This fixes an OpenGL renderer issue for builds with MSYS2/Mingw64 or MSYS2/Mingw32
The issue pertains to MSYS2 adding a default manifest file (default-manifest.o) to the executable
The bug is for PC systems with GPU drivers that were not properly supported for Windows 10
systems, like Intel HD Graphics series 1st and 2nd generations. In those systems, launching a
game in ScummVM (built with MSYS2/Mingw) with the OpenGL renderer would cause the game
screen to be a white blank image, and various warnings would be output to the console, eg.
"WARNING: GL ERROR: GL_INVALID_ENUM on glTexSubImage2D(0x0DE1, 0, 0, area.top, src.w, area.height(), _glFormat, _glType, src.gere.cpp:167)!"
This was due to MSYS2/Mingw builds trying to load the (poorly supported) GPU driver while advertising support for Windows 10 in their
embedded default Manifest file. Hence, the GPU driver DLL (eg ig4icd64.dll) would be unloaded, causing the bug.
More information is available in the following links:
https://github.com/pal1000/save-legacy-intel-graphicshttps://github.com/LWJGL/lwjgl/issues/119https://github.com/msys2/MSYS2-packages/issues/454https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69880
Credits to sluicebox for the VS GenerateManifest flag
Calling reflowLayout() not only updates the layout (as setupCloudTab()
was already doing), but also recalculates scrollbar. The issue was that
old layout didn't need a scrollbar, but updated layout did. But,
ScrollContainer was not notified, and thus scrollbar didn't appear
(until user tried to reselect the Storage via popup).
If user doesn't have a "saves" folder, listing it as a first step of
syncing save files would result in 404 from OneDrive.
OneDriveTokenRefresher handles token-related errors (401), so when it
meets 404, it calls its finishError method. But because there was some
strange behaviour from OneDrive with sending invalid JSON, this method
tries fixing JSON and parsing it again. If it is valid, it calls
non-error method again, and in result we get stack overflow.
In order to fix that, I've added a non-JSON prefix "<irrecoverable>", so
finishError won't be able to parse JSON and thus won't call finishJson
again. Saves syncing callback does check string contents apart from
trying to parse JSON, so it still works and correctly handles the
situation when "saves" directory is missing. But, if needed, code can be
updated to search for the prefix I've added and remove it before parsing
original JSON.
The warning emitted here was due to debug() call with an empty format
string, so this could have been fixed by replacing this with "%s", "".
However, this change should be better since it avoids the duplication
of the number of file check over several lines and reduces the string
duplication i.e. DRY improvements.
This comment for fall through to prevent the compiler warning about
this has to be at the point of fallthrough i.e. outside the grouping
brackets to be effective.
Well, it ain't a fix, because it's not exactly correct for any of the
themes. Yet it's the best for all of them. If I put what seems to be
correct, "modern" theme gets ruined, because it has this mystical 2px
offset in tabs/scrollcontainers.
- removed +1px in ListWidget, added in lordhoto's 2007 commit 68eb28a
(aka r29971 in svn) `Fix for bug #1670082 "GUI: Modern theme gfx glitch
in launcher".`, because it made clip this last line of scrollbar in all
themes, which doesn't look good. In 2007 theme was written in .ini,
which is not the case now. I don't see any glitches after removing this
"fix";
- fixed how scrollbar top and bottom scroll buttons are drawn in
ThemeEngine::drawScrollbar: there were these weird magic numbers, but in
reality extra space that buttons should occupy is hardcoded in
scrollbar.cpp (ScrollBarWidget) and is just +1px.
Previously, the PNGDecoder would always convert images that
have a palette with a transparent color(s) to a full RGBA
surface automatically. There needed to be a way to diable
this and keep the image paletted for the Glk engine, since
some Infocom V6 game Blorb files reuse the palettes from
previous images, so I couldn't have the decoder using the
dummy palette that comes with the image
The new arguments are optional. The lines segments will be close to the same width.
The algorithm is similar to the one we had for Blade Runner but not exactly the same, since that one would wrap a line at a white space after the theoretical split point (quotient of full line text width divided by target lines number)
