Changes

Created an empty Pendulum window.

Replaced our data.csv file with C/C++ inlined data.

Have our Zinc scene all setup.

Added the pendulum model.

Some minor cleaning up.

Some minor cleaning up.

Some minor cleaning up.

Added a time slider.

Render the pendulum the right way up.

Changed the EOL from CR (macOS) to LF (Linux).

Some minor cleaning up.

Now show the time trace of our pendulum.

Provide access to our Pendulum window from our Simulation Experiment view.

Some minor cleaning up.

Make it possible to set some model parameters to our pendulum window.

Just calls and storage, no actual use of those parameters in our
pendulum window at this stage.

Some minor cleaning up.

Have our Simulation Experiment view depend on our Pendulum window.

Update the data of our Pendulum window whenever we get new ones from our Simulation Experiment view.

Provide our Pendulum window with some data from our Simulation Experiment view.

Although we are providing some data, we can’t seem to get the pendulum
scene to reflect it…!?

Some minor cleaning up.

Make sure that all of our Zinc scene is visible.

Moved the contents of PendulumWindowWindow::customizeZincContext() into PendulumWindowWindow::initData().

Moved PendulumWindowWindow::createAndSetZincContext() elsewhere in the file.

Slight GUI improvement.

Some minor cleaning up.

Make sure that our Zinc scene is still visible when adding data to it.

Some minor cleaning up.

Some minor cleaning up.

Some minor cleaning up.

Shorten the length of our SED-ML simulation.

Finally got things to work to a respectable level!

Small hack to get our Pendulum window to load last.

Small hack to get our Pendulum window to load last.

Some minor cleaning up.

Make sure we can build on Windows.

Make sure that we ship our pendulum model.

Remove our dependency on Qt Charts.

Don't limit ourselves to GCC/G++ 4.9.

Oops!

Merge branch 'master' into pendulum.

Make sure that Zinc is always built/retrieved and that our Zinc widget also gets built.

Merge branch 'master' into pendulum.

Pendulum window: slightly changed the colour of our axes.

The idea is that it matches the colour of our plots.

Revert "Pendulum window: slightly changed the colour of our axes."

This reverts commit c897b95ce31fc990167f0d992a20c129b393bdbe.

Pendulum window: slightly changed the colour of our axes and path.

The idea is that our path matches the (new default) colour of our plots.

Clang code model related cleaning up.

Some minor cleaning up.

Merge branch 'master' into pendulum.

Make Clang-Tidy happy.

Merge branch 'master' into pendulum.

Merge branch 'master' into pendulum.

Merge branch 'master' into pendulum.

SUNDIALS: upgraded to version 5.2.0 (#2322).

SUNDIALS: binaries for version 5.2.0 (#2322).

Slight update to our downloads page (#2322).

OpenSSL: upgraded to version 1.1.1g (#2326).

OpenSSL: binaries for version 1.1.1g (#2326).

Slight update to our downloads page (#2326).

Simulation Experiment view: fixed a couple of GUI glitches (#2332).

Slight update to our downloads page (#2332).

Oxygen library: upgraded to version 5.68.0.

Python support: make it possible to retrieve the type of a DataStoreVariable object (#2336).

Slight reordering of our CellML variable type (#2336).

Just so that the order looks "better".

Slight update to our downloads page (#2336).

jQuery: upgraded to version 3.5.1 (#2338).

Slight update to our downloads page (#2338).

libgit2: upgraded to version 1.0.0 (#2340).

libgit2: binaries for version 1.0.0 (#2340).

Slight update to our downloads page (#2340).

Mesa: upgraded to version 20.0.7 (#2342).

Mesa: binary for version 20.0.7 (#2342).

Slight update to our downloads page (#2342).

SUNDIALS: upgraded to version 5.3.0 (#2344).

SUNDIALS: binaries for version 5.3.0 (#2344).

Slight update to our downloads page (#2344).

libgit2: upgraded to version 0.28.5 (#2346).

libgit2: binaries for version 0.28.5 (#2346).

Slight update to our downloads page (#2346).

PMR support: ported (back) our libgit2-related code to version 0.28.5 (#2305).

libgit2: upgraded to version 1.0.1 (#2348).

libgit2: binaries for version 1.0.1 (#2348).

Slight update to our downloads page (#2348).

Some minor cleaning up.

Some minor cleaning up.

Some minor cleaning up.

Some minor cleaning up.

Some minor cleaning up.

Some minor cleaning up.

Some minor cleaning up.

Some minor cleaning up.

Some minor cleaning up.

Some minor cleaning up.

Some minor cleaning up.

Some minor cleaning up.

Some minor cleaning up.

Some minor cleaning up.

New snapshot.

Qt WebKit: new macOS binary.

Not sure what happened, but our previous binary wasn't working on macOS 10.13, only macOS 10.14 and 10.15...!? (macOS 10.13 claims that Qt WebKit was built for macOS 10.15.)

Qt WebKit: new Linux binary.

On Ubuntu 20.04, we now need access to libffi.so.6!

Linux: libffi.so.6 is needed by Python (well, NumPy), not Qt WebKit.

Yet, there is no point in shipping it since it's not one of "our" files that needs libffi.so.6. Our solution may be to upgrade NumPy. Will have to see...

Qt WebKit: new Linux binary.

This is a revert of commit b45dfd332 where we built Qt WebKit with a special RPATH value, which we don't need.

OpenSSL: new macOS binary.

Not sure what happened, but our previous binary wasn't working on macOS 10.13, only macOS 10.14 and 10.15...!? (macOS 10.13 claims that OpenSSL was built for macOS 10.15.)

Unneeded plugins.

At least, they don't seem to be needed anymore.

Revert "Unneeded plugins."

This reverts commit d785793f64abd650721aa31d4ccb231809e19d8a.

macOS: support macOS Mojave 10.14.4 and later (#2354).

Some minor cleaning up.

Linux: deploy libffi.so.6 (#2353).

macOS: support macOS Mojave 10.14.4 and later (#2354).

QtWebKit: reverted back to version 5.212.0 Alpha 3 (#2359).

Build system: no need to set `PYTHON_SUPPORT` since we always have Python support (#2358).

New snapshot.

Properly retrieve OpenCOR's date.

Oops, at some point we added an "\n" to the version file, which means that retrieving the last item returned an empty string...!

Mesa: upgrade to version 20.1.2 (#2363).

Revert "Build system: no need to set `PYTHON_SUPPORT` since we always have Python support (#2358)."

This reverts commit 1e7a08f264dd449a1752306ed6837fcae18e763f.

Indeed, we STILL need to know whether we are building OpenCOR with Python support, in case we need to build ONLY a third-party library (e.g. Mesa).

Mesa: binary for version 20.1.2 (#2363).

QScintilla: upgraded to version 2.11.5 (#2364).

QScintilla: binaries for version 2.11.5 (#2364).

Slight update to our downloads pages (#2364).

Qwt: upgraded to version 6.1.5 (#2365).

Qwt: binaries for version 6.1.5 (#2365).

Slight update to our downloads pages (#2365).

clcachewrapper: don't escape double quotes since we only use paths with no spaces.

Plugin implementation and reading of the variables

Implementation of the plugin Mapping based on SampleView.
The plugin can read and display in a list the variables of the open file.

Sometime crashes when opening a file.

Mapping check file type

Added a type of file check to assure no crash when the file isn't readable

Reads Mesh nodes numbers

better folding of tests mesh files

Mapping properties separated from view

Store state in QMap

Added Zinc Widget

do not display the previous lists etc but instead display an empty list and a transparent zinc widget

zinc widget create scene

remove tpm files

display the figure. Issue with depth

Show correctly large models

del useless displays

Deleted too much last time !

BROKEN moves from widget to zincwidget

MappingViewZincWidget effective

store locally region and scene

Nodes pickable

Yessss
Nodes'id print in error flow

Selected node highlight and lighter click detection

refactoring

Stop trying to copy mesh files.
Renamed meshreader, attributed all zinc work to mappingviewzincwiget.

Qt: use the new connect syntax.

return to .ui + var tree + light refactoring

Some minor cleaning up.

Some minor cleaning up.

got rid of .ui and installed toolbar and wheel

layouts fix

Finally the widget are in the good positions

basics of drag and drop

BROKEN now based on viewplugin

The plugin isn't detected

BROKEN setting mimedata

drag from tree operationnal

LLVM+Clang: upgraded to version 10.0.1 (#2371).

LLVM-Clang: binaries for version 10.0.1 (#2371).

Slight update to our downloads pages (#2371).

Tests: wherever possible, use QTEST_APPLESS_MAIN() rather than QTEST_GUILESS_MAIN().

drag & drop of the variables effective

added heavy mesh files

(too heavy for git, >200Mo)

Revert "clcachewrapper: don't escape double quotes since we only use paths with no spaces."

This reverts commit 7035b6117c120b7135c1a9bb2ebf0f13dc3c8fac.

node size changeable and drag ruled

only legal drags can be done form the tree (no multiple elements + no drag of components)

drop on zinc can't deselect node

Improved the testing of our Python interface.

The idea is to have only one output file per Python test (as opposed to one for each platform and each release/debug mode).

Some minor cleaning up.

Improved the testing of our Python interface.

The idea is to have only one output file per Python test (as opposed to one for each platform and each release/debug mode).

Some minor cleaning up.

Improved the testing of our Python interface.

The idea is to have only one output file per Python test (as opposed to one for each platform and each release/debug mode).

Improved the testing of our Python interface.

The idea is to have only one output file per Python test (as opposed to one for each platform and each release/debug mode).

Improved the testing of our Python interface.

The idea is to have only one output file per Python test (as opposed to one for each platform and each release/debug mode).

Make CodeFactor happy.

Core: slight improvement to our MathML tests.

Revert "Core: slight improvement to our MathML tests."

This reverts commit e22409987af5bb83081999f9cd24fa9ccca8845f.

try to change color of mapped nodes

doesn't work

Renamed MappingView in CellMLZincMappingView

Some minor cleaning up.

Some minor cleaning up.

Improved the testing of our Python interface.

The idea is to have only one output file per Python test (as opposed to one for each platform and each release/debug mode).

no more pointer to zinc objects

Color changing effective

tree vriables no more editable

Core tests: we need our MathML tests to use QTEST_GUILESS_MAIN() on Windows.

Otherwise, we get something like:
```
********* Start testing of MathmlTests *********
Config: Using QtTest library 5.12.9, Qt 5.12.9 (x86_64-little_endian-llp64 shared (dynamic) debug build; by MSVC 2017)
PASS   : MathmlTests::initTestCase()
QFATAL : MathmlTests::plusTests() ASSERT: "!m_baseURI.isRelative()" in file environment\qgenericstaticcontext.cpp, line 82
FAIL!  : MathmlTests::plusTests() Received a fatal error.
Unknown file(0) : failure location
Totals: 1 passed, 1 failed, 0 skipped, 0 blacklisted, 91ms
********* Finished testing of MathmlTests *********
```

Improved the testing of our Python interface.

The idea is to have only one output file per Python test (as opposed to one for each platform and each release/debug mode).

delete button operaitonnal. begin of save button

Don't make external processes time out.

Revert "Some minor cleaning up."

This reverts commit 499dc2dc47b735d3f657700d48e3454bbf4ba76d.

Some minor cleaning up.

Make Clang-Tidy happy.

Map saving operationnal

Some minor cleaning up.

Make Clang-Tidy happy.

Pendulum window: fixed a small problem with the axes' font point size.

Zinc window: skip createAndSetZincContext() when closing OpenCOR.

Zinc window: keep a temporary copy of our .exfile file.

Pendulum window: skip createAndSetZincContext() when closing OpenCOR.

name of the mesh model in the map file, and some commented code

Fixed a typo.

Fixed a typo.

Zinc: don't allocate our Zinc context on the heap.

See the square-bracket comment at https://github.com/opencor/opencor/issues/2369#issuecomment-673350896.

Some minor cleaning up.

Zinc: don't allocate our Zinc context on the heap.

See the square-bracket comment at #2369 (comment).

Some minor cleaning up.

open mesh button operationnal

Pendulum window: make sure that everything still works fine when docking/undocking the window.

Caveat: OpenCOR will crash if there is another OpenGL window (e.g. `ZincWindow`) and we dock/undock `PendulumWindow`.

Some minor cleaning up.

Some minor cleaning up.

solve trilinear and megamesh bug

now read the coordinates field name automatically

light refactoring

mZincContext et mScenePicker are not pointers anymore

find coordinate field properly

Zinc window: automatically retrieve the coordinates field, independent of its name.