Remove WM_SYSKEYDOWN handling at component level, in consistency with other components (ListView,...)
Early embodiement of the fix proposed to WineHQ:
https://bugs.winehq.org/show_bug.cgi?id=49097 in order to remove functional limitation in ReactOS.
Fixes regressions CORE-17020 and CORE-12203 that were once introduced
by SVN r72320 == git 297e33f228
during COMCTL32 WineSync 1.9.16 CORE-11866
Fix picked from 0.4.15-dev-191-g 295ba62820
Do not call EndDialog outside the dialog procedure.
fix picked from commit 0.4.15-dev-511-g 90456e54a4
This hang/crash in 2nd stage setup did *not* affect our official carrier-releaser isos that
we host on sourceforge, because those come bundled with Gecko already and therefore always
skipped the download.
Most users were safe in the past therefore.
Still everyone that would like to build an iso by himself could potentially run
into those crashes, and it did not only happen upon user-driven cancellation,
but also often crashed upon network failure / timeout.
A very important fix therefore.
For some unknown reason the likelihood of this crash gradually increased from
0.4.7 where it happened almost never, until 0.4.14 where it crashed much more likely.
Not well understood, maybe depends on memory layout or something similar.
All of those are super safe commits. Fixes CORE-13508
-excerpt of 0.4.11-dev-909-g 07b120bd56 zh-CN [TRANSLATION] Update Simplified Chinese translation. (#1094)
-excerpt of 0.4.11-dev-149-g 31019e8c0c ro-RO [TRANSLATION] Update Romanian translation (#851)
-excerpt of 0.4.9-dev-126-g 0182dcd71b sounds.c x64 compatibility issue, possible pointer truncation
-0.4.9-dev-104-g 9e9f053eec fixes CORE-13508 fr-FR "Sound settings : narrow elements (french)"
-excerpt of 0.4.8-dev-644-g ba3f074320 tr-TR whitespace noise
-excerpt of 0.4.8-dev-627-g fe50c655aa zh-CN [TRANSLATION] Update Simplified Chinese translations and font settings. (#266)
-excerpt of 0.4.8-dev-533-g 2b91b29642 zh-CN/zh-TW [CPL] Control Panel Applets: Use correct font name in Chinese resources, CORE-9566 (#220).
mmsys.cpl binary size in bytes RosBE2.1.6 GCC4.7.2 dbg I18N=all decreases for 0.4.7 thanks to the chinese changes:
0.4.11 which is the source 741.376
0.4.10 old 741.376 patched 741.376
0.4. 9 old 741.376 patched 741.376
0.4. 8 old 741.376 patched 741.376
0.4. 7 old 741.888 patched 741.376
Syncing the module will simplify future backmerges.
CORE-16935
I merged that back, because at least the play-icon was a regression.
It regressed by SVN r68073 == git 888c2fb967
Ftr we did not always use an icon here, but instead earlier we just had some
font glyph displaying the play-symbol. It was once changed to icon to improve
display for some asian localization (japanese? iirc).
fix picked from commit 0.4.15-dev-86-g bdb4da009a
The patch also decreases binary size
mmsys.cpl binary size in bytes RosBE2.1.6 GCC4.7.2 dbg I18N=all:
0.4.13 old 786.944 patched 782.848
0.4.12 old 786.944 patched 782.848
0.4.11 old 745.472 patched 741.376
0.4.10 old 745.472 patched 741.376
0.4. 9 old 745.472 patched 741.376
0.4. 8 old 745.472 patched 741.376
0.4. 7 old 745.984 patched 741.888
Improves some strings, but actually I ported that back to reduce binary size.
The translated strings are shorter even than the english ones. Win-win!
Source was state of 0.4.12-release-31-g62822df.
desk.cpl built with RosBE2.1.6 on Windows with GCC4.7.2 debug building with I18N=all:
539.136bytes for 0.4.12-release-31-g62822df (was source)
0.4.11 before 539.648bytes after 539.136bytes
0.4.10 before 539.648bytes after 539.136bytes
0.4. 9 before 539.648bytes after 539.136bytes
0.4. 8 before 539.648bytes after 539.136bytes
0.4. 7 before 528.896bytes after 528.384bytes
This improves both: the visual quality of the dialog and also
decreases desk.cpl binary size.
Fix picked from 0.4.11-dev-244-g 6f138e2811
------------------------------
But it also inherently includes
0.4.11-dev-94-g 25f4ed75a0
"[CPL][DESK] Update monitor resource (#816)"
which replaces the CRT bmp with a TFT bitmap
"Let's encrypt" certificate issuer string changed within last 24hours remotely.
Update its name to make RAPPS stop complaining on DB update.
And also fix a bug in rapps mixing A()/W() in the
messagebox that informs about that cert check failure.
Fix is port of
0.4.15-dev-1120-g 12caaece25
0.4.15-dev-1617-g f77f5a30cf
Next time this happens I will relentlessly undefine CMakeLists.txt USE_CERT_PINNING
Fixes CORE-15558 "AbiWord font listbox should display a specific font only once, but displays each of them many times instead"
which regressed by SVN r75491 == 0.4.7-dev-32-g d10c0ec112
by porting back the commits:
0.4.12-dev-320-g 6e4e5a004c
and
0.4.13-dev-107-g ae8417fd90 to prevent introducing another regression CORE_15785 with the initial fix.
and some more addendums to prevent introducing regression CORE_15755
I verified afterwards that all 3 issues: CORE-15558, CORE_15785 and CORE_15755 are in fixed state.
Actually I don't really like the resulting code due to the qsort() that it introduces into
win32ss/gdi/gdi32/objects/font.c
We did not need that at all before 0.4.7-dev-32-g d10c0ec112
In case that qsort() would ever reveal any new trouble like reduced performance or stack overflow,
we should structurally revert all that context to 0.4.7-dev-31 instead.
IntEnumFontFamilies() is called immediately upon process initialization for many applications and the qsorted-list is *huge*,
so this indeed could be relevant for performance.
Hack has been introduced in IntSendChildNCPaint() in r73532 == 6d7fcc4
to attempt to pseudo-fix CORE-12071 and CORE-12085 while breaking other stuff,
amongst which CORE-14763 is one of the latest manifestations, and re-enable the correct original code.
This fixes painting hangs under diverse situations,
while CORE-12071 and CORE-12085 remain fixed.
When it was hanging like CORE-14763, we could not even ACPI shutdown anymore.
fixed by porting back:
0.4.10-dev-324-g 25004bb5b4
0.4.10-dev-323-g ee0511b49d
0.4.10-dev-322-g 4d057cf626
picked in squashed form from
0.4.8-dev-650-g 81eb3bbceb
0.4.9-dev-622-g c704cf5f87
I also ported that into all other release-branches >= 0.4.8 before.
The same MSVC-rls-fixed implementation.
It gets triggered several times during the setup, but also many times
when browsing with Opera 12.18 for example.
picked from 0.4.10-dev-380-g 74adcadbd0
which effectively mutes them.
Silence harmless errors, which can be reported at every boot.
And do get logged even before the "-------------" reboot-delimiter.
inspired by 0.4.8-dev-875-g b5a7765068
I decided to just switch to WARN for the backport,
but to not enlarge the strings contents/lengths.
This fixes:
- CORE-15911 "Scrollbars do not disappear when Maximizing and not needed"
- CORE-10617 "ListView corrupt scrollbar upon resizing the column-header"
and fixes regressions:
- CORE-15429 "Uninitialized scrollbars in 'My Computer' permanently drawn"
- CORE-16466 "Uninitialized scrollbars in 'My Computer' do still flash up for a fraction of a second before getting overpainted"
both unhidden by SVN r75735 == git 0.4.7-dev-168-g 6af37fd54e
by porting back the commits:
0.4.14-dev-312-g b931f643e3
0.4.13-dev-535-g 1158c24194
Many thanks to the patches author JIRA-user "blueghost"
tibor.fuzi@gmail.com
The patch fixes regression CORE-16703, which was unhidden by explorer commit
8ad0eae687 == SVN r74968
although the root-cause was missing functionality in WIN32SS.
The patch also fixes CORE-16705 and CORE-16706.
fix picked from 0.4.14-dev-1137-g 848d7ec633
----------
And while touching winpos.c I decided to also port back the
fix for the format-string-glitch from #1893 authored by Stanislav Motylkov.
fix picked from 0.4.13-dev-1079-g b16ca9cd65
This fixes a strange toggle-back-effect in 'Zeckensacks Glide Wrapper 0.84c'
by importing Wine commit
9dbe886e3f
Thanks to patches author Akihiro Sagawa <sagawa.aki@gmail.com>
Fix picked from 0.4.13-dev-1169-g 841732e45c
This was no regression, but I could not resist porting it back, because it's just
too tempting to fix a bug by deleting code. Win-win!
Addendum to 0.4.7-release-38-g ec85d9c62b
CORE-12752 was already fixed,
but using the "Motorbike"-game from CORE_16707, I could trigger a new
conflict between user32.dll and hid.dll.
Therefore move 3 potential colliders away.
CORE-15433 "Fix assertless BSOD when Opera 12.18 browses duckduckgo.com"
fix ported from 0.4.12-dev-20-g 43e2089 (I ported it back earlier into 0.4.11-RC-7-g 7d0eadd already)
CORE-14775 Allow NULL, Fix repeatedly hit assert when Opera 12.18 browses www.palemoon.org style_name==NULL
by taking parts only of the fix from 0.4.10-dev-284-g ae8762e . I left the verbosed debug-prints aside, as that is not performance-neutral.
They are both related. Actually style_name == NULL is ok, that assert was not correct.
but some of the code was just not handling it properly and dereferenced it.
This is the hard work of Mark Jansen, and in sum reflects the state of 0.4.13-rls-0
when leaving out all work of Jim Tabor and Katayama in this file.
I picked only one unrelated single-line-commit of Timo Kreuzer additionally.
In sum it fixes the following tickets: CORE-16306, CORE-16297, CORE-15863, CORE-15001, CORE-16298.
And is a backport of the following master-commits:
0.4.14-dev-40-g e1984d0 CORE-16306 "Regression, UltraISO shows Languages popup menu at wrong position" (ported back earlier into 0.4.13-RC-12-g 609f2ae)
0.4.13-dev-815-g 6948764 CORE-16297 "Regression, The popup menus position is not perfect for systray icons context menu of Abyss webserver"
0.4.13-dev-814-g 84e263c CORE-16297 "Regression, The popup menus position is not perfect for systray icons context menu of Abyss webserver"
0.4.13-dev-813-g 1c8cdaa CORE-16297 "Regression, The popup menus position is not perfect for systray icons context menu of Abyss webserver"
0.4.13-dev-793-g b5c6af4 CORE-15863 "Regression, popup-menus that expand a sub-popup-menu may erroneously overlay itself"
0.4.13-dev-792-g 7c45a64 CORE-15863 "Regression, popup-menus that expand a sub-popup-menu may erroneously overlay itself"
0.4.13-dev-791-g a59df38 CORE-15863 "Regression, popup-menus that expand a sub-popup-menu may erroneously overlay itself"
0.4.13-dev-713-g 08c6d21 just fixes a compiler warning, unrelated one-liner, only commit of Timo in here
0.4.12-dev-369-g 007ec03 CORE-15001 "Wrong placement of the context menu in systray leads to autoselection, e.g. Abyss WebServer"
0.4.12-dev-346-g d2626f0 CORE-15001 "Wrong placement of the context menu in systray leads to autoselection, e.g. Abyss WebServer"
CORE-16298 "Window system menu does not appear where expected" which is a dupe of one of the tickets above, don't ask me which exactly
There is only one ticket in this context that will NOT get fixed in the final state, but we do accept that,
as it is minor and also MS Windows is not really deterministic in all cases for that:
CORE-15733 "[Win32SS?] dropdown-menu position may be wrong, when the window overflows the screen-borders"
This patch supersedes / is a better replacement for the safe-intermediate-state-fixes that I committed earlier into:
0.4.12-RC-49-g d52cac4 & 0.4.12-RC-48-g 3113d13
0.4.11-RC-35-g a9be77c & 0.4.11-RC-30-g 83e4a61 & 0.4.11-RC-27-g 1d47cfd
It is rare, but I have seen this crash happening as well
when interacting with the taskbar quickly after the boot.
fix picked from 0.4.12-dev-203-g c526f94
Fixes accessing nullptr / invalid handle in Window->SystemMenu
I could trigger this occasionally when moving Opera 12.18 window
in front of Spotify 1.0.29.92 window.
Fix was:
authored by Thomas Faber (Thank you!)
tested by me
reviewed by James Tabor
fix picked from 0.4.11-dev-195-g ef016bfe0a
A squashed port of
0.4.13-dev-702-g 1e91a1690d
0.4.13-dev-809-g 65e4efd475
0.4.14-dev-186-g b3033b81a4
which in sum brings us to the nice state of 0.4.13-release-0-g4ee3778a4ed2d80c96e7a0f7a1e52e78d3a6568e
where I used the same trio successfully already.
CORE-13211 "RIGHT-expansion-arrow in submenus being drawn too large"
did affect many apps, e.g: UltraISO and Abiword
It was fixed by 0.4.13-dev-702-g 1e91a1690d (#1752)
but on its own it did introduce 2 regressions:
The first was that the
"upwards scrolling arrow in language selection submenu of UltraISO was now drawn much too large"
(discussion and pics also in CORE-13211)
That was fixed by the addendum 0.4.13-dev-809-g 65e4efd475 (#1753)
But we had still another regression from the first commit to solve:
CORE-16470 "The checkmarks in Abiwords menubar became black boxes"
That last regression was fixed by 0.4.14-dev-186-g b3033b81a4 (#2009)
Fixes CORE-13986 "Wise Installer draws scrollbars enabled while they should be drawn disabled"
by porting back 0.4.11-dev-586-g 929a2c6
which requires to pick also 0.4.14-dev-1167-g 8bb9434 to prevent the first fix introducing regression
CORE-16721 "Scrolling by pressing the scrollbar arrows does not draw the arrows as pressed"
and finally port back 0.4.14-dev-243-g 19fbdda to fix
CORE-15557 "disabled scrollbars being clickable"
In sum this brings the scrollbar code to the state of 0.4.13-release-0-g 4ee3778
Not to translate!
Fixes CORE-16762 "inability to pass 2nd stage setup in Albanian"
The regression was introduced by SVN r62040 == git 258cfa34a0
fix picked from commit 0.4.14-dev-1224-g
1edfddf07a
The regression was unhidden by MSI-winesync 1.5.10 almost 7years ago:
8939969e21 == SVN r57070
Finally we can fix it by adding MS enumerator extension to Jscript.
Many thanks to the patches author
Andreas Maier <staubim@quantentunnel.de>
JIRA-nick: andy-123
fix picked from commit 0.4.12-RC-18-g 9ebb42a4c0
It went afterward into master in slightly modified form as 0.4.13-dev-890-g 83dcae1
But I decided to take the older state from 0.4.12RC, so that I do not have to risk
touching releases/0.4.12 again to sync to "latest&greatest".
Due to the changed PROPF_* defines in jscript.h, I felt most safe by porting back also:
0.4.10-dev-188-g c6f49f5
0.4.10-dev-152-g 3f071cc
0.4.9-dev-248-g 8dba275
0.4.9-dev-120-g 004d830fe9 [INCLUDE/WINE] Import heap.h from Wine Staging 3.3
0.4.8-dev-722-g 3381ee4
to the state of 0.4.10-release-24-g8c2fc27 to fix
regression CORE-14009 "blank icons in livecd for desktop-lnk cmd & QuickLaunch"
which regressed by 0.4.7-dev-111-g d800840 == SVN r75607
The most relevant part came from 0.4.11-dev-11-g 6eb6aa2 which I ported back earlier already into 0.4.10-RC-13-g 3a1fe9d
(and this fixes/prevents also another regression CORE-14961 which did affect only intermediate states, in other words
was never present in releases/0.4.7 0.4.8 and 0.4.9)
Sync those 2 font-files to the same state we used in all of those:
0.4.10-release-0-g2be9148
0.4.11-release-0-ged7305b
0.4.12-release-0-g5ad5272
0.4.13-release-0-g4ee3778
First and foremost the aim of this backport is to fix
- CORE-12091 [Tahoma font/kerning] has issues when displaying letter combination "fi" & "fli"
Initially we believed that to be a regression, because its exposure to the users eyes
was massively increased by SVN r62982 == git 8552a1297e
Since then it did massively affect many RichEdit controls.
But actually the fonts itself were broken before that commit already
as the following ticket does prove:
- CORE-12172 "Tahoma "fi" & "fli" rendering issue in Firefox 28"
which was happening also beforehand and outside of a RichEdit control.
All of them and also some duplicate-tickets like CORE-10751 get fixed.
In sum this ports back the following commits:
0.4.11-dev-309-g b560491 "Fix underline/underscore position" (#892) (that I ported back earlier into 0.4.10-RC-28-g 86e0a30 already)
0.4.11-dev-3-g 0f308a0 "Improve Tahoma font (14px bitmap and ligatures)" (#809) CORE-12172 / CORE-12091 / CORE-10751 (that I ported back earlier into 0.4.10-RC-12-g 1194b0b already)
0.4.10-dev-622-g e58607e "Add Hebrew vector glyphs to Tahoma Bold font" (#790) CORE-14943
0.4.10-dev-621-g b80171e "Add Hebrew vector glyphs to Tahoma font" (#789) CORE-14943
0.4.10-dev-190-g 2355838 "Delete 16px bitmap from Tahoma Bold"
0.4.10-dev-100-g e0f5ded "Improve Tahoma Bold font (tahomabd.ttf)" (#580)
The visible symptom for us was, that the Apply button in our
desk.cpl was always visible, even in case no user input was done in the dlg yet.
A very old regression. It regressed somewhere around
SVN r42990 – Good (2009-09-03)
SVN r44397 – Fail (2009-12-04)
The fix is a backport of 0.4.13-dev-908-g
344d559935
Unchange the property sheet page after WM_INITDIALOG (#1842)
In my testing on Win2k3, EN_CHANGE is generated in OnInitDialog, that is correct.
The target is property sheet.
The property sheet does unchange the page after WM_INITDIALOG generation.
-fixes CORE-14192 Notepad text rendering performance improves dramatically
-fixes 50% of CORE-13631 GSView Setup, dramatic speedup, but only partially solved because the jumping does persist
-fixes CORE-14002 "It takes 8seconds to open up about dlg of DevCPP 5.11" reduces to 2seconds
-fixes "about-dlg of DoubleCommander takes long to open"
[USP10] Re-use script caches for the same font. CORE-14192
This significantly speeds up WM_SETTEXT in multiline edit controls.
by porting back the fix from 0.4.8-dev-779-g cd992d022f
-----------------------------------------------------------
but since that commit on its own would introduce the regression CORE-14226
I have to additionally port back 0.4.12-dev-151-g c086edf366
That final fixes commit message:
[USP10] Don't mix character and glyph index caches. CORE-14226
Import Wine commits by Nikolay Sivov:
* d1e618d31da usp10: Use glyph index consistently in ScriptGetGlyphABCWidth().
* 8d018d8d1da usp10: Make sure metrics cache is used with glyph indices only.
* eda3a3e37da usp10: Remove extra parentheses.
Fixes font display in UltraISO and GOG and many other installers.
See also https://bugs.winehq.org/show_bug.cgi?id=44410
-------------------------------------------------------------
and I have to sync the
modules\rostests\winetests\usp10\usp10.c to the state of 0.4.8-release-30-gfb23ed8
With these, we generate the WM_CONTEXTMENU and NIN_(KEY)SELECT
shell icon notifications that applications expect when they handle
shell notification icons with uVersion >= 3.
This fixes in particular the previously unresponsive icon of KVIrc 4.x,
and more generally *all* the notification icons of Qt applications.
CORE-10605 #resolve
fix picked from 0.4.9-dev-587-g 4127024b0d
This squashed commit ports back almost all improvements
we did to explorer during 0.4.8-dev.
Main motivation is:
-fixing tons of bugs painting the systray icons, e.g. (but not exclusively) CORE-11378
-and to fix a regression about clock margins CORE-14272
-and the clocks vertical text placement
-Win+E behavior CORE-14199
Here is a list of the commits I ported back:
[PSDK] shellapi.h
0.4.8-dev-1053-g 8b78b0ff98
0.4.8-dev-982-g a64c33e787
0.4.8-dev-970-g 9d2139b7d8
[SHELL32]/[REACTOS]/[EXPLORER]
0.4.8-dev-984-g 131678a025
0.4.8-dev-971-g 5ecabcce59
[EXPLORER]
0.4.8-dev-1065-g f1535c2d6e tiny part of CORE-14306
0.4.8-dev-1064-g abdde0b764
0.4.8-dev-1054-g 3003471e64
0.4.8-dev-974-g 04498f3eb5
0.4.8-dev-944-g f5d679aeb9 fixes CORE-14272 Clock Margins
0.4.8-dev-936-g 556c485c8b
0.4.8-dev-935-g d11e7af5f2
0.4.8-dev-934-g c109dc9f24
0.4.8-dev-928-g 22d1e1017f
0.4.8-dev-925-g c54c5ebb5f
0.4.8-dev-924-g 488477d10d
0.4.8-dev-923-g 85d7dd4f83
0.4.8-dev-922-g 5a1984c998
0.4.8-dev-921-g 6c073e9539
0.4.8-dev-920-g 455bce68b5
0.4.8-dev-908-g 7f35908ad2
0.4.8-dev-860-g 3d75cc0814
0.4.8-dev-853-g 5ec441a867
0.4.8-dev-852-g bc43733e48
0.4.8-dev-851-g bbca71c4a5 [EXPLORER]&[ROSCTRLS]
0.4.8-dev-789-g cb8cc0d098
0.4.8-dev-788-g be2bf9b8c5
0.4.8-dev-787-g 851e0ca5a8
0.4.8-dev-786-g 2abd068398
0.4.8-dev-785-g 8c91a00fcf
0.4.8-dev-784-g 6c9629bb72
0.4.8-dev-783-g d66c6890f7
0.4.8-dev-691-g 2c8ed37605
0.4.8-dev-690-g 74e58d938e
0.4.8-dev-669-g 4b6d929be1 fixes CORE-14199
0.4.8-dev-46-g 2ed535da8a
0.4.8-dev-21-g 82b12ef4cf
0.4.8-dev-20-g 4238f22396
For now I intentionally did leave out
0.4.8-dev-684-g 71e66c69c1
to not make explorer import oleaut32 SysAllocString()/SysFreeString().
I also did not update the translations rc files.
PR #1599 "[COMCTL32] Saturated images: Take mask-images in account."
Many Thanks to the patches author
Andreas Maier <staubim@quantentunnel.de>
JIRA-nick: andy-123
CORE-14209 IDA5free issue with disabled menubar buttons
regressed by SVN r66418 == git
8370f15813
The patch also improves the rendering of disabled menubar buttons for SumatraPDF CORE-10601.
Even after the patch we are not perfect at deciding when to draw masked
and when to draw desaturated in 32bpp, but it definitely improves end users experience.
Fix picked from 0.4.13-dev-881-g
4093681ad0
and as preparation 0.4.13-dev-880-g
49a797543f
CORE-12752 Symptom "multiple programs refuse to start because
base address conflict between MSVCP71.dll
& our user32.dll (that can't be relocated by design)."
E.g:
-Photoshop CS2
-Avira Antivir 6.35.00.243 when selecting "Start update" from systray icon
It regressed by SVN r68838 == git
a9a2119c0c
The new base-address is manually crafted to same address that I used
for 0.4.9-RC-23-g
519cdd3c52
already. It is most likely not as effective in preventing
other relocations as the proper fix that was
created by a full recalculation of all dlls base addresses later in
0.4.10-dev-338-g
80bac79070
But it worked properly for me to solve CORE-12752 and today
I do fear I would have higher chances to unhide other
memory corruption bugs if I would do the proper thing.
So unless this would prove to cause any other issue, I will
stick with this minimal invasive approach for the older releases.
CORE-14044 Symptom "Installing MS Tahoma did not longer replace Tahoma.ttf, just Tahomabd.ttf"
It regressed by SVN r73900 == git
d5f4f99b0e
Thanks to Doug Lyons for pinpointing the issue.
fix picked from commit 0.4.11-dev-311-g
3d44f79e3e
