Why use testStand

[dusty_g]

I have two questions.

1.) Why should I use TestStand instead of developing my own GUI that I have complete control over?  What are the Pros and Cons of using TestStand?

2.) When using testing let's say I create a profile for a Technician to have limited access and to where he can't change the code.  During test I have TestStand configured to stop on failure.  Will I be able to log in as an Admin and begin to troubleshoot or will I have to finish the test as a technician and then log in as an admin after the testing is finished?  This is very import for me to do during a test.  Especially since some of our tests can take a very long time. 

[paulmw]

To question #1, With TestStand I have found that you can still have "complete control" over the GUI.  It all depends on where you want to start.  I think the biggest pro that TestStand has is it gets you a jump start into developing a test without the initial overhead of developing the "common" functionality (reports, loops etc).  This jump start assumes that you are not going to change some of the underlying, out of the box, functionality.  For example, you can start testing right away if your needs can utilize the existing process model and Operator Interface (OI).  With that said, TestStand still provides all the options you would need to create your own test environment built on the TestStand API.  You can create your own custom interface with the operator.  This can be developed in multiple languages (LabVIEW, C++, C#, etc).  You can create your own process model specific to your needs.  So, don't think that TestStand will impede you from having full control.  It will just be in a different format.  As for Cons, I'll leave the meat of this to someone else.  I have not found any Con for using TestStand (except maybe the price)

For question #2, Out of the box, I am not sure if it will perform exactly as you describe.  A well written, custom operator interface and/or test sequence, should possibly perform the tasks.  I question at what level you will need to "change the code" after the test has started or if you are just looking for debug tools (such as pausing, stepping, changing flow and interacting, etc) which TestStand definitely has.  Also, you describe a "stop on failure" but indicate in the next sentence not having to finish the test.  In this example, would it be fair to characterize that are you describing more of a pause on failure condition?  In which case, you login in, while paused, and perform any debug capabilities authored under the permissions of your login level(?).  For the example I just describe I think that changing the user via the login-logout mechanism built into the OI and sequence editor will terminate the current execution.  This doesn't mean that you can not change the user and permissions; just that you will have to develop a new means, again, most likely a customized OI.

[paulmw]

I want to correct my statement about changing the user terminates the current execution. I was very wrong! I was mixing up a custom OI that had been developed that I currently use (which does this, bad design in my opinion) and with the shipped OI that was switching to the sequence view when I changed users during a paused execution. So, yes you can change the user in the middle of a pause execution and the OI will switch to using the permission of the new user. Sorry to anyone for the confusion. P.S. the report will show the user that launched the execution not the one that finished it...

[j_dodek]

Hi dusty,

1.) You should visit this thread

     http://forums.ni.com/ni/board/message?board.id=330&message.id=24709#M24709+

2.) Till the upper post was posted i had  no idea if dynamical user changing in a exectuion is possible

      so you see: yes !

... using Teststand(forum) also means a lot of people will help you solving your requierements.

Greetings

juergen

[dusty_g]

Thank you very much for your replies.  your are correct, I meant pause on failure, and not to really change anything but more for debugging and exploring.  I wouldn't be allowed to change the code anyways with out a proper review.  It's good to know that you can do this though.

I will explore the other forum about using teststand over all.  Paul I know you state that you can pretty much create your own OI in any other language, but that still leaves me to the "Why use teststand".  If you're creating your own OI why jump through the Teststand hoop.  I'll continue to investigate this.  I don't have a coding background so all the times I've used it, it's worked for me.  However all the software engineers on my team seem to dislike it.  I was just trying to understand that more.

[Dennis_Knutson]

The OI is just a small part of TestStand and not close to being the most significant reason, imho.

TestStand allows you to concentrate on writing tests and only tests. You do not need to write any data logging or report generation software. You not have to write any code to check limits. You do not have to write any code to import settings/limits from an external file. You do not have to write any code to test multiple UUTs in parallel.

I have no idea why your software engineers would dislike it. Have they ever given you any reasons?

[dusty_g]

it's always just seemed lilke they would rather program in C, C++, Java, etc.  Typical things they've said is it's slow (which i agree) and it's buggy.  We haven't used version 4.2 yet though.  The team will be having a meeting tomorrow to decide if we're going to go with using teststand or not.  I'll here more reasons why they don't want to use it then.

[Dennis_Knutson]

You mean that they would rather code the UI and all of the auxilliary functions in C, C++, Java, etc. All of the code modules are written in whatever language you like. If you are at all interested in increased productivity, then I think TestStand is an easy choice. I've been using it since version 1 and have found it to be very bug free and not at all slow. The speed of execution seems to be more a factor of how the test steps are written than anything else, imo.

[RayFarmer]

Like Dennis, I have been using TestStand since version 1 and like all software some issues are better than others, such stable version are 2.0f1, 3.5 and 4.1.

Version 4.0 wasn't nice but that's when the new style SeqEditor appeared.

You didn't say what version you are using but if you have problems then you need to report them.

Maybe your engineers need to look at the bigger picture rather than looking at there own tiny worlds.

I dont think you said what you have been using or are going to use TestStand for but some application are more suitable for TestStand than others. Test solutions that generate a lot of results or have a very long runtime need some careful design considerations.

The default test reports is never the best thing to use and it is usually top of the list to change. Enhancement in TestStand 4.2 should make it easier when it comes to organising the reports generated.

As to whether you use it or not, well that depends on whether you want to adopt a standard approach to your Test solutions within your company.

Regards

Ray Farmer 

[dusty_g]

Right now the system is baselined to use TestStand 4.2 and LabWindows 9.0.  We will be using TestStand to call the CVI testscripts that will be running very length tests on our UUT.  Some of our UUTs have over 3000 I/O to exercise, and schedule is a big concern.  The faster and more successful we can implement this testing software the better.