoshkov wrote: > diff --git a/dlls/windowscodecs/icoformat.c b/dlls/windowscodecs/icoformat.c > index 7aec245..da5a733 100644 > --- a/dlls/windowscodecs/icoformat.c > +++ b/dlls/windowscodecs/icoformat.c > @@ -200,8 +200,12 @@ static HRESULT WINAPI IcoFrameDecode_GetColorContexts(IWICBitmapFrameDecode *ifa > static HRESULT WINAPI IcoFrameDecode_GetThumbnail(IWICBitmapFrameDecode *iface, > IWICBitmapSource **ppIThumbnail) > { > - FIXME("(%p,%p)\n", iface, ppIThumbnail); > - return E_NOTIMPL; > + TRACE("(%p,%p)\n", iface, ppIThumbnail); > + > + if (!ppIThumbnail) return E_INVALIDARG; > + > + *ppIThumbnail = NULL; > + return WINCODEC_ERR_CODECNOTHUMBNAIL; > } The test added by 6395af1ae7b0cc5f2f1f82796502e2a605bc5a6b says otherwise, GetThumbnail is supported for ICO frames.

Jan. 30, 2013

Re: [2/6] windowscodecs: Implement ComponentFactory::CreateEncoderPropertyBag

by Alexandre Julliard

Ludger Sprenker <Ludger(a)sprenker.net> writes: > Previous "[2/5]" should have been "[3/6]" ... It doesn't build on Mac: gcc -m32 -c -I. -I. -I../../include -I../../include -D__WINESRC__ -DENTRY_PREFIX=WIC_ -DPROXY_DELEGATION -DWINE_REGISTER_DLL -D_REENTRANT -fPIC -Wall -pipe -fno-strict-aliasing -Wdeclaration-after-statement -Wempty-body -Wstrict-prototypes -Wwrite-strings -gstabs+ -fno-omit-frame-pointer -Wpointer-arith -I/sw/include/freetype2 -I/sw/include -I/sw/include -g -O2 -o icnsformat.o icnsformat.c icnsformat.c: In function 'IcnsEncoder_CreateNewFrame': icnsformat.c:666: warning: passing argument 1 of 'CreatePropertyBag2' from incompatible pointer type icnsformat.c:666: error: too few arguments to function 'CreatePropertyBag2' -- Alexandre Julliard julliard(a)winehq.org

Jan. 29, 2013

Re: gdiplus: Avoid not necessary memory allocation for BITMAPINFO.

by Vincent Povirk

Did you try running the tests in valgrind with this change? You're essentially reverting 0ecd8daa16e4aa3df157a1175c611cdffc66d6fe which I think was done to make valgrind happy.

Jan. 29, 2013

Re: kernel32: dstlen < 0 causes error in codepage conversion

by Alexandre Julliard

Alex Henrie <alexhenrie24(a)gmail.com> writes: > This is a pretty straightforward patch to resolve a couple of test > failures marked todo_wine. Hopefully it's not controversial, but if it > is, I am willing to make any reasonable modifications that are > communicated to me. If it was that easy it would have been done already. There are places in Wine that depend on being able to use negative sizes, so it would need a code review of all callers before we can make that change. -- Alexandre Julliard julliard(a)winehq.org

Jan. 29, 2013

[website] Added simplified chinese

by Jactry

Jan. 29, 2013

Re: winealsa: Have the MIDI recorder wait in poll(), not snd_seq_event_input().

by Johannes Kroll

On Thu, 24 Jan 2013 10:33:58 +0100 <Joerg-Cyril.Hoehle(a)t-systems.com> wrote: > Compare with wineoss which uses poll <= 0 already... > > Johannes Kroll mentioned that MIDIIn takes a long time to end. > Indeed, our testsuite ran 3 times into the "kill after 5s" timeout due to this bug. > Now the tests execute much faster. Just a quick note to confirm that this patch does fix the waiting problem for me. Thanks a lot! - Johannes

Jan. 29, 2013

Re: riched20: ME_StrDup() is no longer used so remove it.

by Huw Davies

On Tue, Jan 29, 2013 at 07:05:13AM +0100, Francois Gouget wrote: > --- > dlls/riched20/editor.h | 1 - > dlls/riched20/string.c | 5 ----- > 2 files changed, 6 deletions(-) I'm about to send in a patch that uses this again, so please hold off on this one. Huw.

Jan. 29, 2013

Re: winemenubuilder: Try to wait for creating icon files (try 2).

by Sergey Guralnik

On 2013-01-23 16:34, Sergey Guralnik wrote: > It supersedes next patch: > > From 4ca51cf16c5be2bcf34dddb7f33fe043c1d777c4 Mon Sep 17 00:00:00 > 2001 > From: Sergey Guralnik <serhio(a)etersoft.ru> > Date: Tue, 22 Jan 2013 15:58:20 +0200 > Subject: winemenubuilder: Use only existing file for icon creating. > > The logic is the same as before - wait for parent process termination > if > required icon file doesn't exist yet. But now used easier way to do > this. Is something wrong with this patch? It's in queue about a week without any feedback. -- Sergey

Jan. 29, 2013

Re: [PATCH] winebuild: Use Clang to assemble if found. (try 5)

by Alexandre Julliard

Charles Davis <cdavis5x(a)gmail.com> writes: > Try 5: Don't use strstr(3). > > The configure script invokes the C compiler to figure out if .cfi > pseudo-ops are supported, on the assumption that it will invoke the > system assembler. Winebuild, on the other hand, invokes the system > assembler directly. This caused a problem on Mac OS hosts when compiling > with Clang: configure would detect that .cfi ops were supported, because > Clang's integrated assembler supports them; however, when winebuild then > emitted assembly with .cfi ops in them, the system assembler would then > choke. This patch makes winebuild invoke Clang's assembler if it is > found in the path. It's causing warnings on my Mac: clang: warning: /usr/bin/clang: 'linker' input unused when '-c' is present clang: warning: /usr/bin/clang: 'linker' input unused when '-c' is present -- Alexandre Julliard julliard(a)winehq.org

Jan. 28, 2013