port back:
0.4.15-dev-6588-g d9cd760173 [TIMEDATE] Fix the way the current time-zone is found in the list. (#5649)
which fixes regression CORE-18666 'Wrong timezone saved/displayed'
which was introduced by 0.4.14-dev-1522-g aa69236646
no increase in binary size luckily:
timedate.cpl 0.4.14rls RosBEWin2.1.6 GCC4.7.2dbg 368.128 -> 368.128
This ports back the following commits about lossless PNG-shrinking:
0.4.15-dev-6410-g cf4bb8fced [WALLPAPERS] Optimize PNG images (#5506)
0.4.15-dev-6398-g e5993f13f0 [REACTOS] Optimize PNG images (#5492)
0.4.15-dev-6376-g 8f3a09722e [MSPAINT] Optimize images for help docs (#5486)
Most of those PNGs have not been touched ever since we switched from SVN to git, so they can safely be ported back by copying over.
Only the Mizu-wallpapers had something specific to take care about:
They were added (still with a different name) by 0.4.12-dev-819-g 181a2b3359 [WALLPAPER] Add ReactOS Soft Blue
Later they were renamed by 0.4.14-dev-1435-g 6a8c894642 [WALLPAPER] Rename Soft Blue to Mizu (#1945)
and with that commit a fourth pic was added for Mizu.
Since they are only copied to iso but are not explicitly referenced by the theme, I decided
to port back the renaming part (without adding the fourth pic) to releases/0.4.12 and 0.4.13 to harmonize their name everywhere.
But did not port them back any further than where they existed already.
Tweak a few related txt/md/html files also with trivial stuff like EOL-whitespace-removal.
by porting back:
0.4.15-dev-6030-g 8a7b5a9009 [SHELL32][USER32] Fix icon regression and one test (#5207) CORE-18897
It regressed by 0.4.14-dev-459-g cfdbccf937
I actually am not a huge fan of the code-flow-splitup and all those ifdefs, but it fixes the regression.
Therefore I will not port that back further than releases/0.4.14 for now.
------------
Also partially pick 0.4.15-dev-3642-g 83be315abf [SHELL32] Use wide char string literals. (only the part in CFSFolder.cpp)
I might pick the other parts later.
------------
Binary size remains under control:
0.4.14 shell32.dll GCC4.7.2 RosBE2.1.6 dbg-cfg 9.345.024 -> 9.345.024
0.4.14 user32.dll GCC4.7.2 RosBE2.1.6 dbg-cfg 1.448.448 -> 1.448.448
This effective one-liner-backport fixes flickering in some apps like:
- CORE-7652 Hyperlink Blinking text while installing vc2008 redist package
and many duplicate-tickets of that same symptom, like:
- CORE-18553 CPUID CPU-Z 1.84 "about" flickers
- CORE-7753 Caesar 3: Autorun.exe flickers endlessly
- CORE-10741 API monitor high CPU usage
- CORE-15300 Worksoft Navigator 0.985 installer is locked in an endless redrawing loop
- CORE-16325 Some of MWSnap's installation windows fail to display content
- CORE-17925 MS Visual C++ 2008 Redistributable Setup takes long time
- CORE-18153 BitComet 1.86 forces the shell to a endless redrawing
- CORE-18434 Ubisoft installers (SetupUbi) - continously refresh of window
- CORE-15546 Shed 1.0 from rapps - setup does not draw some dlgs initially, but redraw can be enforced
- CORE-18582 Emule 0.50a - Interface frozen / flickering until window is moved
And since I do also port back the follow-up-commit of that, we can achieve that without introducing
the following regressions which the initial fixes caused:
- CORE-18817 [NTUSER] Winpos.c modification generates bad first drawing of some elements
- CORE-18830 Explorer Drawing glitch: Missing "X" on "Folder" pane when Right Click > Explore from Desktop icons
Those 2 things did not happen in the older releases beforehand.
In sum this is a squashed backport of:
0.4.15-dev-6507-g 49dbc8fbbb [NTUSER] Whitespace-fixes winpos.c
0.4.15-dev-5684-g 57eb2a6bf4 [NTUSER] Missing "X" on "Folder" pane when Right Click > Explore from Desktop icons & Rapps toolbar errors (#5068) CORE-18830, CORE-18817
the superseded 0.4.15-dev-5618-g 8866b9d7b0 Update win32ss/user/ntuser/winpos.c
the superseded 0.4.15-dev-5617-g 7866eb2886 [NTUSER] Fix winpos bug hyperlink&images loop redrawing CORE-7652 (#4995)
To prevent increased binary size for the added conditions in the code in older branches, strip a single good-path-TRACE()
which had no trailing \n in the older branches,
but where that \n was added in master head meanwhile. So that TRACE was a diff to older branches anyway.
Don't call ClientToScreen in handling NM_RCLICK message because NMMOUSE.pt is already in screen coordinates.
CORE-18966
A backport of:
0.4.15-dev-6298-g b4e946f5b0
and strip EOL whitespace
Ports back:
0.4.15-dev-5910-g 99e5f51d5b
in a slightly adapted form, which respects that we were still
a "Team" in 0.4.7 - 0.4.14 and no "Project" yet.
The idea with the "Project" was born during 0.4.15-dev'ing.
That existing difference does also manifest in media/inf/*.inf and I don't intend to
ever change that for the older branches.
That means also that in the older branches we do not need to scroll at all to
see all the text.
Also adopt ReactOS birth year from 1998 to 1996 which we do also use in the msgina startup-dlgs-pics
all the way from releases/0.4.7 to todays master head.
Actually I was working on a much bigger commit in the same context for releases/0.4.7
when I realized that related glitches do also affect master head
and also other release-branches in some areas.
this ports back (also) the following commit:
0.4.15-dev-6308-g 2d9877470d
0.4.15-dev-6256-g 8ef47d2e5e [NETSTAT] Optimize a bit (#5405)
partially 0.4.15-dev-6245-g 5ee97b9537 [NETSTAT] -b flag implies -o flag on Windows XP/2003. CORE-19006 (#5377)
partially 0.4.15-dev-6211-g 40864bc15c [NETSTAT] Fix crash when parsing the protocol CORE-19005 (#5363)
by porting back:
0.4.15-dev-6231-g a5360f542f [IPHLPAPI] Formatting only, no functional change
partially 0.4.15-dev-4492-g 03422451b3 [IPHLPAPI] Add a missing \n in an ERR() logging within ipstats_reactos.c getNumRoutes(). And strip a good-path TRACE in return in the same func.
fix CORE-18328 'FileZilla 3.8 unable to list content of a remote directory due to undue ECONNRESET'
I tested to list the contents of ftp.heise.de and downloaded some files from its pub.
by porting back:
0.4.15-dev-5870-g 9f9b81e396 [AFD] ReceiveActivity: Don't return STATUS_FILE_CLOSED in case of FCB overread CORE-18328 (#4972)
this fixes:
CORE-18429 Explorer should not have duplicated menus in the File menu
It regressed by
0.4.14-dev-955-g1cf564c 1cf564c25f
And we do now mention here CORE-9467
(and its duplicate CORE-11797 also) to make us
aware there is still something to do in this context. The guilty rev
aimed to fix those. But it wasn't the right thing to do.
by porting back the revert:
0.4.15-dev-6039-g 0fa4edebd9 and addendum
0.4.15-dev-6083-g d3164a0f55
Prevent the window from being destroyed twice
fixes CORE-18821 'WIN32K BSOD 0x1E when opening Build menu in AVR Studio 4.19'
port of 0.4.15-dev-5938-g 77d4653164
And port back also some unrelated formatting improvements,
and kill a block of code that was commented out in all branches,
and was not x64 compatible in 0.4.7/0.4.8, and therefore was a diff already anyway.
Optimize IOCTL_CONSOLE_FILL_OUTPUT_ATTRIBUTE and IOCTL_CONSOLE_FILL_OUTPUT_CHARACTER for speed by using cache.
a part of CORE-18838
It is faster because it minimizes the amount of pointer dereferences within the for-loops.
fix picked from 0.4.15-dev-5881-g 3526ffc094
0.4.15-dev-5985-g 31fdaca8c0 [REGEDIT] Strip redundant include, minor formatting (#5264)
0.4.15-dev-5970-g 9d7d3314b3 [REGEDIT] Check whether pszSelectKey is NULL on CDN_FILEOK (#5255)
0.4.15-dev-5898-g 84e580b67e [REGEDIT] Fix ListView selection and finding (#5150)
(superseded) 0.4.15-dev-5802-g cce3eb9393 Fix buffer read past end problem. (#5146) CORE-15896 and CORE-18230
partially 0.4.15-dev-5753-g 1ee9ea4518 [REGEDIT] Avoid buffer overflow in SelectNode. CORE-18602 (I left the wscpy-replacement at the end for the fixed size strings, as they would increase binary size, and are totally safe)
partially 0.4.15-dev-5308-g 5daf5cd057 [REGEDIT] Use shell icons for the tree view (#4841) CORE-10764
0.4.15-dev-4850-g 568383c9b9 [REGEDIT] Fix issue at Find registry key (#4341)
partially 0.4.15-dev-2265-g bebdfda8b7 [REGEDIT]... Use newer file open dialog (#3571)
0.4.15-dev-1470-g 9de8787187 [REGEDIT] Display search finished messagebox when using Find Next (F3) menu item (#3371) CORE-17368
0.4.15-dev-1302-g 6aae5f4fdd [REGEDIT] Make "finished find" messagebox owned by Regedit window (#3369) CORE-17367
0.4.15-dev-169-g 0461de33c5 [REGEDIT] Fix HeapFree() on the wrong variable (#2736)
a few fragments only of 0.4.15-dev-5304-g 36a7f0dc7c [REGEDIT] framewnd.c resize_frame_rect(), main.h strip externs before func declarations and reordering, treeview.c InitTreeViewImageLists()
also port back a few selected translation improvements
This is a temporary workaround to suppress the last remaining 4 compile-time-warnings for GCC4.7.2 dbg-cfg:
WARNING: /reactos/dll/win32/rsaenh/rsaenh.spec line 26: Exported symbol 'DllRegisterServer' should be PRIVATE
WARNING: /reactos/dll/win32/rsaenh/rsaenh.spec line 27: Exported symbol 'DllUnregisterServer' should be PRIVATE
WARNING: /reactos/dll/win32/msxml3/msxml3.spec line 10: Exported symbol 'DllGetClassObject' should be PRIVATE
WARNING: /reactos/dll/win32/ole32/ole32.spec line 137: Exported symbol 'DllGetClassObject' should be PRIVATE
The new checks were explicitly and intentionally enabled (actually by good intentions!) by:
0.4.13-dev-720-g 2becb000c8
So the muting is no proper fix. But since CORE-16380 is now unfixed for 3.5years (although there were various attempts),
I lost the hope that we will see a proper fix anytime soon.
It is more important to me to reach the "0 warnings at compile time" for the primary toolchain GCC4.7.2 dbg
for releases/0.4.13 and releases/0.4.14 than seeing these currently unaddressable warnings.
And with this commit I will reach that.
Afterwards all releases/0.4.7 to releases/0.4.14 do have 0 compile-time-warnings for their primary toolchain
GCC4.7.2 dbg-cfg RosBEWin2.1.6 and
GCC4.7.2 dbg-cfg RosBELin2.1.2 on Carrier-Releaser.
I intend to defend that state with *claws and teeth*! It took me long to get there!
When needed the checks can still be performed on all releases/0.4.7 till releases/0.4.14 on demand
by building with MSVC2010, which does these checks out-of-the-box without any additional code on our side
and it generates the same warnings for those very same modules then.
Once we will have a proper fix on master, I will reenable the check and port back the final thing.
Master-branch will not receive this workaround and will continue to warn even for GCC.
I verified my workaround to compile successfully with:
RosBEWin 2.1.6 GCC4.7.2 dbg-cfg (tested upfront)
RosBEWin 2.1.6 GCC4.7.2 rls-cfg (tested upfront)
RosBEWin 2.2.2 GCC8.4.0 dbg-cfg (tested upfront)
RosBEWin 2.2.2 GCC8.4.0 rls-cfg (tested upfront)
RosBEWin 2.1.6 MSVC2010SP1 (16.0.40219.1) dbg-cfg (tested upfront)
RosBEWin 2.1.6 MSVC2010SP1 (16.0.40219.1) rls-cfg (tested upfront)
(RosBELin 2.1.2 GCC4.7.2 dbg-cfg carrier-releaser) (I will test that after the push in a few minutes)
I did not test Clang at all.
port back:
0.4.15-dev-5901-g c37addcbf5 [CONSOLE] Harmonize IDC_LBOX_FONTTYPE height to 40 for *.rc (#5211), tweak some headers
and partially 0.4.15-dev-5503-g 5123edd731 [CONSOLE] Improve font preview and Update de-DE.rc (#4933)
and also pick a few other rc improvements here and there, mainly dupe-accelerator-fixes, and whitespace+logmuting in some .c
This mutes the logging:
(reactos/boot/freeldr/freeldr/ntldr/registry.c:381) err: Did not find sub key 'BiosInfo' (full \Registry\Machine\SYSTEM\CurrentControlSet\Control\BiosInfo)
and it was triggered on master by the combination of the following 2 commits:
0.4.14-dev-16-g f551caa840 +
0.4.15-dev-1097-g0689455bb3e66f89b64aecefc2367efd454b9c16 (which was ported back in 0.4.14-release-61-g ef399cd938 )
no older branches than releases/0.4.14 were affected, because they do lack the first commit, and got only the 2nd.
backport inspired by 0.4.15-dev-3794-g 914f14a8b9 CORE-18040 [FREELDR:NTLDR] Adjust some debug traces. RegOpenKey(): demote an ERR to WARN.
also delete some more traces that we do no longer have in master head.
0.4.15-dev-5688-g 460a01b167 [BROWSEUI] CORE-18646 Fix crash while editing a label in the folder tree, add missing break
0.4.15-dev-5686-g a777cc2cc4 [BROWSEUI] CORE-18646 Fix crash on backspace while editing a label in the folder panel (#5059)
0.4.15-dev-4497-g beefb07d18 [BROWSEUI] CORE-12804 Remove useless variable and unreachable code (#4483) and strip EOL whitespace
0.4.15-dev-5890-g 56bf3969cd [SYSSETUP] Tweak IDD_COMPUTERPAGE positioning (#5158)
0.4.15-dev-5875-g 7c755d9b84 [SYSSETUP] Delete obsolete lt-LT.rc
0.4.15-dev-5874-g c762234546 [SYSSETUP] Fix font glitch in ja-JP.rc (#5192)
0.4.15-dev-5871-g 3963d7d5f2 [SYSSETUP] Delete obsolete fi-FI.rc
0.4.15-dev-5772-g 784946db8c [SYSSETUP] Tweak resources CORE-18870
0.4.15-dev-5732-g 909b165c4a [SYSSETUP] Removed qemu credit entirely (#4971)
0.4.15-dev-5678-g cc54ecd779 [SYSSETUP] Fix a space-typo in IDS_WZD_PASSWORDEMPTY for *.rc (#5049)
0.4.15-dev-1102-g ad4c174eb6 [SYSSETUP] fr-FR: Change keyboard accelerator for Owner Organization field (#3282)
and a few other gems from other partially picked master-commits that not explicitly mentioned here.
--------
Also strip the IDC_LOCALETEXT and IDC_LAYOUTTEXT from the resources as that was only implemented in releases/0.4.15,
is even today still not properly localized there, they changed their dimensions again with that
and I don't intend to port that back anytime soon. And if I ever will, then all the resources would need to be touched again anyway.
port back:
0.4.15-dev-5785-g c766ee9ebb CORE-18874 [NETSHELL] Shrink radio buttons vertically (#5157)
and port back also a few other small RC-improvements from master here and there
0.4.15-dev-5102-g 18d40ebe81 [BOOTDATA] hivesys.inf: Remove redundant fastfat "Start" entry. The one in the [AddReg] section has the very same value.
-------
0.4.15-dev-5012-g e3f8f52f3d [BOOTDATA] Delete "Layout Id" from Japanese/Korean Keyboard Layout
Katayama explained: These registry values are for pre-installed IMEs we don't have. It is safe to remove them also in the older branches.
-------
0.4.15-dev-4352-g cfe387a9b0 [BOOTDATA] Add two missing separating commas in the ServiceGroupOrder List multi-string.
-------
PARTIALLY PICK: 0.4.15-dev-4412-g 94d144f8a5 [BOOTDATA] Fix some services groups
- AudioSrv belongs to "AudioGroup" (and not "Audio" that doesn't exist).
- lanmanworkstation belongs to "NetworkProvider" (without a space).
I dared only to pick the "NetworkProvider" and "AudioGroup" part for now. The MUP part I might do later when actually being able to retest CORE12107 "VMWare tools" myself.
-------
0.4.15-dev-1097-g 0689455bb3 [BOOTDATA] Comment out biosinfo.inf (#3266)
This allows ros to boot with its LiveCD files using MS NTLDR!
I stripped the line here, instead of commenting it as the master commit did. The file does not exist.
-------
Additionally to those backports I cleaned up some of the stuff that was previously commented out and/or broken:
- instead of following 0.4.15-dev-3322-g c6e002bfb3 to fix the Latvian kbdlv1.dll QUERTY->QWERTY typo, I just stripped those lines, as
they are commented out anyway. And were added in the broken and already commented way by 0.4.9-dev-825-g c2089d5319
We never had them active. No such file even.
- Also strip some 'translated' service strings that were just copy-pasted from the english strings section
without actually getting translated at all. The english ones will be used anyway.
It is enough to have that redundant stuff in master to allow translators to see it.
- Also strip the disabled kbdfo.dll Faeroese keyboard layout. Enough to have those commented lines in master as a TODO reminder. Was never active ever. No such file even.
- Also strip the disabled BENGALI_INSCRIPT kbdinben.dll. Where the existing lines even had bugs in the older branches as proven by 0.4.15-dev-3322-g c6e002bfb3
It was always disabled and was neither localizable, nor did it have the correct layout code even. No such file even.
None of those changes will introduce any additional diff or break any existing features.
They just slightly move the already existing diffs position in time towards newer release-branches.
Format-strings, but no args.
Those were fixed during 0.4.15-dev'ing on master.
Just touched this branch, because I want to update some
older branches and wanted to take that with me as well.
VC1 Linker fills BSS.SizeOfRawData with data in IMAGE_OPTIONAL_HEADER.SizeOfUninitializedData but keeps PointerToRawData to zero.
Fixes CORE-18797 and some duplicates of that. (Apps, e.g. Doom95, compiled with Watcom-C-compilers wouldn't run otherwise.)
which regressed by
0.4.10-dev-136-g c7eb46d9fd
fix picked from
0.4.15-dev-5589-g 0d4a8d0ea6
----------
Port back also some unrelated formatting-improvements within section.c, e.g.: return(xyz) -> return xyz
A *partial* port of
0.4.15-dev-1793-g b7eb0fddf3
Win+Left, Win+Right, Win+Up, and Win+Down can be disabled by registry value WindowArrangementActive REG_SZ 0 in
HKEY_CURRENT_USER\Control Panel\Desktop
Snapping mouse can be also disabled.
by porting back:
0.4.15-dev-5703-g d04e148d1c Optimize window snap disabling (#5061)
0.4.15-dev-5641-g ea55101aad Allow Window Snap to be disabled (#5014) CORE-16379 (#5014)
Also port back some minor unrelated improvements:
0.4.15-dev-3341-g 4bc95812fd defwnd.c Mark default as unreachable.
0.4.15-dev-316-g 938df97b54 nonclient.c NC_IconForWindow(): Remove redundant if() (#2842)
and:
-defwnd.c: delete a malformed TRACE() which had a format string, but no argument.
This is a true unit test and runs the same code on Windows as it does on
ROS, so this uses unconditional todos rather than todo_ros.
This avoids test failures on Windows and hopefully makes things less
confusing.
Those tests were once introduced by 0.4.14-dev-1383-g9bcd835 9bcd835074
fix ported back from 0.4.15-dev-5649-g d5999c6b66
It regressed in this branch by 0.4.14-release-48-g e5c9a50c57
( in master it was by 0.4.15-dev-5613-g 7a17c7d9ad )
It was the only place within the ros sources where message WM_SHOWSTATUSDLG
was sent also from code. And therefore the only one that needed to be updated as well.
Fix picked from:
0.4.15-dev-5712-g a5cd42c1ea
by porting back:
0.4.15-dev-5676-g 5c22ce3742 [SHELL32] Ignore unused flag SEE_MASK_UNICODE (#5033)
to mute logspam of the kind
fixme:(dll/win32/shell32/shlexec.cpp:1934) flags ignored: 0x00004000
e.g. when starting Rapps.
---------------
0.4.15-dev-2957-g c30b1fe893 de-DE.rc Update regarding IDC_BROWSE_FOR_FOLDER_NEW_FOLDER,
and pick other small tweaks in that file from master
---------------
Then delete an unused #ifndef __REACTOS__ block within brsfolder.c and strip all EOL-whitespace in that file.
This part is done as the file is part of a backport into even older branches. It is a partial pick of:
0.4.15-dev-3642-g 83be315abf
brsfolder.c will be 100% in sync between releases/0.4.7 - releases/0.4.14 afterwards.
It regressed by the introduction of the SendTo-menu in:
0.4.14-dev-238-g 5308a606dc
Since then we had some unintended verbs in the start buttons context menu.
Fix that by *partially* porting back:
0.4.15-dev-5611-g 8e036fe74d [SHELL32] Improve StartButton context menu (#5008)
0.4.15-dev-5614-g f156f9a137 [SHELL32] Follow-up
I picked only the parts that I needed to hide the CSendToMenu, because the older branches do not
have the CCopyToMenu, CMoveToMenu. And I do not intend to ever add the new source files implementing them.
The symptom was unhidden by (regression from users perspective):
0.4.12-dev-389-g 5f2bebf7a5
by porting back:
0.4.15-dev-5608-g afb953ae4d [BUTTERFLIES] Do not release the DC passed to OpenGL. CORE-18498
0.4.15-dev-5607-g 8b19b6ec5a [GDI32] Fix memory leak in case of public DC. CORE-18498
For now I will not port it back further than releases/0.4.12
although the code in both modules might be suboptimal even in earlier release-branches.
But we see no symptoms in those yet.
Prevents a BSOD 0x1E upon dereference.
CORE-18091 CORE-18558
By partially porting back:
0.4.15-dev-5587-ga999c43 a999c43746
and take also some cosmetic changes from master head, like EOL-whitespace-removal.
by porting back:
0.4.15-dev-5588-g e89675768b [SHELL32] Add 3 accelerators for IDM_DRAGFILE MENU de-DE.rc (#5001)
pick a single line only of 0.4.15-dev-984-g a5a30fc249 [SHELL32] CFSDropTarget.cpp ERR->TRACE to mute logspam
and strip some EOL-whitespaces in 3 related source-files thats contents I will port back into even older branches, like it was done in master.
by porting back 2 fixes:
0.4.15-dev-5355-g 0484beb04b [NTUSER] Don't redraw scrollbar if it is hidden (#4849) CORE-18593
0.4.15-dev-3875-g 977c129f33 [USER32] SetScrollPos should not bypass UserApiHook (#4372) CORE-18048
Both regressions were caused/unhidden by 0.4.14-dev-1134-g 00adb1a3f9
Switches from "FreeMono" and "Dejavu Sans Mono" to "Courier New" and "Lucida Console"
and therefore brings us closer to 2k3sp2 stock fonts. Saves some RAM also.
Deletes some "Ubuntu" font left-overs in some Themes.
CORE-18605 (and partially CORE-16264)
By porting back the following commits:
0.4.15-dev-5362-g c47758b102 [SETUP][INF][MEDIA] Get rid of the last Ubuntu references CORE-18607 (#4871)
0.4.15-dev-5336-g bd40c13f6f [SETUP][BOOTDATA][FONTS][INF] Delete DejaVu Sans Mono and re-map (#4856)
0.4.15-dev-5335-g a4193ade03 [SETUP][FONTS][INF][GDI32_APITEST] Delete FreeMono and re-map (#4852)
0.4.15-dev-5324-g d035452148 [SETUP][INF][ROSTESTS] Delete font substitutions and tests for "DejaVu Sans" (#4829)
by hand!
Motivation is to fix the following warnings at configure-time for *MSVC x86 dbg* :
-- umandlg has no base address
-- hotplug has no base address
-- qcap has no base address
-- mydocs has no base address
-- fusion has no base address
-- fusion_1_1 has no base address
-- fusion_2_0 has no base address
-- storprop has no base address
-- localspl_apitest.dll has no base address
This fixes CORE-14831 for *MSVC dbg* for this release-branch at least.
Meaning: It hides the effect, without fixing the cause.
It is not a sustainable solution for master yet.
We'd need to improve the python script for that.
I tested afterwards the bootup of a build configured via:
-DENABLE_ROSTESTS=1 -DENABLE_ROSAPPS=1
and built with:
MSVC 2010SP1 (16.0.40219.1) x86 target in dbg configuration
and verified that we had no relocation happening until 3rd stage desktop VBox4.3.12.
MSVC 2010SP1 (16.0.40219.1) x86 target in *rls* configuration was not
tested yet, as that up to now did never reach the desktop yet for unrelated reasons.
I will not upload an iso to sourceforge after this commit,
because it will not have any impact for our primary GCC 4.7.2 toolchain.
Based on patch by I_Kill_Bugs.
Fixes the regression which was introduced to master by 0.4.15-dev-1126-g 58b0558f94
fix picked from 0.4.15-dev-5532-g ca9ded7af8
WS_EX_NOACTIVATE flag forbids the window to be activated
fix picked from 0.4.15-dev-5137-g 826bd41d88 (1 of 3 commits for CORE-18417)
My reason for picking it is that it fixes CORE-13257 'Reboot leaving Evernote 5.9.8 in tray causes WIN32K BSOD 0x50'
I also had to pick together with it the
0.4.15-dev-5491-g 1fbed1710e to prevent it from introducing (+2 failures in user32:msg_focus) unittests.
----------------
The above patch allows/encourages us to pick then also:
0.4.15-dev-5492-g 47f3a4e144 [OSK] Delete WS_EX_NOACTIVATE workaround code
(which referenced CORE-18528, but is not fixing this minor GUI glitch yet, it just removes workaround-code that is no longer needed,
I just mention the ticket here, because the commit on master did that as well)
----------------
----------------
The following 2 things were small *unrelated* improvements nearby, and I decided to take them with me:
0.4.15-dev-4323-g 9f5cde9cbe [WIN32K:NTUSER] window.c Remove broken assert in NtUserCreateWindowEx. CORE-18123 fixed
and
a tiny part of 0.4.15-dev-3476-g 55a1c29341 [WIN32K] appswitch.c: delete unused func DWORD wtodw(const WCHAR *psz)
