Anyone here used the Automation Tool module from Epicor? We’re thinking of using it for patch testing and our migration from 10.0 to 10.1 series.
Link to the white paper on it on EpicWeb:
https://epicweb.epicor.com/resources/MRCCustomers/Epicor-ATE-A4-FS-ENS.pdf#search=epicor%20automation%20tool
I’ve been looking that for the same purpose.
I’d be interested in hearing people’s experiences with it.
We’re looking for references as well.
I have used in E9 and have plans to use it in E10 again. Not sure what to explain; it works as advertised. You record your steps, throw in some scripting/code and do that for all of your 400 other Customizations 
and the next time you upgrade, you simply press Play All and get some coffee, come back and see results.
However it requires maintenance, you tweak your customization, you must also tweak the tests to account for that.
If you are using Epicor Web Access you could also make use of Selenium http://www.seleniumhq.org/ since it doesnt come with a hefty price tag.
Also Epicor uses ATE according to them in their development continues testing cycle. @Bart_Elia is that a true statement?
We use ATE internally every day. Build finishes and it kicks off as a part of the continuous build / deployment system we are always improving.
-If you are not doing CI/CD these days you won’t stay in business long. #AutomateAllTheThings
We use TFS for versions control, MS Unit Test and XUnit based on the unit testing needs, TFS build, TFS Code Analysis, SonarQube, ATE for most of the integration testing and I am probably missing some things.
I have not had a chance to review the Mobile CRM dev process but I know EDD was heavily using Postman to test the REST Endpoints for it.
Thanks for the replies.
Our older testing tool was very dependent on display resolution. If the “script” was created at 1280x1024 and over time we upgraded to widescreen and tried to run the test, it’d fail. So all our old scripts were useless.
I understand as the workflow/process changes the scripts would have to be changed accordingly but having to re-record the entire thing for a new monitor got old.
What do you guys think; Is it better to record and entire Quote-to-Cash session or in smaller groups?
Having spent my youth running little automated testing processes (CalPers - a 19 million user test gag)…
Organize your tests like good code - small encapsulated units. Assemble those together for your workflow tests. If something changes in one area - only update the one unit instead of having to update dozens or hundreds of tests.
In my spare time, (bahahahah, ha, ha, whew, yeah, that was a good one), I’ve been looking at BDD, behavior-based development testing tools. The most famous of the BDD tools is Cucumber but it’s really a Ruby-based tool. However, someone made a .Net version called SpecFlow. This testing system uses a testing language called Gherkin where you describe what behavior you expect our of your software. What I like about this, I’m not just unit-testing code but documenting HOW we use Epicor and writing the test for that expected behavior. I have no control over the source code (especially since we’re probably going to move from ST to DT SaaS), so if something stops working, it may be code working as designed but not as we used it. This is the kind of tool I want for the new release cadence. And the really cool thing is that the test descriptions can be turned into HTML or PDF and your documentation is already done.
We’ll see…
Mark W.
Good advice Bart. We signed the PO for ATE, so we’ll have a chance now to build new test scripts from the ground up instead of inheriting someone else’s mess work. 
Does ATE use Epicor field names or mouse positions when running the scripts? AKA, is it dependent on screen resolution?
honestly… have never seen ATE. They just give me requirements and I have enough to keep on top of and we have a whole team that runs that effort. I keep sticking my nose into other subsystems and then I can’t do my day job. And you folks really want me to finish the vNext toys I am playing with. 
ATE uses field names, for example
EnterText(“MESMenu/windowDockingArea1/dockableWindow3/mainPanel1/txtEmployeeID1”, “100”)
So it should not be dependent on screen resolution
That’s how it’s been done since the early tools figured that out - SQA Robot (I was a 1.0 beta tester), MS Test (gawd was that even the 90s when that started?). Anything relying on screen position is too fragile to bother to start. The tools have advanced a ton since the early days. The biggest challenge was goofy UI components that did not expose anything to tooling. Luckily those kinds of components have mostly died off.
How is everyone’s experience with the Automation Tool? We are looking to use it to speed up our testing efforts for testing patches and releases by continually build a library of scripts to test when each patch. This in hope to make our testing faster, and let our users focus more on new functionality testing than regression testing.
To be honest we dropped the module when the company was acquired. The new boss’s said no to getting it.
I was trying to find someone with experience using it awhile back too. Many have asked about it and some even bought it but not aware of anyone actively using it. In theory I think it would be a great tool to use for upgrading and overall testing. I hope to try and get someone’s ear on it in Nashville to pursue further.
Kelly
Hi Kelly,
We are currently evaluating it. They don’t do trial versions but I know somebody who has it and was able to show me more details than the demo. It actually looks pretty good IMHO. It has a recorder part where it creates a script and then a player where you can replay scripts and have it check for expectations.
Mark W.
P.S. And since you’ll be in Nashville, I’ll be talking about this very topic on Thursday morning at 9:05.
@Mark_Wonsil - I will be in Nashvill and will look for your session. Having a hard time locating it on the agenda 
What’s the title and what time?
I just got the automation tool installed on our 10.2.200.4 test server and will spend as much time as I can in it before Insights so I can get any questions I have answered while there.
We haven’t purchased ATE yet because we write software here and we’re comparing ATE with what our Software QA people use.
Mark W.
I will dabble in it until Insights. We want to determine if ATE can be one tool to streamline regression testing for patch and release updates so we can shorten testing time and update more often than the original plan of once a year. If we can script a number of scenarios and let the users focus more on what was fixed and what is new then it is a win win for all of us. Will be interesting to hear your take on it.
Don’t wait. Send me any Q’s as they come up. mwonsil@perceptron.com.
I don’t think there’s a single tool to do all things but an automation tool should be one of them.
Oh, one thing I hear a lot in QA and will say at Insights,:
“Don’t waste time on low-value tests”
Mark W.
