August 2015 - Wine-devel - list.winehq.org

Prelink usage
by Michael Cronenworth 14 Aug '15

14 Aug '15

Does prelink still provide any benefits in wine? The package is being retired in Fedora so starting with 1.7.48 the wine package will no longer be compiled with it. A few applications I have run under 1.7.48 still function, but I am curious to know the possible failure cases. Thanks, Michael

4 6

Re: [PATCH] mshtml: Handle nsIDOMWindow_GetFrames returning NULL
by Jacek Caban 14 Aug '15

14 Aug '15

Hi Indrek, On 08/09/15 09:09, Indrek Altpere wrote: > + if(!nsframes) { > + WARN("nsIDOMWindow_GetFrames returned NULL nsframes: %p\n", This->nswindow); > + return E_FAIL; The patch looks mostly good, but I think we should return DISP_E_MEMBERNOTFOUND in this case. Thanks, Jacek

1 0

Re: [PATCH] msvcr120: Implement strtof and _strtof_l (try 2)
by Henri Verbeet 13 Aug '15

13 Aug '15

On 13 August 2015 at 02:23, Bernhard Übelacker <bernhardu(a)vr-web.de> wrote: > +static inline BOOL almost_equal(float d1, float d2) { > + if(d1-d2>-1e-30 && d1-d2<1e-30) > + return TRUE; > + return FALSE; > +} > + You almost never want to compare floating point values like that. See e.g. compare_float() in dlls/ddraw/tests/ddraw7.c for a better (if perhaps not quite perfect) way. E.g. for the FLT_MAX case this only works because you happen to get an exact match. "1e-30" is a double instead of a float. Of course this is probably copied from the existing tests, so I suppose I can't complain too much.

2 1

Re: [PATCH] msvcr120: implement strtof and _strtof_l
by Piotr Caban 12 Aug '15

12 Aug '15

On 08/12/15 00:42, Bernhard Übelacker wrote: > + ok(*(unsigned*)&f == __inf_bytes && sizeof(unsigned) == sizeof(float), "f = %x\n", *(unsigned*)&f); It would be nicer to use _finitef from msvcr120 here. Thanks to it you can avoid __inf_bytes and __ninf_bytes declaration. Thanks, Piotr

1 0

Re: [PATCH 1/2] reg: simplify root key search
by Sebastian Lackner 11 Aug '15

11 Aug '15

On 06.08.2015 23:14, xantares 09 wrote: > From: Jonathan Vollebregt <jnvsor(a)gmail.com> > > --- > programs/reg/reg.c | 85 +++++++++++++++++++++++++++--------------------------- > 1 file changed, 43 insertions(+), 42 deletions(-) It is a bit pointless to send patches anonymously to the wine-patches mailing list. No matter if you are the author or not, when you are sending the patches you are the person responsible for it. To establish trust, all developers should use their real names here on the mailing lists. (see the rules for submitting patches: http://wiki.winehq.org/SubmittingPatches ) Besides that I see much bigger chances when Jonathan would send these patches himself. He is experienced with his own code, can respond to feedback, can fix regressions if it causes problems, and so on. For people new to Wine development its always easier to start with some self-written patches. @Jonathan: I would like to encourage you again to give it another try. If you need any help, feel free to ask me. I assume Stefan Dösinger would also be willing to help again. Regards, Sebastian

3 2

Re: [1/3] usp10/tests: Add tests for control character glyphs.
by Marvin 11 Aug '15

11 Aug '15

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at https://testbot.winehq.org/JobDetails.pl?Key=15955 Your paranoid android. === wvistau64 (32 bit usp10) === Failure running script in VM: timed out waiting for the child process

1 0

Re: quartz: waveparser support wav file generated by ffmpeg (try3)
by Dmitry Timoshkov 11 Aug '15

11 Aug '15

"Changhui LIU" <liuchanghui(a)linuxdeepin.com> wrote: > Superseded patch 113457. > > > Change log: > remove todo_wine win_skip The patch looks good, and it actually more closely resembles mmioDescend behaviour (except of alignment to even boundary, but that's a separate issue). -- Dmitry.

2 1

Re: comctl32: ILC_COLORDDB imagelists can be created with 0 sizes
by Huw Davies 11 Aug '15

11 Aug '15

On Mon, Aug 10, 2015 at 08:21:52PM -0500, Aric Stewart wrote: > diff --git a/dlls/comctl32/tests/imagelist.c b/dlls/comctl32/tests/imagelist.c > index 6f13846..fcf5162 100644 > --- a/dlls/comctl32/tests/imagelist.c > +++ b/dlls/comctl32/tests/imagelist.c > @@ -1994,6 +1994,47 @@ static void test_create_destroy(void) > ok(rc == FALSE, "ImageList_Destroy(0xdeadbeef) should fail and not crash\n"); > } > > +static void test_0size_DIB(void) > +{ > + INT cx, cy; > + HIMAGELIST himl; > + IImageList *imgl; > + HRESULT hr; > + INT ret; > + > + himl = ImageList_Create(0, 14, ILC_COLORDDB, 4, 4); This seems an odd choice of function name when you're actually testing DDBs. Also, I think zero_size is much nicer than 0size. Better yet, move all of these into test_create_destroy() where it'll make the difference in behaviour between these and the DIBs more apparent. Huw.

1 0

Re: comctl32: ILC_COLORDDB imagelists can be created with 0 sizes
by Marvin 11 Aug '15

11 Aug '15

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at https://testbot.winehq.org/JobDetails.pl?Key=15944 Your paranoid android. === w2000pro (32 bit imagelist) === imagelist: unhandled exception c0000005 at 0041CED0

1 0

Re: [PATCH 2/5] dwrite: Use OS/2 table data to set font style, make it preferable
by Dmitry Timoshkov 10 Aug '15

10 Aug '15

Nikolay Sivov <nsivov(a)codeweavers.com> wrote: > if (tt_os2) { > + USHORT version = GET_BE_WORD(tt_os2->version); > + USHORT fsSelection = GET_BE_WORD(tt_os2->fsSelection); > + > if (GET_BE_WORD(tt_os2->usWidthClass) <= DWRITE_FONT_STRETCH_ULTRA_EXPANDED) > props->stretch = GET_BE_WORD(tt_os2->usWidthClass); > > props->weight = GET_BE_WORD(tt_os2->usWeightClass); > + if (version >= 4 && (fsSelection & OS2_FSSELECTION_OBLIQUE)) > + props->style = DWRITE_FONT_STYLE_OBLIQUE; > + else if (fsSelection & OS2_FSSELECTION_ITALIC) > + props->style = DWRITE_FONT_STYLE_ITALIC; It's much better to use freetype flags for bold and italic, that way styles will match between gdi32 and dwrite. Besides I'd prefer to trust freetype instead of some handmade code which was not tested by thousands of fonts out there. -- Dmitry.

2 4