Forums

I'm in the process of writing tests for existing code and using Yeti to run through them all. The trouble is, when an error is encountered in a test, the browser has already blown past it and Yeti is back to its listening page giving me no chance to see the loaded scripts, debug them and walk through the code. The only thing I can do is guess where the failure occurred and throw debugger statement in there to force Yeti to wait. This sometimes also crashes Yeti, but that's besides the point.

Would it be possible to include an option to pause before each test or a specific test (via special comments). That way, once the page is loaded, I can set breakpoints and walk the code.

I was running 0.2.0 (just upgraded to 0.2.1) and it would crash but not because of the debugger statements. It tends to crash more often when I abuse it. I'm not 100% sure how to replicate it yet (haven't been paying close attention, been focused on writing tests) but it tends to happen when I'm abusing it.

It seems to happen when my tests have problems (bad paths to files, errors in scripts, no browsers pointing to yeti server when I start a test run) and then the test hangs and I break out. Then I run it again after fixing stuff and at the end of the run yeti server crashes. Something like that.

Now that you mention it, I have been having trouble with Firefox. For some reason, it's the only browser (regardless of OS) that ends up with an "undefined" in the URL at the end of a test and fails to go back to the waiting state. Seemed like to big and obvious of a problem so I figured it's just broken for me. Not sure though how to debug that. For the time being I'm just running tests on other browsers.