Sunday, March 26, 2006

Recorder Controversy

I have not posted in quite some time, because life and real work seem to get in the way. We are still using WATIR as our primary automated testing tool; however, some of the limitations are forcing us to reconsider this decision. I know the team is working hard to get to release 1.5, but that is at least 5 months out. Modal and pop-up interactions are key to the success of WATIR interacting with our applications.

I thought I would take a few minutes to post on the controversial thread regarding the WATIR recorder. First and foremost, the recorder does not work at all for me. Everytime I attempt to record a basic sequence of steps the recorder locks up with I believe a WIN32 OLE error and I have to manually end the application. So it is not a viable option at all for us. The second reason I do not use the recorder is that the syntax is different than all of our previous development, which makes it just plan confusing to persons new to WATIR.

On the forum there were some critical posts. A recorder versus the QA purist are like oil and water. My adventures with WATIR and Webload have moved me to the code first record later approach. The more I understand the code the more power I can get out of the tool.

Webload came out with a new version of their tool to be more GUI focused. Although the new release had some needed enhancements the upgrade created havoc and we could not easily leverage all of the work we had done the previous two years. The tool became useless. It probably was not the tool itself, but I could not figure out the dotted lines between the legacy code and the new tool. I was truly frustrated. I probably told this tale before, but it is something that will stick with me.

For me the WATIR recorder poses similar challenges, so I am opting to stick with the coding approach and not use the recorder. I still have a ton to learn, but as time permits I truly enjoy taking on the challenges. The one thing that does frustrate me about WATIR is that I am not a programmer nor do I have enough knowledge to actively assist in the efforts to get WATIR 1.5 released. My hat is off to all of those persons attempting to make this happen.

I am also amazed that the WATIR community is attempting to stay of the curve and looking at IE 7.0 or Firefox. The need is coming and my guess is the more expensive products on the market are behind the curve.

I digress a bit from the actual topic which is the Recorder controversy. I regret that this flared up to the extent that it did. We the users of the WATIR forum could have helped to alleviate this fury by simply being more specific and descriptive in our posts to the forum. I am guilty of not being clear in my posts, but I am learning. One simple solution would be to put in the header of your post the word, "Recorder", if it was involved at all in your script development. Marcus could have quickly glanced through the forum and immediately got involved.

We obviously have at least two different WATIR communities those using Recorder and those not. It is a precarious situation when open-source collides with tools that must be purchased.

I commend Marcus for attempting to develop a tool useful to WATIR. I also respect the domain and efforts of the open-source community. Basically I am caught in the middle.

Taking the baseball bat away from the playing field was not the correct response nor was personal attacks. I hope everyone takes a step back and works toward a compromise. Separate forums may certainly make sense. Hopefully you forum guru's will create a separate forum, but allow these forums to mutually coexist.

I for one appreciate everyone's efforts into the WATIR community. My vote is to live in peace and continue to develop the greatest thing for the Agile QA community.

Sunday, February 12, 2006

WATIR Recorder

If anyone stumbled on this blog I apologize for not posting more often. I hope to publish some troubleshooting techniques soon. Here is a tid bit off the users group that might help some folks get started using Ruby.

I have not fully evaluated the tool, but it seems intuitive and can give some a quick framework to work within.

I am still hampered from fully implementing our suite of automated tests. I am not able to efficiently interact with windows to download and export files. As soon as we accomplish that I think we will be in good shape with WATIR.

I also ran into a problem this Friday were I can not run multiple suites of test cases. On error occurs then all of the scripts crash opeing up numerous browser instances. Seems like I can run them a suite at a time, so that should be sufficient.

It is late! If you have not tried WATIR for functional testing you need to give it a shot.

Good night!