Forum Discussion
Ok - I have been through many different scenarios, modifying various combinations of settings that you have suggested. I have the following findings:
- In general, any changes in performance are slightly noticeable, if at all. There is nothing that I have tried yet that has a significant impact.
- I checked and unchecked the "Extended Find" option in the Option->Engines->NameMapping panel, with no effect. There did not seem to be any performance change. I have not tried checking the individual "Extended Find" for any objects under the NameMapping object of the project.
- Towards the end of the day today, I removed all objects from the NameMapping object except the main application. This had little or no effect.
- I can only get the procedure to run to completion when and if the Auto-Wait parameter is set to higher than 10 seconds. I have it set a 100 seconds (100000 millisecs), which is higher than I need. The worst case wait is about 30 seconds.
- Many of the suggestions seem counter-intuitive to me. For example, would not a breadth-first search be faster than a depth-first search if I am manipulating objects at the GUI level? Would not a search for visible objects first be faster when manipulating at the GUI level?
- You mentioned adding "...additional "parent" objects in your tree to reduce the amount of searching needed." I think I understand the idea behind it, but how do I add a parent object to a strategic position in the tree? What type of object is being added and how do I attach the children?
I am working on a network that is isolated from all other networks. Getting things like screenshots for someone to look at is problematic at best. However, I am going through the "Testing Desktop Applications" section of the "Getting Started With TestComplete 12: tutorial from SmartBear. I have tried to step through the example in the tutorial with little or no changes. The one thing that I am doing differently is that I have a .bat file that launches the C# version of the Orders program. Otherwise, I believe I have been consistent in following the tutorial.
I am about to head out for the rest of the week, but will be back early next week to hit this again.
Thank you for your help, Robert.
I think I am finally getting to understand this product a little better. It would appear that it runs faster as it "learns" through name mapping. Once that name mapping structure is defined and saved, things run a lot faster. So the trick is to define the structure up front, which may take more time. However, once defined, subsequent test runs are much faster.
No?
Related Content
- 10 months ago
- 5 years ago