H5 Scripts
H5 scripts, written by describing them.
An H5 script is the cheapest change you can make to Infor M3, and the one that usually waits longest for a developer. LoupeCode writes it from a sentence, and because scripts live inside M3 screens, you show it what you mean by clicking the screen itself.
Point at the screen. Say what it should do.
Describing a location in an M3 screen in words is miserable work. Panel names, grid positions, field ids that look nothing like the label above them. So you do not describe it. You open the screen and click.
-
Place buttons by clicking. Want a button next to the item number? Open the screen, click the spot, and LoupeCode puts it there. No panel names, no coordinates.
-
Link real fields from the screen. Pick the fields your script should read or set by clicking them in M3. The technical names and the labels come along for the ride, so the script refers to the same things you do.
-
Say where it applies. One panel, one program, or everywhere. A script can be scoped as tightly as the change deserves.
-
Preview before you publish. Test the script against your own M3 and watch it work on the live screen. Publish when it does exactly what you meant.
The jobs scripts are for
Small, specific, and usually urgent. The kind of thing that never justifies a project and never stops costing you.
Stop the mistake
Warn or block when a discount goes past what the customer is allowed, before the order is saved.
Show what is missing
Pull a value from another program and put it on the screen the user is already looking at.
Save the clicks
A button that calls a transaction and fills in the six fields the user types every time.
Add a column to a list
Put the field the business keeps asking for into the list where they need to see it.
Scripts that survive contact with H5
H5 scripting has sharp edges. Get the command handling wrong and refresh stops working, or paging quietly breaks, and the script looks fine while it does it. LoupeCode is built around the patterns that hold up in a real client, and every script it writes is checked against them before you see it. Where a generic AI gives you code that reads well, this gives you code that keeps working next month.
Belongs in the engine?
Validation that must hold for every channel is XtendM3 work, not script work.
How does it compare?
Against a consultant writing the same script, or a platform renting you one.
Run it in your environment.
Bring an environment. We'll show you what one sentence can build. Already on Loupe? Just ask and we'll switch it on.