May 2012 - Wine-devel - list.winehq.org

Re: [PATCH 5/5] d3dcompiler: Partially implement D3DCompile function.
by Henri Verbeet 08 May '12

08 May '12

I realize compile_shader() is mostly a copy of assemble_shader(), but nevertheless: On 8 May 2012 16:17, Matteo Bruni <mbruni(a)codeweavers.com> wrote: > +struct bwriter_shader *parse_hlsl_shader(const char *text, enum shader_type type, DWORD version, If this isn't called outside of compiler.c it should be static. > + LPD3DBLOB buffer; "ID3DBlob *" > + CopyMemory(pos, preproc_messages, strlen(preproc_messages) + 1); memcpy() should do fine. > + hr = SlWriteBytecode(shader, 9, &res); ... > + size = HeapSize(GetProcessHeap(), 0, res); > + hr = D3DCreateBlob(size, &buffer); That's not very pretty, SlWriteBytecode() should probably just return the size as well.

2 1

Re: mscoree: Implement vtable fixups.
by Alexandre Julliard 08 May '12

08 May '12

Vincent Povirk <madewokherd(a)gmail.com> writes: > + if (fixup->fixup->type & COR_VTABLE_32BIT) > + { > + DWORD *vtable = fixup->vtable; > + DWORD *tokens = fixup->tokens; > + for (i=0; i<fixup->fixup->count; i++) > + { > + TRACE("%x\n", tokens[i]); > + vtable[i] = (DWORD)host->mono->mono_marshal_get_vtfixup_ftnptr( > + image, tokens[i], fixup->fixup->type); This will cause warnings on 64-bit. -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: kernel32: Add UTF-7 support
by Alex Henrie 08 May '12

08 May '12

2012/5/7 Ben Klein <shacklein(a)gmail.com>: > Without being a regular contributor, here are a few tips from a cursory glance: > 1) Whitespace change on at least one line The only change to whitespace I made was: - if( flags) + if (flags) I don't think this is a big deal. > 2) Why the change from ERROR_INVALID_PARAMETER to ERROR_INVALID_FLAGS? To match the behavior of Windows. I verified this with both MSDN and empirical tests. If the flags parameter is not null, all versions of Windows tested report ERROR_INVALID_FLAGS with the possible exception of Windows NT4. See http://testbot.winehq.org/JobDetails.pl?Key=18148 > 3) New function defs do not match format of surrounding defs in unicode.h This was a hard one. I didn't know if I should keep Katayama's use of INT or change the functions to use int like the UTF-8 functions do. So you're saying it should all be int? Do LPCSTR and LPCWSTR need to be changed to something else as well? > 4) No new tests to confirm the behaviour is correct (or mention of > current tests succeeding) dlls/kernel32/tests/locale.c already has just as many UTF-7 tests as it has UTF-8 tests. What tests do we need for UTF-7 which are not needed for UTF-8? Thanks for your quick response, -Alex

2 1

Re: msi: Tweak msiexec's usage message.
by Julian Rüger 07 May '12

07 May '12

Hi Francois and wine-devel, while we're at it, shouldn't these >"Advertise a product:\n" >"\t/j[u|m] package [/t transform] [/g languageid]\n" >"Apply a patch:\n" >"\t/p patch_package [property]\n" >"\t/p patch_package /a package [property]\n" also be put in curly brackets? Like this: \t/j[u|m] {package} [/t transform] ... \t/p {patch_package} [property]\n ... Best, Julian

3 3

Re: [Programs] Notepad: sync with ReactOS notepad (2)
by carlo.bramix＠libero.it 07 May '12

07 May '12

Hello friends, >> Because you do seriously believe that ReactOS developers reversed the >> Windows notepad?! Along with Solitaire & calculator, I suppose? >> Common sense is a powerful tool, let's make use of it :-). > >It's a matter of trust. If a person is known to use illegal practices >to write ReactOS code then even adding one line of code to notepad >makes her/his contribution dangerous for the entity accepting that code. > >> For the record, ReactOS as well disallows any person who has seen the >> Microsoft Windows code to comment, contribute, commit to the project. > >That's just a blanket statement, nothing more. It's hard to believe >to it if the person(s) publicly accused in using leaked code still >commit to ReactOS without any limitation. as I already wrote, I totally agree. Please reject the patches. Sincerely, Carlo Bramini.

1 0

Re: [Programs] Notepad: sync with ReactOS notepad (2)
by carlo.bramix＠libero.it 07 May '12

07 May '12

Hello friends, >The issue is whether the appropriate guidelines were followed >(http://wiki.winehq.org/CleanRoomGuidelines) One issue is that Wine >does not allow people that have ever viewed Microsoft source code to >contribute to Wine, but there are other issues. You'll have to talk >to Alexandre and see how he feels about this specific issue, but if >you've followed all of the guidelines and contribute only your own >changes then he might be ok with your contribution. I see. Perhaps it would be better to reject all these patches since I'm not the only author of all the changes. I understand that this is done to protect yourself and the project, because you do not know the people who touched the sources of the notepad, who they are and how they collected their information, so I agree with you. Sincerely, Carlo Bramini.

3 2

Re: [Programs] Notepad: sync with ReactOS notepad (2)
by carlo.bramix＠libero.it 06 May '12

06 May '12

Hello friends, >thank you for working on this. I think you will improve the chances for >your patches to be accepted if you give them individually meaningful >names, like > >"notepad: Don't use reserved MSVC keyword." > >for patch #2. Actually, I followed the tip I read into the wiki of WINE in "Patch submission guidelines", so I tried to imitate the style of other patches for WINE and for this reason I used an identical title with a progressive number as subject. If something different is prefered, no problem, I will follow your suggestion and I will use a more focused title. >Also, like Jerome said, we may have a problem if different people >commited these changes to ReactOS. I'm no expert on this, but as I >understand, Wine needs the patch to be credited to the author who >actually wrote the code. Well, ReactOS notepad had been imported from WINE long time ago and various people did some changes on it and I'm included in them since, in this specific case, the status bar implementation had been almost rewritten by me. I'm not an expert too, but since it is distribuited under the same license of WINE, with same coding restrictions and all people who sent patches have agreed to the same GPL as well, my intuition would say that there is no problem on this... because the owner who can providing this code to you is the ReactOS project itself. That's just my opinion of course... I will wait for your suggestions/orders... ;-) Sincerely, Carlo Bramini.

4 3

Re: [Programs] Notepad: sync with ReactOS notepad (2)
by Julian Rüger 06 May '12

06 May '12

Hi Carlo, thank you for working on this. I think you will improve the chances for your patches to be accepted if you give them individually meaningful names, like "notepad: Don't use reserved MSVC keyword." for patch #2. Also, like Jerome said, we may have a problem if different people commited these changes to ReactOS. I'm no expert on this, but as I understand, Wine needs the patch to be credited to the author who actually wrote the code. (Can someone else please comment on this, if I am mistaken?) Best regards, Julian

1 0

Re: rpcrt4/tests: Fix build with PSDK headers.
by Marvin 05 May '12

05 May '12

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 http://testbot.winehq.org/JobDetails.pl?Key=18238 Your paranoid android. === WNT4WSSP6 (32 bit ndr_marshall) === ndr_marshall: unhandled exception c0000005 at 004081C6

1 0

Re: [Programs] Notepad: sync with ReactOS notepad
by Jerome Leclanche 05 May '12

05 May '12

On Sat, May 5, 2012 at 2:15 PM, carlo.bramix(a)libero.it < carlo.bramix(a)libero.it> wrote: > Attached patch synces the WINE's notepad with the improvements added into > ReactOS' Notepad. > > - Imported status bar implementation, it fixes Bug #18715. New menu voice > has > been introduced into the resource file. > This is the biggest part of the patch, that it caused to touch all files > modified by the patch. > Other minor fixes included here are: > > - Implemented GOTO function. Since its menu voice has been added into the > resource file when adding support the statusbar, the related dialog box > and the > source code have been introduced too. > > - Find/Replace functions have an extremely similar behavior, so the job > can be > implemented with a single function. > > - Find and Find next commands must be grayed if there is not a text to > search > into. > > - The "class" word is a reserved keyword in MSVC, so I also imported the > fix > into the creation of the class of the window: "class" has been replaced > with > "wndclass", no code change. > > Sincerely, > > Carlo Bramini > > It's not going to be accepted as one big patch. It needs to be split. Also, are you the author of all the patches or are these other contributors for this? J. Leclanche

1 0