Improve crawling error reporting

[christophfriedrich]

When data is old, this can be for several reasons:

• the refresh of a resource didn't succeed because
  • the backend was offline
  • the Hub didn't understand the backend's (invalid) response
• the crawling wasn't performed as scheduled
• the database was not reachable
• ...

Currently, the only error reporting is:

• the crawl script outputs to the console
• the frontend renders warning notices like this:

The user should receive better error messages for the different cases.

Also see #13.

[christophfriedrich]

In the meantime, crawling error reporting was massively improved as a side-effect of the normal development process. Especially in the commits that lead to the v0.5 release, e.g. f013bb1. So when this issue comes onto the agenda again, the current status should be re-assessed.