hi pylon53, I have the same issue. If I try the example for frozen headers from  Base SAS: Creating a Data Grid Like VB.NET, it works from SAS Eguide (program running running on unix server). But if I use it in a web-application STP (file=_webout), it doesn't. In the generated source is NO javascript. Did you get any further? thanks.   ... View more

I know this is a year and a half late, but I had a similar issue and just found a solution! Create a new 'text' prompt _metauser - I've made it hidden, required, and used throughout the project.  Then in the default field put: &_metauser.  Now in the dynamic prompt, you can make it dependent on the _metauser prompt.  Only the departments that metauser is associated with in your registered table should show up. I'm using SAS 9.4 and EG 6.1 - not sure if it works the same for other versions.  It seems to work in both EG and the stored process web application for me. ... View more

"...or DI Studio doesn't behave the same as SAS" I'm picking this bit from your post because it's important that you understand what DI Studio does and what "SAS" does. DI Studio is a client where one can define ETL jobs in metadata. Based on this metadata DI studio then creates SAS code which gets executed on the SAS server. So it's always the same "SAS" which actually executes. What you need to do is to look at the code generated - and eventually the log - because this will tell you what and why things are not working. Creating columns via code (eg. using Proc Transpose) is quite easy in SAS but it's kind-of an issue with any ETL (and also databases by the way). You need (and want) stable table structures to map columns from source to target. You can of course re-create the table always when you're running the code and change the table structure (number and names of columns) every single time. If you then also need to have table metadata for this you can execute Proc Metadata which can create or update the corresponding table metadata for you. But what are  you going to do with this table metadata then? The new columns won't be mapped to anything in your flow. You might need to re-think your approach in an ETL context and try to go for a long table structure instead of a wide one (=have a year (or date) column and several rows instead of several columns). ... View more

I think you only have to use the -noterminal option if you are running batch on Unix. ... View more