Browser
Selection
To shortlist browsers & their OS versions, we can consider a
combination of data from
- Browser Trend Analysis
- Technical analysis to remove browser redundancy
- Risk Based Browser/OS matrix
Output of Browser Trend Analysis can be used to analyze the similarities
between different browsers based on
·
Layout Engine
·
Shell Features
·
JavaScript engine
·
Plug-ins
Different browser versions can be grouped and few of them can be
considered for the execution scope, based on the risk attributes of the browser/OS combination &
prioritizing them.
Risk Based OS – Browser Priority
To identify each of the supported browser’s priority, Priority = (probability of the Bug occurring) x (Expected loss)
Probability of the Bug
Variety of factors has to be considered for estimating the “probability” of a bug being discovered in a website while using a particular browser.
· Age of the Browser
Age of the browser is
meant, when the most stable browser version was released to the market/users. Ex:
- IE 6.0 is released in 2001 and is the default browser shipped with Windows XP,
which used to have a large market share.
Probability of Bug in an IE 6.x version is higher, due to the lack of
compatibility with modern web standards.
·
Default Browser used by the Dev
Team(Example – Firefox 3.6.x)
·
Similarities between
Default(Example – Firefox 3.6.x) & the Targeted browsers(Example – Firefox
3.6.y, Safari 5)
Expected Loss
Estimating the “expected loss” can be found out by the popularity of a browser on which the app is mostly run (No. of users & Type of usage) and the stability of the browser with respect to the client technologies used.
An application can be used by users who are Internal/External/Both.
This type of usage along with the probable number of users who access the application decides the impact of the browser /OS version (Expected Loss)
The analysis ((probability of the Bug occurring) x (expected loss)) will result in the identification of priority as “high”, “medium” and “low” categories for a particular application.
|
Browser**
|
Probability of Bug
|
Expected Loss
|
Priority
|
|
IE 6
|
High
|
Low
|
Low
|
|
IE 7
|
High
|
Medium
|
High
|
|
IE 8
|
Medium
|
High
|
High
|
|
IE 9
|
Medium
|
Medium
|
Medium
|
|
Firefox
3.x
|
Low
|
Medium
|
Low*
|
|
Firefox 4.x
|
Low
|
Low
|
Low*
|
|
Firefox 5.x+
|
Low
|
High
|
Low
|
|
Firefox 11
|
Medium
|
Medium
|
Medium
|
|
Firefox 12
|
Medium
|
Medium
|
Medium
|
|
Apple Safari 5
|
Medium
|
Medium
|
Medium
|
|
Google Chrome 12
|
Low
|
Medium
|
Low*
|
|
Google Chrome 13+
|
Low
|
High
|
Low
|
|
Google Chrome 18
|
Low
|
High
|
Low
|
Low* - Low Priority/Importance browser.
** - For the above analysis, Microsoft Windows OS is considered & new versions of Browser/OS/devices might have been released by now.+ - Since Firefox & Google Chrome browser versions are getting iterated fast, Test coverage is recommended for the latest major, stable version of the browser & analytics data from your web application.
All the credits for this post go
to,
Matt Archer - mattarcherblog.wordpress.com
Brent Strange - http://QAInsight.net
as I have just collated the information from their sites, to give a basic idea on browser
selection.
Both of them have posted detailed & more analyzed
information on Cross Browser Testing.
For info on, Risk Based, Cross Browser Testing with Scrum, Please refer
For info on, various browser’s technology
background, Please refer
No comments:
Post a Comment