Automation of IBM systems ("green screen")

I currently work (in part) on IBM based systems - the “green screen” style, back office, type applications.

Wondering if anyone works in similar field?
Specifically interested in whether anyone has successfully managed to automate in this space?

We have some automation mainly around XML processing (create XML file with specific data, drop file to MQ, read database condition post-processing to check data as expected), but not yet managed to actually drive the screen. Anyone had any joy?

Hello @pmfrench!

The transaction you described wouldn’t seem to need a screen if you have access to MQ and the database. Unless there is a need to validate screen content, other tools could be considered for this transaction.


Hi Joe,

You may have misunderstood (or perhaps I wasn’t clear). I meant that we have successfully automated certain elements of our applications (xml processing) but not yet managed to automate screen interactions.


Thanks @pmfrench!

If the elements of your applications (xml processing) have been evaluated through automation, what is the purpose of automating screen interactions?


1 Like

Hi @devtotest
There’s a few reasons this might be useful.
A lot of the user interactions are still UI driven so just having some automated checks of these would be useful. But we also have some batch process triggers which require manual actions (e.g. xml message places an order but a scheduled batch job allocates and processes it).

What kind of green screen do you mean? Terminals? I would love to look at the value of using a screen scraper to simply scan and write a wrapper to validate the workflows you care most about are showing correct data as you are seeking. I know very little of what your test case in/out criteria look like, so I might have this all wrong, but its teletype(ansi code driven), or is it direct screen writes that need scraping if you want to test the screen?