Spawn a new build...

Codecov Report

Merging#1872 intomaster willincrease coverage by0.02%.
The diff coverage isn/a.

@@            Coverage Diff             @@##           master    #1872      +/-   ##==========================================+ Coverage   74.44%   74.46%   +0.02%==========================================  Files         282      282                Lines       13253    13253                Branches     2382     2382              ==========================================+ Hits         9866     9869       +3+ Misses       3257     3253       -4- Partials      130      131       +1

Continue to review full report at Codecov.

Legend -Click here to learn more
Δ = absolute <relative> (impact),ø = not affected,? = missing data
Powered byCodecov. Last update7f48a19...6753b27. Read thecomment docs.

Try this out on the new CI definition in VSTS...

Another crack at VSTS CI

Once again...

Once again...

@DonJayamanne this change actually doesn't seem to have any effect at all. If you look here at the code, and then search the 'testSingleWorkspace' output for that test, you can see theIS_CI_SERVER value is being completely ignored. Perhaps this is due to a shortcoming on the VSTS side, I'll check it out further... but maybe how I've used the environment variable is to blame?

Agreed, something wrong in VSTS, cuz it works as expected on Travis and appveyor

I found it. The VSTS env varTF_BUILD is set to"True" in true MSBuild style.

My expression was failing due to case mismatch!

I found that in the VSTS tooling it is enough to test for theexistance of theTF_BUILD environment variable so I've decided to follow that pattern. (I doubt very much a VSTS build agent would ever define that value as "False" or anything else...).

@DonJayamanne yep this fixes it! I'm shoving it in... will watch it and revert tonight if it turns out problematic...