Next I'll try to find a way to avoid hardcoding the branch name and shorten the paths,
but for now this should avoid accidental overwriting of the demos
of the release.
The generated code tries to compute the box layout info for an empty
repeater, where we end up calling `&*std::begin(cells)` and MSVC
asserts that this is dereferencing an invalid iterator. As Olivier spotted,
`std::data` comes to the assert-free rescue :-)
Text's behavior right now is to clip on a line basis, so with the given
font-size for these two elements, the glyphs aren't shown at all because
the ascent + descent that's use for the clip test is bigger than the
maximum height, despite the actual glyphs fitting.
For now, reduce the font-size slightly to make the glyphs fit.
In this case, the last line will have elipsis on the last line
when no more line can be printed
Note that the behavior is different with the Qt backend
Unfortunately we require cmake right now, due to fontkit wanting to compile
in a copy of freetype. That could be fixed though by making it use
ttf_parser perhaps.
With the nightly rust compiler, we include the langref.md in the
rust documentation, but we don't want these snippets to be understood
as rust code and tested
Glutin's desire to default to srgb mysteriously creates a black window
on Windows 10 under Vmware. I tested this with many glutin apps and examples,
they all exhibit the same: black window,
working when query for srgb is turned off.
This used to work, so I suspect some change underneath on the vmware svga
driver perhaps, or Windows.
The interpreter created an intermediate window for the component but
the regular parent window is used for rendering. This resulted in the situation
where the layout computation was done before rendering with the intermediate window,
the text layout was calculated and cached in the Text's cached rendering data.
That in turn is just an index into the window's rendering cache (arena),
which is also later used for rendering. The index is local to the window, so
during layout the index was assigned for the intermediate window and
when rendering it seemed valid and happened to also refer to a Text item in the
cache of the real window, but with the wrong size/data.
Fixes#278
Doxygen would still generate xml for private_api::ArrayModel and would
try to resolve a reference to iself, which is not possible because we've
excluded the namespace:
```
stderr: (!) Critical error while generating the file for [/home/vagrant/sixtyfps/target/cppdocs/api/classsixtyfps_1_1Model.rst].Traceback (most recent call last):
File "/home/vagrant/.local/share/virtualenvs/docs-4wjFCyEr/lib/python3.8/site-packages/exhale/graph.py", line 2591, in generateSingleNodeRST
gen_file.write("{0}\n".format(node.baseOrDerivedListString(
File "/home/vagrant/.local/share/virtualenvs/docs-4wjFCyEr/lib/python3.8/site-packages/exhale/graph.py", line 393, in baseOrDerivedListString
link=nodeByRefid[refid].link_name
KeyError: 'classsixtyfps_1_1private__api_1_1ArrayModel'
```
Making the entire class as \private works around it.
Oddly, this isn't needed for IntModel, perhaps because it's not a template?
This reverts the Pipenv pinning of
f9ef21bffd and thus disables the logo, as
subsequent changes in the pinned sphinx fork cause unrelated breakages
when building docs. Let's re-enable this when the fixes are in an
upstream release of sphinx and the rtd theme.
