I am wondering if anyone has ever encountered this situation, or have any input into possibilities of working through this very quizzical issue I am seeing.

The application I am creating WR functions against is an IE application where the pages are dynamically created from a format file (every window has one of these files except the main log in screen as the application allows the user to create the interfaces as they desire for their business workflow). Development language is C++/C#.

The issue that I am encountering is one for the identification of the control object on the IE page. I have 2 systems available to me for script and function development. I am in the process of creating a data driven test suite, with also the goal of allowing non-WR users to design automated test through parsing a txt file that can be supplied, so I have one system to develop on, and another to test/run what I have built prior to committing the work to VSS.

The first system is a standard desktop, screen resolution of 1024X768. The second is a laptop with a "HD" screen and a "forced, suggested" resolution of 1680X1050. The objects when learned or spy'd on the first system, they are recognized correctly. But then the application is looked at on the second (laptop)system, when spying or finding an object through the GUI map, moving the mouse over the object (any object - button, edit, list, etc), it highlights the object elsewhere on the screen and not where it is drawn on the page. (Imagine on the screen that clicking FIND in the GUI Map editor and clicking on a button in the lower right of the screen will report to find the object, but highlight it more to the middle of the screen).

This puts a stop to recording scripts, or even playing back scripts or functions as the objects are interacted with at their "ghost" location and not the location where they physically reside on the page. I am ensuring that physical coordinates or index positions are not being used to define any objects in the application, so it is isn't related to the GUI learning the object in a physical location on the screen.

I have also forced the screen resolution on the Laptop to 1024X768, but the same issue persists (and granted makes the laptop almost unusable with the display).

Anyone ever experienced this, or have any ideas as to how to overcome it?


[ 02-26-2004, 12:10 PM: Message edited by: awdavis ]