Henrik Feldt
Wouldn't it be more correct and easier to reason about types of measurements? You could provide equality comparers for euclidian distance, similarity, groupedness, or whatever you consider is almost equal....
There's a cost to everything, including breaking things up. There's a cost for learning to pull down both Expecto.Core and Expecto.Expect via the meta-package Expecto. It lowers cohesion ever so...
Is this ready to merge?
My guess is that Microsoft has special-cased xUnit; perhaps if you compare the exact stacktrace output and/or any lines that come just before it, you'll find why and could send...
They both print to the stdout, and one is recognised, the other one is not. I'd bet one is recognised due to its unique fingerprint; that is, the particular formatting...
Perhaps it would be worth it for you to write an Azure pipeline printer? We already have a Teamcity one; https://github.com/haf/expecto/blob/2462fb8614d4190206155b193dcde81de3a6dd4a/Expecto/Expecto.fs#L786-L866 that gets automatically used if necessary https://github.com/haf/expecto/blob/2462fb8614d4190206155b193dcde81de3a6dd4a/Expecto/Expecto.fs#L941-L945
@cmeeren Perhaps you can file a support issue with MSFT and ask them? I don't know either
If it's based on the test results file; we support it https://github.com/haf/expecto#testresults-file
Hi, I just thought that if a TestResults.xml file was generated, Azure would find it, but I guess you're successfully running Expecto and generating such a file already. I'm actually...
But isn't our discussion about what a "proper stack trace is"? If you find what makes it proper, then PR that and we'll run with it ;) That said, inside...