> Any chance we could write a test to cover this class of problems; for example,
> verifying that all instance relationships are accessible without triggering SA
> errors? (Perhaps with an eye to extending this to all models in the future...)
Unfortunately I don't think I'll be able to for a couple reasons:
1) There's currently no DB testing suite (it's all faked), and I wouldn't have time to write that this week during the ending of our Sprint cycle.
2) I'm unable to reproduce this bug after spinning a clean build environment. My thoughts are that it was a one-off situation in our environment that triggered this exception.
> Any chance we could write a test to cover this class of problems; for example,
> verifying that all instance relationships are accessible without triggering SA
> errors? (Perhaps with an eye to extending this to all models in the future...)
Unfortunately I don't think I'll be able to for a couple reasons:
1) There's currently no DB testing suite (it's all faked), and I wouldn't have time to write that this week during the ending of our Sprint cycle.
2) I'm unable to reproduce this bug after spinning a clean build environment. My thoughts are that it was a one-off situation in our environment that triggered this exception.