I have usually solved this by running my tests one category at a time. In order to ensure I allways run all tests, my last test run executes all categories
except the ones that were explicitly executed in the first run.
You then have mbunit store the test result to xml files. (gallio.utitilies.exe)
Once the all the tests have executed, i genereate one html file per category.
Finally - and this is the reason why I create XML first, then html after - I create a "condensed" report from ALL the xml files. The condensed report only contains tests with problems - ignored and fail - and is typically the only report you are interested in. I like to keep both reports as build artifacts on my server.
I have attached the .bat file I use for accomplishing this - I give no guarantees that it will actuall work - but at least this has done the trick for me...
Regards
Espen Albrektsen
This is the script I use: