February 2010 - Wine-devel - list.winehq.org

Did today's git compile properly?
by Susan Cragin 11 Feb '10

11 Feb '10

It's probably me... But I got the following: .1 make[1]: Leaving directory `/home/susan/wine/loader' make[1]: Entering directory `/home/susan/wine/programs' rm -f /usr/local/bin/`dirname programs/msiexec/__installprog__` && /usr/bin/install -c wineapploader /usr/local/bin/`dirname programs/msiexec/__installprog__` /usr/bin/install: cannot create regular file `/usr/local/bin/programs/msiexec': No such file or directory make[1]: *** [programs/msiexec/__installprog__] Error 1 make[1]: Leaving directory `/home/susan/wine/programs' make: *** [programs/__install__] Error 2 susan(a)ubuntu:~/wine$

2 1

Re: shell32[1/2]: implements shell link's GetCurFile (with tests, fixes bug #21297)
by Austin English 11 Feb '10

11 Feb '10

2010/2/10 Mikołaj Zalewski <mikolaj(a)zalewski.pl>: > Some installers require this. + expect_eq_x(S_FALSE, IPersistFile_IsDirty(pf)); + expect_eq_x(S_FALSE, IPersistFile_GetCurFile(pf, &string)); + expect_eq_p(NULL, string); + expect_eq_x(S_OK, IShellLinkA_GetHotkey(link, &hotkey)); + expect_eq_x(0, hotkey); + expect_eq_x(S_FALSE, IPersistFile_IsDirty(pf)); + // Setting the same value will not set IsDirty... + expect_eq_x(S_OK, IShellLinkA_SetHotkey(link, 0)); + todo_wine expect_eq_x(S_FALSE, IPersistFile_IsDirty(pf)); + // ... but a different one will. You've got C++ comments in there... -- -Austin

3 2

Make issue with latest git
by chris ahrendt 11 Feb '10

11 Feb '10

Just did the latest git and got the following : all -c wineapploader /usr/local/bin/`dirname programs/msiexec/__installprog__` /usr/bin/install: cannot create regular file `/usr/local/bin/programs/msiexec': No such file or directory make[1]: *** [programs/msiexec/__installprog__] Error 1 make[1]: Leaving directory `/work/wine-git/programs' make: *** [programs/__install__] Error 2 if I went into the /usr/local/bin there is no programs directory so made the directory and reran make install or tools/install and worked fine... Did not need this until todays git as yesterdays worked just fine without the directory... chris

2 1

Re: [PATCH 1/7] msctf: Add langbarmgr.c
by Justin Chevrier 11 Feb '10

11 Feb '10

Thanks for the feedback. Would submitting this as one patch be acceptable? My intention was really to have all this applied in one go. If not, I'll split as recommended. I've got some fixing up to do and I'll resubmit. Justin On Thu, Feb 11, 2010 at 1:27 PM, Justin Chevrier <jchevrier(a)gmail.com> wrote: > Hi Justin, > > It's good that you're splitting patches, but you did it wrong. No patch may > depend on another patch that is later in series. After applying each patch > the tree must be in good state. That said, Wine won't compile after patch > #1 until patch #6 is applied and patch #2 there will cause compile warning > until patch #7 is applied. Also it's not good to add code that is dead, > such as your patch #1 until patch #3 is applied and part of patch #3 > without patch #4 (although such patches are sometimes accepted). I'd > suggest sending in order: > - patch #6 (note that the interface is called ITfLangBarMgr) > - patch #5 > - patch #4 (with better changelog) > - joined patches 1,2,3,7 > > Also langbarmgr.c seems to be better place for TF_CreateLangBarMgrJacek > implementation. > > > Jacek >

2 1

Re: [PATCH 1/7] msctf: Add langbarmgr.c
by Nikolay Sivov 11 Feb '10

11 Feb '10

On 2/11/2010 19:51, Justin Chevrier wrote: > --- > dlls/msctf/langbarmgr.c | 193 +++++++++++++++++++++++++++++++++++++++++++++++ > 1 files changed, 193 insertions(+), 0 deletions(-) > create mode 100644 dlls/msctf/langbarmgr.c > > > + This = HeapAlloc(GetProcessHeap(),HEAP_ZERO_MEMORY,sizeof(LangBarMgr)); > > You don't need HEAP_ZERO_MEMORY here, cause you will set all fields anyway. Also it looks like you specially split to several patches to have only one file per patch, that's not needed.

2 2

Re: [PATCH 1/7] msctf: Add langbarmgr.c
by Jacek Caban 11 Feb '10

11 Feb '10

Hi Justin, It's good that you're splitting patches, but you did it wrong. No patch may depend on another patch that is later in series. After applying each patch the tree must be in good state. That said, Wine won't compile after patch #1 until patch #6 is applied and patch #2 there will cause compile warning until patch #7 is applied. Also it's not good to add code that is dead, such as your patch #1 until patch #3 is applied and part of patch #3 without patch #4 (although such patches are sometimes accepted). I'd suggest sending in order: - patch #6 (note that the interface is called ITfLangBarMgr) - patch #5 - patch #4 (with better changelog) - joined patches 1,2,3,7 Also langbarmgr.c seems to be better place for TF_CreateLangBarMgrJacek implementation. Jacek

1 0

Re: oleaut32: Fix rounding
by Alexandre Julliard 11 Feb '10

11 Feb '10

André Hentschel <nerv(a)dawncrow.de> writes: > --- > dlls/oleaut32/tests/varformat.c | 10 ++++++++-- > dlls/oleaut32/varformat.c | 3 ++- > dlls/oleaut32/variant.c | 2 +- > 3 files changed, 11 insertions(+), 4 deletions(-) It doesn't work: ../../../tools/runtest -q -P wine -M oleaut32.dll -T ../../.. -p oleaut32_test.exe.so vartest.c && touch vartest.ok vartest.c:1347: Test failed: Expected cDig = 1, got 2 vartest.c:1348: Test failed: Digit[0], expected 1, got 0 vartest.c:1348: Test failed: Digit[1], expected 255, got 1 make[1]: *** [vartest.ok] Error 3 -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: user32: Remove leftover HINSTANCE16 code from LoadCursor/LoadIcon.
by Alexandre Julliard 11 Feb '10

11 Feb '10

Michael Stefaniuc <mstefani(a)redhat.de> writes: > The Wine code doesn't seem to pass a 16bit hinstance to > LoadCursor/LoadIcon; the only calls to those functions from 16bit code > use a 0 hinstance which is fine. 16-bit instances can come from 32-bit calls too, if we are dealing with a 16-bit window. -- Alexandre Julliard julliard(a)winehq.org

2 2

Re: runtest: search also for Applicationtests
by Alexandre Julliard 11 Feb '10

11 Feb '10

André Hentschel <nerv(a)dawncrow.de> writes: > So it works with cmd.exe_test.exe.so now You should compute the right name based on the directory instead of guessing. -- Alexandre Julliard julliard(a)winehq.org

1 0

Adding wikipedia links to appdb?
by Joerg-Cyril.Hoehle＠t-systems.com 11 Feb '10

11 Feb '10

-1 1. AFAIK, one can add (after submission, i.e. maintainers only) as many URL's as one wants. So one could well add one labeled "Wikipedia article" 2. I already have half a dozen of AppDB entries yet to write simply because it costs me a lot of time :-( - Research a possible URL for that 10 year old app - What name to use: publisher, distributor or developer (where do I find this), what URL? E.g. Is Ubi Soft a distributor or a publisher or a reseller? Is it appropriate at all? - Scanning for duplicates - Are English and German releases different versions or different entries? - What's the original/English title (e.g. for what's known as "Backpacker" in Germany but seems to come from Sweden)? - Mindboggingly buggy JavaScript editor (adding white space here and there as well as spurious - signs at the end of texts), requiring at least 6 preview iterations. - Are my installation receipies & caveats reproducible? - Try to test and play long enough (esp. adventure games) to be able to submit thrustworthy results, before the version of wine used for installation fades out of AppDB and I'm refused the addition of the version I tested. I don't want to feel obliged to supply even more data. Currently, I'm really p.o.'ed that I cannot submit test data for older releases of Wine. I've apps that used to work and currently don't. I very much prefer an old gold entry in AppDB, along with bugzilla regressions rather than a single garbage entry for the current version, perhaps saying "this used to work in wine-1.1.18". But I digress. Regards, Jörg Höhle

1 0