Test "NumberFormatOptionsImplSpecific-Win10.js" failed on Windows Server 2012 R2

[KarenHuang2016]

Environment: Windows Server 2012 R2 + VS2017 + Chakra master branch latest revision aeb7117
Repro steps:

1. git clone -c core.autocrlf=true https://github.com/microsoft/ChakraCore D:\Chakra\src
2. Open a VS 2015 x86 command prompt and browse to D:\Chakra\src
3. msbuild /m /p:Platform=x86 /p:Configuration=Test Build\Chakra.Core.sln /t:Rebuild
4. pushd test
5. runtests -x86test

Test "NumberFormatOptionsImplSpecific-Win10.js" run failed as below:
*** Running test #1 (0): Test Valid Options - Formatting Currency with Significant Digits
Test threw exception: assert.areEqual failed: expected: USD 123 actual: USD 123: currency code: USD
FAILED
*** Running test #2 (1): Test Valid Options - Formatting Percentage with Significant Digits
Test threw exception: assert.areEqual failed: expected: 12,300 % actual: 12,300 %: style: percent
FAILED
Summary of tests: total executed: 2; passed: 0; failed: 2

This test passed on Win10 but failed on Windows Server 2012 R2. Could you please take a look at this? Thanks in advance!

[dilijev]

That's correct. That test is tagged in the rlexe.xml in that directory as exclude_winBlue (which, for our purposes, is equivalent to Server 2012 R2).

• If you call rl.exe directly, please pass -nottags exclude_winBlue
• If you call runtests.cmd, please pass the -winBlue parameter which will cause the above to be passed to rl.exe

Our Jenkins VMs run Server 2012 R2 and netci.groovy passes this flag to testone.cmd which internally calls runtests.cmd. This is necessary because we want to have tests that cover all platforms, and sometimes there are behavior differences between platforms. This test was split into two copies and tagged so that only the correct variant would run on each platform. (Also, the complexity of the various test runner scripts is because of all the platforms and environments in which we need to support running tests.)

Sorry for the confusion.

See: https://github.com/Microsoft/ChakraCore/pull/2664/files#diff-2d212ad84506951c8f928e21aed24580

[KarenHuang2016]

@dilijev, thanks for your patient explanation! I will change the way to run test.