The simplicity thesis
The simplicity thesis
My thoughts on automation engineering
History
Tools and recommendations
• I‘m involved in the Selenium project since 2008 when I have started as a beta tester and then just jumped in to the community committing new stuff and fixing some scriptaculous bugs in the selenium core. Since then I‘m am an active member of the Selenium community. In the end of 2011 I have started the project of founding the first Selenium Meetup Group in Cologne Germany. After some difficulties in getting sponsorship and an ideal location, it seems that the group will finally go live in the 4th quarter of 2012.
Limitations of Selenium RC
• Even while Selenium RC is a great tool, it has its limitations. It its a javascript based automation engine with a lot of security drawbacks imposed by the new browsers. I believe that Selenium has reached its limitations and it is not able to follow special new features that new browsers provide. For this reason I do not encourage any new implementation of a test framework using Selenium RC.
• I strongly recommend the use of Watir-WebDriver as a tool that implements WebDriver in its Kernel. Watir-WebDriver is written by the same guy that implements the Ruby bindings and it just works like a charm when using Cucumber as BDD-Framework.
• For easy and elegant support of the page object pattern I strongly recommend the use of the page-object gem
• Another tool that has evolved a lot in the past few months is Geb. It is written in Groovy and also built on top of the Java bindings for WebDriver. It was demonstrated in the second Selenium Conference 2012 in London. It looks like a very promising tool for advanced web automation tests. I would also keep my eyes on that.
• Something like GhostDriver should be used for tests that doesn‘t require any heavy javascript validation.
• Test scripts are test scripts and should be readable by anyone involved in the development process.