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

Call for help with the new WineTestBot
by Francois Gouget June 16, 2013

June 16, 2013

So while the translators are busy updating Wine's translations, what better time than a freeze for developers to fix Wine's conformance tests, particularly on the new WineTestBot: http://newtestbot.winehq.org It does not have as many VMs as the old one but it does have some interesting new configurations: * Two Windows 8 VMs: 32 and 64 bits. These have a lot of failures. * A Windows 7 VM with IE 10 (and one with IE 9 for comparison). * A Vista VM configured for Arabic, a right-to-left language. There's also a Japanese Windows 7 VM. * Two VMs with no sound card (the 64-bit Vista and Windows 2008 ones). These new configurations have uncovered quite a few test failures. But the more traditional configurations have some test failures too which is annoying and warrants investigation to figure out what's wrong: otherwise there will be trouble when patches have to go through the new WineTestBot. So I'm hoping some kind souls (or bored developers) will take it onto themselves to chime in and fix some of these test failures, or figure out what is causing them. Here are some known issues: * Bug 24631 - listview NM_HOVER tests sometimes fail http://bugs.winehq.org/show_bug.cgi?id=24631 * Bug 33718 - comctl32:propsheet Add button test failure http://bugs.winehq.org/show_bug.cgi?id=33718 * Bug 33719 - comctl32:propsheet custom window proc test failure http://bugs.winehq.org/show_bug.cgi?id=33719 You can also find test results from the new WineTestBot on the WineTest site: look for tags starting with 'newtb-'. http://test.winehq.org/data/ And for more information on the VMs configuration and some other known test failures go to this Wiki page: http://wiki.winehq.org/WineTestBotVMs Note that the VMs are missing some Windows dlls (atl100, localspl, msvcp100, qmgr). Information on the best way to get those is welcome. -- Francois Gouget <fgouget(a)free.fr> http://fgouget.free.fr/ Stolen from an Internet user: "f u cn rd ths, u cn gt a gd jb n cmptr prgrmmng !"

1 0

Re: [PATCH 3/3] wined3d: Handle pre-transformed vertices in the GLSL vertex pipe.
by Stefan Dösinger June 15, 2013

June 15, 2013

Am 14.06.2013 um 09:07 schrieb Henri Verbeet <hverbeet(a)codeweavers.com>: > + shader_addline(buffer, "if (gl_Vertex.w != 0.0) gl_Position /= gl_Vertex.w;\n"); I think you have to disable the rhw divide in drawStridedSlow in case the application falls back to it despite this patch, e.g. because of specular fog or float16 emulation.

1 0

Re: gdiplus: fix a clipping regression in GdipDrawString (try 3)
by Dmitry Timoshkov June 14, 2013

June 14, 2013

Rafał Mużyło <galtgendo(a)gmail.com> wrote: > Going on an assumption, that 0 is a magic value (standing for 'point'), > compare against it. Fixes 6ab04040e52465e77558a067309e8a54bdc0f32c regression, > so would be nice if this got into wine 1.6. It would be nice to see some test cases first. -- Dmitry.

1 0

Fwd: [GSoC Mentors Announce] GSoC 2013: Mentor Summit Travel and Details
by Austin English June 14, 2013

June 14, 2013

FYI for GSoC mentors. It's a ways away, but have the date in mind if you'd like to attend. Remember we have two slots, so if there are more than 2 people, we'll give preference to people that haven't gone before. -Austin ---------- Forwarded message ---------- From: Carol Smith <carols(a)google.com> Date: Thu, Jun 13, 2013 at 11:23 AM Subject: [GSoC Mentors Announce] GSoC 2013: Mentor Summit Travel and Details To: gsoc-mentors-announce <gsoc-mentors-announce(a)googlegroups.com> Hi everyone, The Google Summer of Code Mentor Summit is approaching! When: 19-20 October, 2013 Where: Google Headquarters, Mountain View, California Some brief background: The Google Summer of Code Mentor Summit is a two-day "unconference" [0] that brings together two mentors (or org admins) from each of the mentoring organizations that has participated in GSoC that year. Google pays for a two-night hotel stay, transportation costs, conference facilities, and food for the weekend, and the attendees make the agenda. We've consistently heard from mentors over the years that the summit was one of the best conferences of the year for them. The opportunity to meet face-to-face with other FOSS community members is a valuable experience, as is the time spent discussing ideas in a casual forum where everyone's voice can be heard. I'd encourage your organization to send delegates to the summit even if you don't quite understand of the utility of it now - I feel confident you won't be disappointed with the results. Attendees stay at the Wild Palms hotel [1] while they're in town and we serve a traditional dinner of thai food on Friday night and have a pool party on Saturday night. Here's how to proceed with booking and attending: 1) Your organization will need to decide on two delegates [2] to send to the event. Any people who want to attend above and beyond the two from your org will need to list themselves on the Waiting List [3] and wait for word from me about extra space as we get closer to the event. NB: The wiki requires a login when visiting for the first time. The username is "mentor" and the password is "yourock." I highly recommend creating your own account with a unique password once you have logged in for the first time. 2) Once your organization has decided on the two delegates it would like to send to this year's event, those individuals can feel free to book their hotel rooms at the Wild Palms [4] now. Please note that if the Wild Palms fills up you will need to make your reservation with the Domain Hotel instead [5]. If you really want to stay at the Wild Palms, make sure to book early! We'll have shuttles running between the hotels and to campus for the whole weekend. NB: The hotel registration deadline is 6 September. If you really want to stay at the Wild Palms, I'd recommend you book early. :-) 3) Please fill out the Mentor Summit Preparation form [6] as soon as you have confirmed your hotel room. 4) We include the reimbursement for travel to the mentor summit and the mentor stipends in one PO to your organization. Please see the payment info page [7] for more detail on this. These POs will be issued to your org soon after it has made the payment request. 5) Flights should be purchased once you have a confirmed PO. All orgs pay for plane tickets ahead of time and then request reimbursement via invoice against the PO once you have receipts for the flights. 6) Please check out the wiki [8] and fill in your name if you're attending. Look around the rest of the site - it has lots of useful information for you about the event and what to expect. 7) In the unlikely event that your organization does not plan to attend the mentor summit at all, please fill out this form [9] to let me know. Please only fill this out if you are not sending *any* delegates to the summit. 8) We are doing session scheduling for this year's summit via Google Moderator. Please fill in your ideas for sessions on the webpage [10] by 27 September and vote for your favorite ideas! You can read more about how we're running session scheduling this year on the wiki [11]. 9) Please remember that any organization that misses two or more evaluation deadlines may not attend the summit. If you book travel and then get uninvited from the summit after the either the midterm or the final evaluations you will not be reimbursed for travel costs and your hotel stays will be cancelled. So don't miss those evaluations or you jeopardize missing the event! [0] - http://en.wikipedia.org/wiki/Unconference [1] - http://www.jdvhotels.com/hotels/siliconvalley/wild_palms [2] - A "delegate" is defined as either a mentor or org admin who participated in this year's GSoC program. We're sorry, but we can't accomodate other people from your organization who did not mentor or administer this year. Also, we will not accept any mentors who miss an evaluation deadline in this year's program. Please contact me if you have questions about if a person from your organization is on this list. [3] - https://docs.google.com/a/google.com/forms/d/1bfuIJ_U76rznjfqBYdRKfFlfmvhqr… [4] - https://www.jdvhotels.com/book/reservations/wld?start_date=2013-10-14&end_d… [5] - https://www.jdvhotels.com/book/reservations/dom?start_date=2013-10-18&end_d… [6] - http://goo.gl/oyM2y [7] - http://www.google-melange.com/gsoc/document/show/gsoc_program/google/gsoc20… [8] - https://gsoc-wiki.osuosl.org/index.php/2013 [9] - http://goo.gl/OTxuo [10] - http://goo.gl/EH0ZH [11] - https://gsoc-wiki.osuosl.org/index.php/Proposed_Session_Topics_2013 As always, please contact me directly with any questions or concerns you may have about any of these instructions. Cheers, Carol -- You received this message because you are subscribed to the Google Groups "Google Summer of Code Mentors Announce List" group. To unsubscribe from this group and stop receiving emails from it, send an email to gsoc-mentors-announce+unsubscribe(a)googlegroups.com. Visit this group at http://groups.google.com/group/gsoc-mentors-announce. For more options, visit https://groups.google.com/groups/opt_out. -- -Austin

2 1

Re: oledb32: Add IErrorInfo Support
by Nikolay Sivov June 14, 2013

June 14, 2013

On 6/14/2013 14:12, Alistair Leslie-Hughes wrote: > +typedef struct ErrorInfoImpl > +{ > + IErrorInfo IErrorInfo_iface; > + LONG ref; > + > + GUID m_Guid; > + WCHAR *source; > + WCHAR *description; > + WCHAR *help_file; > + DWORD m_dwHelpContext; > +} ErrorInfoImpl; Please use consistent naming, for example avoid prefixed names. Also it would be cleaner to use BSTR here, but it will work either way. > + TRACE("(%p)->(pBstrSource=%p)\n",This,pBstrSource); No need for parameter names in traces.

1 0

Re: ole32: Postpone removing a being destroyed apartment from the global list.
by Alexandre Julliard June 13, 2013

June 13, 2013

Dmitry Timoshkov <dmitry(a)baikal.ru> writes: > StdMarshalImpl_ReleaseMarshalData called from apartment_release() needs to > access apartment data, but fails because an apartment has already been removed > from the global list. That leads to lost connections and memory leaks. But now you can have someone find a half-destroyed apartment in the list... -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [PATCH] [cmd, try2] move filename . deletes file
by Alexandre Julliard June 13, 2013

June 13, 2013

Jason Edmeades <jason(a)edmeades.me.uk> writes: > move is not tollerent of the source and destination being the same > and happily removes the destination (after prompting) ready to > move the file, only to appear shocked that the file is no longer > there... > > This was sent originally where I compared filenames, and was given > feedback to use a better system of equality. This now uses something > the MSDN implies is the best way to tell for equivalent destinations This should be done in MoveFile, and cmd should use that instead of duplicating the checks. -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: oleaut32: Add DECIMAL tests
by Marvin June 13, 2013

June 13, 2013

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=25961 Your paranoid android. === W2KPROSP4 (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === WXPPROSP2 (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === WXPX64 (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === W2K3R2SESP2 (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === WVISTAX64 (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === W2K8SE (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === W7PROX64 (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === TEST64_W7SP1 (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === WXPX64 (64 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === WVISTAX64 (64 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === W7PROX64 (64 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === TEST64_W7SP1 (64 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000

1 0

Re: oleaut32:
by Marvin June 13, 2013

June 13, 2013

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=25960 Your paranoid android. === W2KPROSP4 (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === WXPPROSP2 (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === WXPX64 (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === W2K3R2SESP2 (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === WVISTAX64 (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === W2K8SE (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === W7PROX64 (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === TEST64_W7SP1 (32 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === WXPX64 (64 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === WVISTAX64 (64 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === W7PROX64 (64 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000 === TEST64_W7SP1 (64 bit vartype) === vartype.c:4339: Test failed: expected (0,0,-1,(ffffff84 ffffffff)), got (0,0,-1,(ffffffff ffffff84)) hres 0x00000000

1 0

Re: gdiplus: fix a clipping regression in GdipDrawString (try 2)
by Vincent Povirk June 12, 2013

June 12, 2013

What if rect.Width is 0.25? Shouldn't we still clip then?

1 0