3 The relationship between SnowPlow.js and Piwik.js

HOME > [SNOWPLOW TECHNICAL DOCUMENTATION](Snowplow technical documentation) > Trackers > Javascript Tracker > The relationship between Snowplow.js and Piwik.js

snowplow.js is based on Anthon Pang's excellent [piwik.js] piwikjs, the JavaScript tracker for the open-source Piwik piwik project, and is distributed under the same license ([Simplified BSD] bsd). For completeness, the main differences between snowplow.js and piwik.js are documented below.

Main differences between snowplow.js and piwik.js

Some of the differences are as follows (this list is very out-of-date):

• Simplified the set of querystring name-value pairs (removing Piwik-specific values and values which CloudFront logging gives us for free)
• Tracking is now configured with an account ID (CloudFront subdomain) rather than a full tracker URL
• Added new trackEvent functionality
• Added new trackImpression functionality
• Added browser language and visit ID to the querystring (because not available via CloudFront logging)
• Removed POST functionality (because S3 logging does not support POST)
• Removed goal tracking functionality
• Removed custom variables
• Removed ecommerce tracking functionality (ecommerce tracking is now handled by events)
• Removed piwik.js's own deprecated 'legacy' functionality

We expect these two scripts to diverge further as we continue to evolve Snowplow (see the next section for more detail).

Roadmap

piwik.js provides an excellent starting point for snowplow.js - and we would encourage any Snowplow users to check out Piwik piwik as an open-source alternative to using Google Analytics.

However, we fully expect snowplow.js to diverge from piwik.js, for three main reasons:

• Tracking technology: there are some differences in what is advisable/possible using a PHP tracker (like Piwik) versus using an Amazon S3 pixel (like Snowplow)
• Approach to data aggregation: Piwik performs 'pre-aggregation' on the incoming data in both piwik.js and the [PHP tracker] piwikphp prior to logging to database. The Snowplow approach is to defer all such aggregations to the MapReduce phase - which should reduce the complexity of snowplow.js
• Philosophy on premature analysis: Piwik follows the 'classical' model of web analytics, where the sensible analyses are agreed in advance, formalised by being integrated into the site (e.g. by tracking goals and conversion funnels) and then analysed. Snowplow views this as 'premature analysis', and encourages logging lots of intent-agnostic events and then figuring out what they mean later

Planned items on the roadmap are as follows (this list is very out-of-date):

• Remove site ID functionality
• Remove unused campaign marketing variable code (as no longer used)
• Rewrite in CoffeeScript (joke!)

Copyright and license

Significant portions of snowplow.js copyright 2010 Anthon Pang. Remainder copyright 2012 Snowplow Analytics Ltd.

Licensed under the [Simplified BSD] bsd license.