November 27, 2015
I am pleased to announce that I have started developing a database back-end for the Barcrawl web app and have been making good progress. Since I have deployed Barcrawl using Elastic Beanstalk on Amazon AWS, I am limited to a MySQL database. But that is not a problem since my only previous experience is in MySQL. This database is meant to reduce the application’s dependency on API calls in the Python back-end. This decision comes in order to reduce the impact on the APIs’ request quotas and to help reduce latency in the back-end since the rate of requests to the Google Maps API have to be limited.
The database setup currently has three tables for cities, bars and distance vectors. In order to keep results up-to-date, bars are only requested for cities that have not been updated in the past month, or that do not have enough bar entries saved in the database. Distance vectors between bars are also saved to reduce the need to ask Google Maps for directions data. Once the database has been fully implemented and tested I plan to fully explain its layout.