We have some projects that are very expensive to build. Since I would like to see the output of the last test, I do not want to disable NCrunch completely, but the build engine keeps building the projects even when the tests are all set to run manually and I haven't yet started a test. I would like the option to defer the build until a test is scheduled for execution.

Hi.

I want to be able to mark that nCrunch only use/load the projects that are loaded in VS.
Since our solution are becoming big, I use Funnel to load only the projects I need for my current task. But nCrunch still load all projects in my solution file.

User can hide excluded items in Visual Studio by setting MetricsTreeShowExcludedMetrics value to false. But if you use command line tool for CI it outputs all test and sets html class to "excluded childNode".
It would be nice to have setting ShowExcludedMetrics argument in command line tool

As discussed on Twitter (https://twitter.com/remcomulder/status/782892104485703681), I'd like NCrunch to support custom test cases.

xUnit has the support for custom test cases, as the STAExamples of the xUnit Samples demonstrate (https://github.com/xunit/samples.xunit/tree/master/STAExamples). That relies on using your own test discoverer, and specifying a custom test case. When you run that in NCrunch, it fails with a cast exception (https://gist.github.com/tiesmaster/5ecb8bf76aab6ee3cb2bfcc9d4f08e72#file-stacktrace-of-staexamples-on-stable-xunit-packages-downgraded-to-2-1-0).

I think this is a niche feature, since this something that won't be used that often, but still a powerful feature, so it would be really nice to have this.

In the code coverage report, we would like to be able to navigate into the source code files and see which lines are covered, much like OpenCover does. Here is a screenshot of that feature in OpenCover:

https://www.dropbox.com/s/38kj2wr5bczbtls/OpenCover%20Source%20Code%20Coverage%20Example.PNG?dl=0

Our developers love using NCrunch. We would like to use NCrunch for our CI process, especially because it would unify the development and CI environments. The organization is hesitant to use NCrunch for CI without the ability to drill down into the source code in the report to see what lines of code were covered when the CI job ran.

Integrate with the upcoming IDE from jetbrains to support project rider. Project rider is already available for download. Based on intellij. NCrunch couldn't do exactly what WallabyJS is doing and providing a number of IDE integrations.

MSTest allows us to tag tests with [TestProperty] attributes and their values are then available as this.TestContext.Properties.
However, ncrunch is not populating this dictionary, so a test using these attributes works in mstest/resharper, but fails in ncrunch.

it all in the title :)

I'm using NCrunch for more than 2 years.
IMHO it's must-have extension.
When I come to new company that do not use NCrunch I need to fix/stabilize some test to run with NCruch engine.
But still I stay with few flaky test.
I need to rerun them few times to filter trully failing test.

It would be nice introduce option for my case with flanky test.
It could be rerun count before test is treated as failed
Other option to introduce 3rd test state: flanky.

I use Nunit, with ExplicitAttribute(string reason) on TestFixture class.
The tests are correctly ignored but there no reason visible in NCrunch Tests window.

It would be nice to have the reason, when I select an ignored test. Perhaps in Trace Output (which is empty for ignored tests).

Today, NCrunch shows the execution time across all threads on its performance markers.

If code is run in parallel, e.g.

Parallel.For(0, 100000, DoSomething);

the execution time is somehow misleading as it is not the real execution time (eg. 5 seconds) but instead the overall time (eg. 1 minute).

It would be nice if NCrunch would show both times, the actual time needed for the loop as well as the overall time accross all threads.

I would like NCrunch to display the actual time taken for an async method to have been called, when using await, instead of showing 0ms, just like they do for ordinary synchronous calls.
See http://screencast.com/t/1mwMqiK0bi

Example code:
[TestFixture]
internal class TestFixture
{
private async Task FooAsync()
{
await Task.Delay(1000);
}
[Test]
public async Task AsyncTest()
{
await FooAsync(); //Coverage marker indicates 0ms duration.
}
[Test]
public void NonAsyncTest()
{
FooAsync().Wait(); //Coverage marker indicates 1s duration.
}

}

Perform Mutation Testing on the test cases to identify incorrect test cases.

Currently the "Test Filter Builder" screen in custom engine modes renders very badly. (This is on my 4K laptop screen). See http://imgur.com/Vw5Q1qZ.

This makes the feature barely usable - would be great (and hopefully simple) to fix it.

MSpec uses a [Subject] attribute to nominate the unit-under-test and optionally allows a "concern" to be specified.

In R# these items appear in the test runner window as part of the test name, however in NCrunch there doesn't seem to be any way to get at this information, which is unfortunate.

There just seems to be a level of organization missing when working with MSpec tests. It would be really helpful to be able to group the tests either by the [Subject] attribute, or by the file name that the test is defined in.

Please consider adding support for Gauge (http://getgauge.io). Gauge is an open source test platform, started by ThoughtWorks, for writing specification by example. It is similar to Gherkin, only it uses Markdown. Steps are bound to code using StepAttribute binding.

When I click on a coverage marker circle, a windows pops up with Test name and execution time information.
It would be nice to have also a third column indicating how much times the test hit the current line

Ncrunch does not appear to check all conditions in an if statement for 100% coverage.
i.e. if (x || y) only requires that x is tested, not y.