Nerd-o-meter — from what write tweets

In free time from work we together with Dima Medvinsky and Phillip Nirulin, wishing to improve knowledge in JS, HTML and CSS, have created an interesting project for admirers of statistics.

Nerd-o-meter — it is a service, that allows know about statistics of devices, from which tweets are written. It is possible to search by requests and by specific users.

Idea

The idea of the project came up while visiting "Yandex.weekend": watching for that how many people were writing in twitter about events, it came to my mind, that it would be interesting to learn about statistics of the devices, from which they were writing and see how it vary from one conference to another.

Later this thought was transformed in an idea of a small service that shows statistics by twitter to clients: this information can be easily taken out of twitter API.

Creation

From the very beginning, it was imagined to be in view of classical computer application, for example on Rails, but at some time it became obvious that server part here is just redundant: all what is necessary in particular access to Twitter API and building of diagrams, are available directly in browser. So, a small and light-weighted single-page app appeared.

For building of diagrams was used Google Chart Tools, passingly looked aside of several alternative decisions.

While searching for necessary libraries for developing, an excellent collection microjs.com (though in this small project it was possible probably to go without them) helped me very much.

• Zepto.js:is a light-weighted replacing of jquery, intended first of all on mobile webkit-browsers. Nevertheless it works very well as an "adult library". During the work was found out a small bug while working with FF, but it was updated very quickly by patch, which was send later (and adopted) back in zepto project on GitHub. Used for DOM manipulation and AJAX requests to API.
• Underscore.js:one of the programmers at work used Scala, that is why he felt quite helpless without convenient functional ways of work with collections (at least map/filter). Underscore and some other FP-at gave such a possibility.
• SugarSkull: : is the first version of application that used self-recordering mechanism for permanent links,which are able to work entirely on the client’s side. With entering one more kind of statistics (by tweets of specific user), the URL scheme became more complicated and was added a library. It represents by itself a simple mechanism of dispatching client-side URL.

One of the work advantages entirely in brower was an opportunity to host on GitHub not only software repository but the application itself.

As a result we got work tool for gathering of statistics by twitter-clients.