The "agile approach" is fundamentally different from the classical "throw the software over the wall to the test team and hope for the best" method. Which never really worked, so good riddance. By testing all the time you're developing you maximize the chance of early detection (and repair) of the application -- and we all know that achieving early detection is a big payoff.

The article is light and well done, and a good read. Several good references, too.

In addition to detailing how to get the validation done correctly, this example illustrates some aspects of achieving fully-synchronized playback, and how to work with the eValid PageMap to investigate the contents of the DOM.

Here is the complete description on Validating a Pulldown Menu.

It may be important to note that Nagios has released a new version of the NRPE component that claims "...more graceful shutdown..." among other improvements. Details on how to download and use the upgraded NRPE 2.6 can be found on the Nagios website.

Where this is of interest to us here at eValid is that, as eValid users may know, it too is a kind of customized version of the IE browser -- but with extensive test and analysis functionality built in. But -- and here is the main point -- eValid was not done with IEAK7 and in fact the kinds of technical extensions that are found in eValid are specifically not possible with Microsoft's IEAK7. For example, the eValid feature that allows direct access to eValid documentation online direct from the eValid browser face is just one example of an extension that falls outside the IEAK7 capability.

The important point is that it is very good to see others adopting the idea of a browser with specialized purposes. The eValid engine -- an IE compatible browser with comprehensive test and analysis enablement -- is an ideal engine for website analysis and quality assurance.

Yet mechanically confirming the content of a mouseover popup is often not such a simple matter, as this new Technical Support Note Validating a Mouseover Popup Passage makes clear.

The trick here to "doing two things at once" during a complex recording session is to use the Window Save & Validation Wait First option. You set up the screen capture, and then activate the popup before the timer runs out. Then eValid captures your screen, and you're done.

• Testing of Software and Communications Systems (TESCOM / FATES 07)
  Set for June 2007, this conference is tied to the 7th International Workshop on Formal Approaches to Testing of Software. It aims to consolidate theory and practice in the full spectrum of testing -- empirical and formal methods alike.
• Second Workshop on Automation of Software Test.
  Set for Minneapolis in May 2007, as part of the ICSE event, this workshop aims to "...bridge the gap between theory and practice..." of software test automation.
• Testing of Software: From Research to Practice (in German), set for March 2007 in Hamburg, Germany, this workshop aims to "...describe efforts at applying testing theory to testing practice."
• Testing: Academic and Industrial Conference (TAIC) / Practice and Research Techniques (PART), scheduled for Windsor, UK on 12-14 September 2007, "...aims to forge collaboration between industry and academia" on the problem of software testing.
• TAIC/PART is co-located with the Third Workshop on Mutation Analysis (Mutation 2007), a technique that manipulates program representations to investigate how well tests of programs might fine extant errors.

Sounds like a full plate here, and certainly a worthy goal!

The BCW scores involved 20% on Technical Quality (evaluations done with eValid) and 80% of Quality of Online Experience as measured in a series of three scenarios that contribute 3/4 of the 80%, and an email responsiveness score that contributes the remaining 1/4 of the 80%. The email responsiveness analysis involved a matrix of some 20 different factors.

The final BCW 2006 scores ranged from a low of about 30/100 to a high of 63/100. This material will be summarized in a Financial Mail article that will appear in January 2007 that will name the winners.

The results were based on a survey of 1,058 online shoppers surveyed in the first half of 2006. Additional results showed that one third of shoppers with a bad experience abandoned the site entirely, and 75 percent were likely NOT to shop on that site again.

While poor website performance has always [previously] been a worry, these results appear to be pretty conclusive. Webmasters probably need to take note!

The increasingly popular Google Web Toolkit (GWT) includes a set of pre-defined objects, a "widget set" called the "Kitchen Sink". There is a corresponding Google GWT Kitchen Sink Demo.

As an example of applying eValid to this kind of AJAX-supported object, we've prepared a sample set of tests. Testing the GWT Kitchen Sink shows how eValid handles all of the GWT example objects without difficulty and without any special programming or scripting expertise. The three scripts show:

1. A simple walk through of all 12 groups with simple validations on each one.
2. A complex test of the Trees group, with full validations.
3. An illustration of DOM validation including Adaptive Playback.

These scripts are all "AUTOPLAY" enabled, so you can try them out with any eValid browser (you don't need a license key).

Now Microsoft has jumped onto the AJAX bandwagon big with their formal announcement of ASP.NET AJAX.

Like the Google Web Toolkit (GWT) and Morfik WebOS AppsBuilder suite, among other technologies that are evolving into this market space, Microsoft's offering provides a variety of resources to help developers create a "rich client side UI experience". All of these share the notion that "the browser is the computer" that the users prefer -- a notion made popular in the Web 2.0 developments of the last couple of years.

This feature is primarily intended for applications where eValid is used in monitoring. For complete details, please see the description in Header Processing in Playback.

• 1st Asian Working Conference on Verified Software , Macao SAR, China, 29-31 October 2006. A response to Tony Hoare's grand challenge on "The Verifying Compiler" this conference aims to attack the challenge head on.

• 6th International Conference on Quality Software , Beijing, China, 26-28 October 2006. Aimed primarily at an Asia/Pacific audience, this is one of the continuing series of Quality Software conferences (the two prior instances were in Melbourne, Australia and Brounschweig, Germany).

• Haifa Verification Conference 2006 , IBM Research Lab, Haifa, Israel, 23-26 October 2006. Broadly structured, this workshop aims to discuss both hardware and software verification and testing.

• IEEE International Symposium on Workload Characterization (IISWC 2006) Hilton Hotel, San Jose, California, 25-28 October 2006. In performance engineering, he who holds the benchmark can claim the result. This symposium aims to refine the art of creating how the workload benchmark works.