Conversation
|
@ocefpaf Sorry I didn't see this! Must've gone past my radar. From the error: This might be caused by the database instance not firing up. |
No problem.
I guess it was broken for a while then and we did not notice. (I have |
Yea. One sec. Lemme get this setup. I'll get back to you asap. |
|
@ocefpaf It seems like the mysql setup script might requires sudo, and travis is not allowed to use sudo: https://travis-ci.org/ODM2/ODM2PythonAPI/jobs/263084731#L643 |
|
I'll investigate the alternatives. Thanks @lsetiawan! |
|
When I tried using |
Cool. I did that for now so we can focus on those errors. I am also reading this https://docs.travis-ci.com/user/database-setup/ Maybe we can avoid |
I saw that documentation too, but my question is can we stop and start the mysql server without sudo in travis? |
|
@ocefpaf the problem comes when we have to add a line of configuration for mysql-server https://github.com/ODM2/ODM2PythonAPI/blob/master/ci-helpers/mysql_setup.sh#L10. I will take a look if there's a way that this can be done through SQL rather than manually through script! |
|
Looks like the API change is that
I'm reading more on that too but |
lsetiawan
left a comment
There was a problem hiding this comment.
The travis-ci failure is now related to the actual API rather than database. I am going to merge this for now and address the tests API failures.
|
Thanks @lsetiawan! |
No. I didn't make the updates to the API, but I know how it works.
Should I not merge yet then? |
Awesome! I would never find that b/c I don't really use
Got for it. This is not a code change but a tool to fix the CI 😄 My guess is that we are close to make it green again. |
|
BTW, my reasoning when made #90 (comment) I meant that the test code change should've be part of the PR that implemented the API change. However, because Travis-CI has been broken for ages, we started ignoring it and accumulating changes without really knowing the consequences. Only local tests of the type "works for me." |
Should be a fairly easy fix. I think I got it 😉 , thanks for offer. I'll make a PR once it's done. |
This PR removes the tests for the code removed in #81
There are now some other failures that may be real or due to our testing setup (I also suspect some unicode/decoding issues, not sure).
@lsetiawan can you review this and take a look at the new failures?