I just remembered = there's a QP2SHELL2 - this one runs in the caller's
activation group - with ILE this will matter, as you know.
QP2SHELL starts PASE in a new activation group, it says.
I also looked up QP2TERM - seems it actually starts up a batch job, but
it uses the display for stdin, stdout, and stderr.
Not sure that matters!!
On 6/26/2014 9:56 AM, Richard Schoen wrote:
I have an interesting one.
We have a CL command that calls an OPM CLP that then calls a pase shell command via a call to QP2TERM.
This seems to work fine if run multiple times concurrently from a CL process and the PASE session seems to end fine which is what I expect to happen.
Customer has written an RPG APP where they are building and calling the CL command concurrently from within an RPG ILE subprocedure.
On the first call the command returns fine, but subsequent calls tell them that PASE is already active.
If they call QP2EndPase from the RPG before calling our API then things work fine.
Does anyone have an idea why the Pase environment would stay active in this scenario ?
I'm guessing the Qp2EndPase method was made available for a reason :)
Any thoughts welcome.
RJS Software Systems Inc.
Where Information Meets Innovation
Document Management, Workflow, Report Delivery, Forms and Business Intelligence
Web Site: http://www.rjssoftware.com<http://www.rjssoftware.com/>
Tel: (952) 736-5800
Fax: (952) 736-5801
Toll Free: (888) RJSSOFT