Major new features of the Newtest 8.0 Robot
This article describes in detail two major new features of the Newtest 8.0 robot. These are the new video diagnostic format NVF, and the new function ClickOnObject.
Facilitate the analysis of errors
The new video format, NVF, is a proprietary video format which makes differential captures of the Windows desktop in .bmp format by the robot while the script is running. This format is associated with an ip-label proprietary video player, the NVFviewer. This format was designed to replace the existing AVI video format.
Because the capture format is .bmp, no compression is applied to the image, unlike AVI format. This has the double advantage of allowing the image to be exploited in its original resolution (for scripting, for example), thus alleviating the workload of the robot’s CPU (see the benchmark below). In exchange, however, the NVF format can result in a larger file (+30-50%) than classic AVI format. As soon as it is activated, this should be counted toward diagnostics storage space on the NMC server.
The concept of ‘differential’ capture optimizes the volume of data specific to the video. A user’s desktop is gridded off virtually into segments of the same size. The segments of the desktop which undergo no visual impact between two captures is not stored. The entire NVF video is rebuilt by a special viewer, the NVFviewer, which simulates real-time playing of the video at the moment it was recorded. The NVFviewer client is made available to NMC users in the ‘Download’ area.
Beyond its basic function as a video player, the new NVF reader can embed application logs and play them in sync with the video, as the example below shows: trace logs scroll at the same time as the video plays in a separate window. They also resynchronize when forwarding or rewinding step by step (Prev. – Next).
Optimization in the use of robot resources
Furthermore, this new NVF video format brings an important optimization in the use of robot resources. This is especially true for multiple-session server usages. Experience demonstrated that the 7.6 robot, when using this solution, consumes significant server resources when the video diagnostic in AVI format was activated, to the point of affecting the response times under measurement. Below is a brief comparison showing consumption and performance in multi-session on a VM running in AWS of type M4.4xlarge (16 vCPU/64 GB RAM) with 15 Newtest robots (maximum number of robots for a server) simultaneously playing a scenario connecting www.google.fr every minute without diagnostic, and then with diagnostics enabled in AVI and NVF formats.
As may be seen from the new video capture method corresponding to the last test, the gain is very significant in that it does not affect the reference times (without video), while giving a very good return on vCPU consumption (about 1vCPU per robot).
The strength of the Newtest robot is that it can perform a same action in lots of different ways. One new function now enriches the panel of the Newtest robot’s existing functions for interactions with the Windows environment. The function ClickOnObject() enables simple, precise interaction with any kind of object in an application.
When the wizard of function ClickOnObject() opens on the step of your choice in the scenario, the method for detecting objects consists of dragging the target iconto the application window.
If an object is detected when you drag the target icon, an outline will appear on object and its reference will appear in the field ‘Title’. See example below for the IE address bar.
The possibility of interacting with the button by clicking it can now be configured.
Also note that the (X,Y) coordinates corresponding to the offset in pixels with respect to the center of the found object are saved so that the click is always at the same position. This can turn out to be very useful in cases when the object consists of links or buttons that were not detected individually, as in the case of a menu bar, shown below: