Use raphael.js for some mildly interesting timezone visualizations. Instead of using dropdowns or just raw text to determine city offset, we "plot" cities along an infinite timeline.
This allows one to quickly and visually quantify the difference in hours and time of day among various cities.
Dragging the the timeline allows you to quickly compute "what-if" scenarios: "If it's 9am in NYC, what time is it in London?" and vice versa.
Hovering over a city will give you relative calculations to other cities; how many hours the other cities are in relation to the city you're hovering.
A demo can be seen here: http://www.roflopoly.com/zoneslider/
browserify is used to combine all source JS files into a single file. grunt
needs to be installed globally so that the
grunt binary is available on your
npm install -g grunt
grunt task is
browserify watch, which will both combine the
The outputted file is
Sinatra is used to serve up everything within
public. This includes the
main index.html file and all static assets.
2 simple API endpoints exist:
/utc returns the current UTC time as reported by the server.
/search?q= returns search results for city searches.
Where is the data coming from?
Pretty much all Unix systems make use of the tz database (aka Olson database or zoneinfo)
located somewhere near /usr/share/zoneinfo. These set of files store all the rules of
timezones and associated daylight savings adjustments for various cities in the format of
geonames.org provides Creative Commons licensed data files that mesh up city names with the zoneinfo timezone name. http://download.geonames.org/export/dump/readme.txt
data/generate_db.rb is a small ruby script that reads in
data/cities15000.txt, a tab-
delimited file containing cities and their zoneinfo-formatted timezone name, and loads
cities.db, a Sqlite3 database.
Sinatra then just does a search within cities.db to find correspodning timezone information. The ruby library TZInfo is used to compute actual timezone-related calculations.