Judges' feedback

[ghost]

Placeholder issue for judges' feedback.

[ghost]

Verified score of 114 points on the automated scoring, based on the current master branch.

[ghost]

Models and DB:

• I see that you're using code as the foreign key to Emergency from Responder. Is this column an oversight from an earlier implementation?
• Fine use of scopes on Responder which get leveraged elsewhere.
• Doesn't this destroy the related Responder records? If it does then this is not what was intended, and that would mean there's a gap in the test coverage (our bad). The intended implementation would simply unassign the responders from the emergency when it is resolved, freeing them for another, future emergency.
• Straightforward capacity reporting

Routing:

• Simple, clean routes. Easy 404 handling.

Controllers:

• Simple centralization that handles the various concerns with controller actions: routing errors, 404 pages, and unpermitted parameters.
• Your EmergencyController and RespondersController seem immediately understandable. No real complexity, and judicious sprinkling of things like create_params and update_params, though I think those could potentially be inlined since they are specific just one action each.
• Succinct dispatcher code - really like this. I also see that you put it in lib where a lot of people put it in app/models as PORO objects. Perhaps this is a side effect of when I first started using Rails, but the lib folder seems right to me.

---

Thanks for your participation in the challenge and good luck in final scoring!

[atomaka]

Thanks for taking the time to review my solution!

I've corrected the extraneous column; it was left over from earlier work where I considered using id over code as they key.

As far as I can tell, responders.clear is not destructive and should only remove the association. I confirmed this with a pry before and after the posts in the dispatch test file, verifying that Responder.all.count remained unchanged. Have I misunderstood your note?

[ghost]

If you've confirmed it with pry then that's good enough for me, especially since it passes the tests.

Thanks for the follow up.

[ghost]

:trophy: Congrats on placing a very close 2nd in the first railschallenge!

You had the best dispatch component (the hardest single component) according to the judges. We'd like to invite you to write up something about your approach to the dispatch code, as well as your overall approach, to be included in the feedback for all participants.

Please reach out to us at info@railschallenge.com

[ghost]

Your individual score breakdown:

311.5 points total

• 114 points for passing tests
• 197.5 points for the judge's section (actually the highest anyone got)

You were unfortunately edged out of 1st place by the bonus points awarded to @sb8244 for this pull request, a very small collaborative code bonus that we didn't expect to alter the outcome of the contest. So, really 1st and 2nd place were decided by a spread of just 3.5 points. Your amazing work on the dispatch code, which will be highlighted in our writeup, was superb and deserves special recognition since it was the most difficult part of the challenge. All of the judges commented on your clean, minimal code. I'd hire you in a heartbeat (if we were hiring).

Really outstanding job.