1. Internet Explorer
2. Firefox
3. Chrome
4. Safari
5. Opera

But then within each of these browsers there are different versions available to the general public.  Just because the company comes out with a new version of the browser doesn’t mean that people upgrade to it.  So lots of people will be visiting your website using different versions of these browsers.  So what do I test and support?  Two things you should monitor: #1 What are the Global Internet Statistics and #2 What are your actual customers using.

#1 What are the Global Internet Statistics

It is always advisable to monitor what is going on with technology and where its headed.  So you should monitor what the global traffic is like so that you can gage the direction of technology usage.  In doing so you should use multiple sources in order to get a non-biased result.  I use 5 different sites, take their stats and identify an average.  The sites I use are:

StatCounter GlobalStats
Stats are based on aggregate data collected by StatCounter on a sample exceeding 15 billion pageviews per month collected from across the StatCounter network of more than 3 million websites. Stats are updated and made available every 4 hours, however are subject to quality assurance testing and revision for 7 days from publication.

NetMarketShare
We collect data from the browsers of site visitors to our exclusive on-demand network of live stats customers. The data is compiled from approximately 160 million visitors per month. The information published is an aggregate of the data from this network of hosted website statistics.

W3Counter
This report was generated … based on the last 15,000 page views to each website tracked by W3Counter. W3Counter’s sample currently includes 47,262 websites. The browser market share graph includes data from all versions of the named browser families, not only the top 10 as listed below.

StatOwl
Our data is made up of an average of 28 million unique visitors per month to the network of web sites that we collect data from. We do our best to collect data only from web sites that have a broad appeal taking things like target market, audience, geographic location and type of web site into consideration. We attempt to cover a good mix of all of these indicators to provide as accurate a look as possible into Internet browsing trends.

Clicky
Marketshare is calculated from over 300 million daily page views across the 300,000+ web sites that use Clicky Web Analytics.

w3schools.com is another site however their monitoring is based on their traffic usage and since they are a technical site they admit their stats are skewed towards a technical audience which is willing to experient with different browsers and use tools that the average user will not use.  This holds true as you will see their stats are very different than they other 5 sites, therefore I don’t use their stats in my analysis at my fulltime job since my customer base is non-technical.

As of September 2011, the global stats from each of the 5 sites are:

Averaging them together presents this view of the Global Average Percentage on Browser Usage:

Now it is important that you track your customer usage and add that to your monitoring.  This stat can be gathered from tools like AWStats, Google Analytics and your own logging in your database if you process user records.  When you add this information to the graphing grid you will get a good idea what your customers are using and whether they are trending like the general internet traffic or if they are slightly different.  It is very important to remember that what your customers are using is the most important stat.

Browser Versions: Now the next bit of analysis is to drilldown into the browser versions.  In looking at the global averages, Internet Explorer and Firefox make up about 70% of all the usage.  Since you can come up with all sorts of older version combinations out there, I only deal with older versions for Internet Explorer and Firefox.  The other browsers (Chrome, Safari and Opera) I just test the latest version available on the market.

If we look at Internet Explorer you will notice that once you hit IE6 the usage really drops, therefore we only support back to IE7.

In looking at Firefox the two most used versions are Firefox 6 and Firefox 3.6.  The odd thing is that Firefox 3.6 has so much traffic but that can be attributed to the fact that Firefox is available on both Windows OS and Linux.  I’m not positive on this but the opinion floating around is that until recently, the latest “common” version on Linux was 3.6, but with the release of Firefox 7 in October, Firefox upgraded Linux too so this stat should shift once we get in the monthly stats for October.

Testing: now how do I handle supporting these various browsers.  Here is my strategy – our development team is Linux based therefore they only use Firefox.  So during their coding and unit testing they will see how the site looks in Firefox so this takes care of testing on the latest version of Firefox.  Since the QA team is primarily testing the UI through browsers (backend testing is different) and not tied to an OS, the entire QA Team is on Windows and uses Internet Explorer (we also do this because we use the Quality Center and it only runs on IE).  So we have Firefox and IE taken care of.  Now for the other browsers and older versions I have a PC test lab consisting of many PCs with different configurations and containing various browser versions on each.  On the test lab wall is a grid listing what is installed on each PC so if someone wants to test IE7 they can look on the grid and see which PC has that version and they can then log into that machine to test the site on that older version.  Likewise I have each PC’s configuration also taped to its monitor for easy reference.   Depending on the OS, I have several of the systems setup so that the test team can perform a Remote Desktop into the machines remotely from their cubes.  This way they don’t have to physically go over to the Test Lab.  I’m using physical machines because I was able to utilize hardware that was rotated out of regular inventory but fine for our use since they are primarily used just for internet access.  If you don’t have physical hardware to use, another great option is to something like VMWare on a server that can be setup with numerous virtual instances to simulate numerous machines.  This is a very effective way of “creating” the effect of having many different configurations, remote access and easy maintenance for the IT team.

If you need older browser versions, a great site that contains tons of versions to pick from for FREE is OldApps.com (note: I have not had a problem with that site but you download apps at your own risk.  Always scan them for viruses and adware prior to running the install program).

Issue: All of a sudden QuickTest Pro stopped recognizing the web browser and saw it only as a winobject (like a regular GUI window).  So none of my actions on the web page would get recorded, and my test script would be empty.  Obviously this is a major problem.

Solution: The solution to the problem was turning the setting “Enable Protected Mode” off in Internet Explorer.  To do this perform the following steps:

1. Open Internet Explorer
2. Select “Tools > Internet Options”
3. Click the Security tab
4. Make sure “Enable Protected Mode” is unchecked so that it is off
5. Click the OK button
6. Restart IE