Easy off on crawling

[ojongerius]

We crawl once a day, for bigger sites this can imply visiting all the job urls, of which me might have seen most. Some solutions:

• Fetch only new jobs (new that 24 or 28 hours).
• Save job urls (we can search, or add functionality to search on URL to the REST API), and check when it was last seen.

• merge PR with a solution
• re-enble scheduled jobs

[ojongerius]

Until we can update objects in #26 , I decided to not revisit URLs that are already associated with an existing job. This will solve most of our use cases; we save on a lot of unnecessary traffic, and I don't expect objects to change often.