October 2008 - Wine-Devel - list.winehq.org

Re: comctl32 tree problem?

by Juan Lang

> Here's an app that says they had to work around a problem in our tree control: > http://www.webyog.com/blog/2008/10/30/sqlyog-712-has-been-released/ Maybe you could contact the developer and ask him to log a bug? --Juan

Oct. 30, 2008

Present Wine at CeBit?

by Jeremy White

Anyone want to give a presentation on Wine at CeBit? http://www.linux-magazine.com/online/news/cebit_open_source_linux_magazine_… If you're interested, email me privately, and I'll connect you with Britta. Cheers, Jeremy

Oct. 30, 2008

Re: Vertex pipeline replacement

by Chris Spencer

Hi Stefan, I have tested this with Age of Myhology (Titans 1.03) on wine-1.1.7-106-g38815e1, patches 1-11 don't seem to cause any problems (haven't checked extensively though), but upon applying the 12th (0012-WineD3D-ARBvp-ffp-replacement.patch) the water lost it's transparency [1]. I'm not sure what other information you need, so just let me know. I'm happy to test any changes you make. Chris. [1] http://i37.tinypic.com/dcyjac.png

Oct. 30, 2008

comctl32 tree problem?

by Dan Kegel

Here's an app that says they had to work around a problem in our tree control: http://www.webyog.com/blog/2008/10/30/sqlyog-712-has-been-released/

Oct. 30, 2008

Re: DIrectSound on top of openAL ?

by Roderick Colenbrander

> If all of the wish list gets added, it would appear the next > specification of OpenAL would appear to me to go a long way to allowing > wine to implement the Windows sound API's via that, rather than having > to support various drivers. > > -- > Darragh Personally I think that even if OpenAL added the functionality we need that the user experience for most users still won't be good. Most users want to play lets say mp3's while they are gaming. This means that in order to work properly OpenAL under the hood again needs to use alsa+dmix or even worse pulseaudio. For that reason I'm not sure if going directly to OpenAL instead of going through wineoss/winealsa is any better. Roderick -- "Feel free" - 5 GB Mailbox, 50 FreeSMS/Monat ... Jetzt GMX ProMail testen: http://www.gmx.net/de/go/promail

Oct. 30, 2008

Re: Changed more malloc to HeapAlloc calls

by Michael Stefaniuc

Pete Myers wrote: >> In addition to what James said please split the patch up; one patch per >> module. A dll or program is a "module". > > Sorry, I'm inundated with noobie errors here I think. No problem. Janitorial tasks are really good to accustom new people to the patch submission process. >> Also while you are at it please remove the superfluous casts in the lines >> you are changing. HeapAlloc (malloc too) returns a void pointer which >> doesn't needs to be casted to an other pointer type. E.g. > > Will do. Didn't want to change coding style. Well that's not coding style. Superfluous casts are evil. I have even created a new janitorial tasks for that: http://wiki.winehq.org/SuperfluousCasts Also if one has to touch a line it is expected from him to fix all errors in that line. That includes dangling whitespace at the end of the line and mixed tab and space indentation. > Actually my big struggle at the moment is trying to figure out where > malloc() memory gets free(). What's the policy here? Should Just search for free in that file. > HeapAlloc() always be set free... Is the fact that some malloc() > memory not free()'d an error, or a policy? Or should I work harder to You need to replace the free calls with HeapFree calls. > find out where free() gets called? grep -w free dlls/iphlpapi/tests/iphlpapi.c dlls/wnaspi32/winaspi16.c That will do. bye michael

Oct. 30, 2008

Re: Changed more malloc to HeapAlloc calls

by Pete Myers

> In addition to what James said please split the patch up; one patch per > module. A dll or program is a "module". Sorry, I'm inundated with noobie errors here I think. > Also while you are at it please remove the superfluous casts in the lines > you are changing. HeapAlloc (malloc too) returns a void pointer which > doesn't needs to be casted to an other pointer type. E.g. Will do. Didn't want to change coding style. Actually my big struggle at the moment is trying to figure out where malloc() memory gets free(). What's the policy here? Should HeapAlloc() always be set free... Is the fact that some malloc() memory not free()'d an error, or a policy? Or should I work harder to find out where free() gets called? Pete

Oct. 30, 2008

Re: mshtml: Implement IHTMLElement get_document

by Jacek Caban

Hi Alistair, Alistair Leslie-Hughes wrote: > Hi, > > Changelog: > mshtml: Implement IHTMLElement get_document > + *p = NULL; + + if(!This->node.doc->nsdoc) { + WARN("NULL nsdoc\n"); + return E_UNEXPECTED; + } You don't use nsdoc so there is no reason to test it. + hr = IHTMLDocument2_QueryInterface(HTMLDOC(This->node.doc), &IID_IHTMLDocument, (void**)p); + + return hr; You could just cast This->node.doc to IHTMLDocument and call AddRef on result. Jacek

Oct. 30, 2008

Re: [PATCH 1/3] fusion: Implement the IAssemblyEnum interface. [try3]

by Jacek Caban

Hi James, James Hawkins wrote: > --- > dlls/fusion/Makefile.in | 1 + > dlls/fusion/asmcache.c | 83 ---------- > dlls/fusion/asmenum.c | 385 ++++++++++++++++++++++++++++++++++++++++++++++ > dlls/fusion/asmname.c | 15 +-- > dlls/fusion/fusion.c | 12 -- > dlls/fusion/fusionpriv.h | 15 ++ > 6 files changed, 402 insertions(+), 109 deletions(-) > create mode 100644 dlls/fusion/asmenum.c > + +done: + FindClose(hfind); + return S_OK; +} + You always set hr in error paths so I guess you want to return it here. Also, it's a matter of style, but these gotos look strange. You could replace them all by break statements. Jacek

Oct. 30, 2008

Re: Changed more malloc to HeapAlloc calls

by Michael Stefaniuc

Hello Pete, James Hawkins wrote: > On Wed, Oct 29, 2008 at 8:02 PM, Pete Myers > <peterdanielmyers(a)googlemail.com> wrote: >> It does match the style of the way that HeapAlloc is used elsewhere in the >> file, though not malloc admittedly. I wasn't sure which convention had >> precedent. >> > > Please bottom-post when replying to a post on the wine lists. There > is one test function that uses a space after opening parentheses, but > it's the exception, not the rule. > >> I'm afraid that I don't understand how my changelog doesn't match the >> patch. I'm new here, can you help me out? There's obviously something I'm >> missing! >> > > You had several 'changelog' entries in the email, but the last one was > > Changelog: > * malloc calls in dlls/kernel32/process.c have been change to HeapAlloc > > which does not match the patch. In addition to what James said please split the patch up; one patch per module. A dll or program is a "module". Also while you are at it please remove the superfluous casts in the lines you are changing. HeapAlloc (malloc too) returns a void pointer which doesn't needs to be casted to an other pointer type. E.g. --- a/dlls/iphlpapi/tests/iphlpapi.c +++ b/dlls/iphlpapi/tests/iphlpapi.c @@ -225,7 +225,7 @@ static void testGetIpAddrTable(void) "GetIpAddrTable(NULL, &dwSize, FALSE) returned %d, expected ERROR_INSUFFICIENT_BUFFER\n", apiReturn); if (apiReturn == ERROR_INSUFFICIENT_BUFFER) { - PMIB_IPADDRTABLE buf = (PMIB_IPADDRTABLE)malloc(dwSize); + PMIB_IPADDRTABLE buf = (PMIB_IPADDRTABLE)HeapAlloc( GetProcessHeap(), 0, dwSize); apiReturn = gGetIpAddrTable(buf, &dwSize, FALSE); ok(apiReturn == NO_ERROR, should really be just: PMIB_IPADDRTABLE buf = HeapAlloc(GetProcessHeap(), 0, dwSize); thanks bye michael

Oct. 30, 2008