PGObject and PGObject::Simple
Hi, apart from these, I've started quite a few other modules too. One problem I'm seeing is that, using OpenBSD's PostgreSQL testing module (which I'm only just now getting aquainted with) is that I'm having to patch the tests for get past the skip_all. They pass the tests just fine, except that they both try to drop the test database while it's still open. I'll ask about this on the OpenBSD mailling list. I may be doing something wrong. I want to be sure and test everything with the database. Kind of important. :-} Is there a kink in the test programs or am I just doing something wrong? I can just patch things on the OpenBSD side, which is pretty commonly needed. That's not a problem. I'll get back. Been working on things in spurts. Many modules started. Now I'm working on finishing up on them one by one. Regards, Chris Bennett
Hi, On Sun, Dec 15, 2019 at 9:56 PM Chris Bennett <chris@bennettconstruction.us> wrote:
Hi, apart from these, I've started quite a few other modules too.
One problem I'm seeing is that, using OpenBSD's PostgreSQL testing module (which I'm only just now getting aquainted with) is that I'm having to patch the tests for get past the skip_all.
I'm not sure what you mean, here. Can you post a patch that you're applying? Looking at http://matrix.cpantesters.org/?dist=PGObject+2.000002 I'm seeing that the tests of this module run fine on OpenBSD. Am I missing something?
They pass the tests just fine, except that they both try to drop the test database while it's still open.
Is this a local problem? Or a problem on a remote OpenBSD build server for ports?
I'll ask about this on the OpenBSD mailling list. I may be doing something wrong. I want to be sure and test everything with the database. Kind of important. :-}
You mean you're patching the modules to test *with* the database on some central build infrastructure? The tests that run against a database are marked AUTHOR_TESTS which means that they're only meant to be run by the developer (or maybe by the one doing the ports, which in this case would be you). I don't think it's generally good practice to promote an AUTHOR_TEST to a test being run on system installation.
Is there a kink in the test programs or am I just doing something wrong? I can just patch things on the OpenBSD side, which is pretty commonly needed. That's not a problem.
Well, I don't really know about the kink, but I'm wondering: is this problem you describe also a problem you experience when testing locally? If not, then you found your problem, I think.
I'll get back. Been working on things in spurts. Many modules started. Now I'm working on finishing up on them one by one.
Very nice ! Hope you get through the list soon so we can see a release on OpenBSD soonish! Regards, -- Bye, Erik. http://efficito.com -- Hosted accounting and ERP. Robust and Flexible. No vendor lock-in.
participants (2)
-
Chris Bennett
-
Erik Huelsmann