I developed a entire process with CV Scope activities and hot keys. Here the thing is in my laptop it is working fine but when i try to run in other system it is going to fail by clicking on some elements.
In my laptop also sometimes it is working fine but sometimes it is throwing errors, i tried modified the descriptor values how much can do, i did all for that but still it is going to fail.
Here i want to know that
Can we expect 100% bug free outputs for this process?
Can we change the machines to run this process without changing any resolutions?(in QA machine, Staging Machine, Production Machine)
These all issues are faced when i am run execution from studio and attended mode only.
CV Scope and hotkeys are resolution-dependent. If you change machines without matching the screen resolution and scaling, the automation may fail.
Make sure you match the resolution while deployment into production server. Add 1920 in resolution width & 1080 in resolution depth for full HD. If your dev server is also of same resolution.
In screen i have a table in that i have to click on each row then i will get a pop up, again i have to close the pop up after extracting the data. Data extraction is going smooth.
But the problem is in 1st run entire execution will run without any issues but after 2nd run or 3rd it will throw error to click on each row item.
worked with combination element exists and delay at that time only its working.
is there any other solutions.
And i need proper functionality points to explain to my team.
How CV will work and how much complexity it have.
Element exists is the basic and must have(in my view) to have robust processes.
For my processes(Cv or not), I always include element exists.
Delay also is required to use sometimes and that is totally fine. I usually try to create an asset for delay(instead of hard coded value) so I can come trip it from orchestrator and enable/disable it any time.
Also, anchors is a good add. I sometimes use multiple anchors to identify a target elements. More chances of success.
Additionally, all CV activities have input/output regions that can be used.
For e.g. if before clicking you are checking element exists, you can create an output region for it and then in next activity, simply use that value in the input region.
And highlight also comes handy because you can see where the control is flowing on the screen.
I have used CV in 2 projects as of now and results have been fine. I do face few issues sometimes but retrying those transactions usually work.
My observations so far:
process with CV selectors takes little more time to run than the one with regular selectors.
less number of failures and retries required.
CV relies on screen resolution so we have to consider that and develop process in a way that any temporary banner additions in the site doesn’t mess with its working.
