Test failures in develop with msvc-14.1

[awulkiew]

This is the list of failures I get right now with msvc-14.1 (VS2017)

b2 toolset=msvc-14.1 --abbreviate-paths cxxflags="\wd4244 \wd4996" -j4 test

test\algorithms\detail\algorithms_approximately_equals

test/algorithms/detail/approximately_equals.cpp(75): error: in "test_main_caller( argc_ argv )": Expected: 54 Detected: 43
test/algorithms/detail/approximately_equals.cpp(75): error: in "test_main_caller( argc_ argv )": Expected: 44 Detected: 33
test/algorithms/detail/approximately_equals.cpp(75): error: in "test_main_caller( argc_ argv )": Expected: 34 Detected: 23

test\algorithms\overlay\algorithms_get_clusters

test/algorithms/overlay/get_clusters.cpp(67): error: in "test_main_caller( argc_ argv )": Case: borderx_no ctype: e expected: 0 detected: 1
test/algorithms/overlay/get_clusters.cpp(67): error: in "test_main_caller( argc_ argv )": Case: bordery_no ctype: e expected: 0 detected: 1

test\algorithms\overlay\algorithms_sort_by_side

test/algorithms/overlay/sort_by_side.cpp(215): error: in "test_main_caller( argc_ argv )":   caseid=case_recursive_boxes_46_union open count: expected=[2 , 1 , 2 , 1 , 1 , 2 , 1] detected=[2 , 1 , 1 , 2 , 1 , 2 , 1]

---

FIXED:

test\algorithms\algorithms_line_interpolate https://github.com/boostorg/geometry/pull/854

test/algorithms/line_interpolate.cpp(60): error: in "test_main_caller( argc_ argv )": MULTIPOINT((1.4 1),(1.8 1),(2 1.2),(2 1.6),(2 2),(1.6 2),(1.2 2),(1 2.2),(1 2.6),(1 3)) != MULTIPOINT((1.4 1),(1.8 1),(2 1.2),(2 1.6),(2 2),(1.6 2),(1.2 2),(1 2.2),(1 2.6))
test/algorithms/line_interpolate.cpp(60): error: in "test_main_caller( argc_ argv )": MULTIPOINT((1.8 1),(2 1.6),(1.6 2),(1 2.2),(1 3)) != MULTIPOINT((1.8 1),(2 1.6),(1.6 2),(1 2.2))
test/algorithms/line_interpolate.cpp(60): error: in "test_main_caller( argc_ argv )": MULTIPOINT((1.79997 1.00002),(2 1.59985),(1.60006 2.00007),(1 2.20015),(1 3)) != MULTIPOINT((1.79997 1.00002),(2 1.59985),(1.60006 2.00007),(1 2.20015))

test\algorithms\overlay\algorithms_get_clusters https://github.com/boostorg/geometry/pull/878

test\algorithms\overlay\get_clusters.cpp(82): error C2398: Element '3': conversion from 'initializer list' to '_Ty' requires a narrowing conversion
        with
        [
            _Ty=fp
        ]
test\algorithms\overlay\get_clusters.cpp(82): error C2398: Element '1': conversion from 'double' to 'const CoordinateType &' requires a narrowing conversion
        with
        [
            CoordinateType=float
        ]
test\algorithms\overlay\get_clusters.cpp(83): error C2398: Element '3': conversion from 'initializer list' to '_Ty' requires a narrowing conversion
        with
        [
            _Ty=fp
        ]
test\algorithms\overlay\get_clusters.cpp(83): error C2398: Element '1': conversion from 'double' to 'const CoordinateType &' requires a narrowing conversion
        with
        [
            CoordinateType=float
        ]
test\algorithms\overlay\get_clusters.cpp(84): error C2398: Element '3': conversion from 'initializer list' to '_Ty' requires a narrowing conversion
        with
        [
            _Ty=fp
        ]
test\algorithms\overlay\get_clusters.cpp(84): error C2398: Element '1': conversion from 'double' to 'const CoordinateType &' requires a narrowing conversion
        with
        [
            CoordinateType=float
        ]
test\algorithms\overlay\get_clusters.cpp(85): error C2398: Element '3': conversion from 'initializer list' to '_Ty' requires a narrowing conversion
        with
        [
            _Ty=fp
        ]
test\algorithms\overlay\get_clusters.cpp(85): error C2398: Element '1': conversion from 'double' to 'const CoordinateType &' requires a narrowing conversion
        with
        [
            CoordinateType=float
        ]

[mloskot]

Does variant=debug vs variant=release make a difference?

[awulkiew]

Does variant=debug vs variant=release make a difference?

@mloskot No, the failures are the same in both cases. The last one is compilation error so it doesn't matter.

algorithms_line_interpolate fails since it was implemented. AFAIU it's a numerical issue. Right @vissarion ?

I created this issue to let @barendgehrels know about these failures because I think he doesn't test with msvc locally.

[vissarion]

algorithms_line_interpolate fails since it was implemented. AFAIU it's a numerical issue. Right @vissarion ?

right

[awulkiew]

@vissarion I don't know why exactly but it seems that after merging https://github.com/boostorg/geometry/pull/840 line_interpolate test also started to fail with gcc-6.3 the same way: https://app.circleci.com/pipelines/github/boostorg/geometry/165/workflows/5d3af105-4f5c-49b5-a4b6-7a128f94347c/jobs/12439 It also fails for me locally with gcc-9 as well as with clang-10.

Do you have an idea why this may be happening? Does it make sense to alter the expected results?

[awulkiew]

The issue was automatically closed after merging a fix for one of the failures. So I reopened it and edited the issue description.