Closed DavidLeoni closed 8 years ago
Closed in staging-0.3
Seems like sending to Refine a schema with a non-cached etype doesn't actually breaks the system, it is only the error message which is not that clear. To improve things a bit, I implemented a better (although still crude) exception report in #216 , so at least in the ui it is possible to understand the problem cause by looking at the stack trace. This is of course very rough but I think that given how much caching is problematic, doing error reporting nicely for caching-related issues is premature.
Searching etypes should either only show cached etypes or be resilient to changes, at least in the case of new etypes being found in the list.