Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[package tests] CI mode not working? #161

Open
mcbain opened this issue Apr 28, 2015 · 3 comments
Open

[package tests] CI mode not working? #161

mcbain opened this issue Apr 28, 2015 · 3 comments

Comments

@mcbain
Copy link

mcbain commented Apr 28, 2015

meteor test-packages --velocity --driver-package respondly:test-reporter YOUR:PACKAGE

What is the difference of the mentioned CI mode, to a normal run?
Currently it does not stop after execution, nor generating any information usable in a CI system (e.g. xunit output etc).

@zimme
Copy link

zimme commented May 19, 2015

meteor test-packages --velocity --driver-package velocity:console-reporter ./

@rissem
Copy link
Collaborator

rissem commented Jun 20, 2015

The --velocity flag outputs test information (pass/fail) to the terminal. Without this flag, passes and fails are not printed to the terminal console. If it's not stopping after execution that's a bug. Do you have a reproduction repo I could look at?

@chmac
Copy link

chmac commented Jun 21, 2015

@rissem I spent some hours on it last night. Ripping out parts of my app to try and pin down an issue. I found a few things.

First, if you only import mike:meteor-moch on the server, the tests will never run, so the process will never exit. I created an example.

I've been through my whole application, and if I load mike:mocha-package on both client and server, but only load the package under test and my test files on the server, the tests properly exit.

It's interesting, because I found a few bugs in my Package.onTest() calls (oh javascript, you and your fussy brackets...), which when fixed, seem to have resolved my CI issues. But, all of my tests were correctly run, and the results correctly reported, without the --velocity flag. It seems like there's some kind of difference between how tests are handled when there's a browser involved and when not.

I'll keep digging, and if I'm able to find a simple reproducible case, I'll post it here. Until we need client side testing, it looks like my issue is resolved for now.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants