Resolved: Condition result differs for attributes and containers!

[Mike Pokraka]

Hi Jeff,
Thanks, though we're using an evaluation path directly, and the 
evaluation path is incorrectly (?) defined to include persons in the 
result. It's quite straightforward, but finding the blighter was another 
story!
I don't know if it helps you any but whilst trawling OSS for answers I 
did spot a note or two for the WF Admin / Work Center problem you 
mention, don't have them in front of me ATM, but should be easy enough 
to find.
Cheers
Mike


Jeff Rappaport wrote:
> Mike,
>  
>     This sounds similar to the problem with using a Work Center with 
> mixed types as the WF Admin (SWDC). See if it's calling FM: 
> *SWI_GET_USERS_OF_ORG_UNIT* & this could explains why it's happening.
>  
> Jeffrey A. Rappaport
> Business Workflow, LLC
> Voice: (732) 355-0123
> Fax:    (509) 757-4144
> Email: Jeff at Business-Workflow.com <mailto:Jeff at Business-Workflow.com>
> Web:  www.Business-Workflow.com <http://www.Business-Workflow.com>
> 
>     ----- Original Message -----
>     *From:* Mike Pokraka <mailto:wug.replies at workflowconnections.com>
>     *To:* SAP Workflow Users' Group <mailto:sap-wug at mit.edu>
>     *Sent:* Monday, September 12, 2005 12:09 PM
>     *Subject:* Resolved: Condition result differs for attributes and
>     containers!
> 
>     Grumble... after testing this on another system I found the answer to
>     seeing OBJ.ATTR having a few entries but still testing false for their
>     existence. Since it's a sneaky combo of two buglets (= "not sure if
>     it's a
>     bug or not") I thought I'd share it:
> 
>     There's an evaluation path WF_ORGUS ("All users for an object of type A,
>     C, O, S, or US"). Fairy nuff methinks, you feed it any org object and it
>     spits out US-users, right? Not quite: we tested it no prob, then we
>     introduced persons. A position with a person also returns the "user" 'P
>     12345678' as well as the US-userid. Hmmm... evaluation path doesn't
>     quite
>     do what it says on the tin.
> 
>     Enter a multiline attribute of USR01 objects: a piece of code uses
>     WF_ORGUS, doesn't check for 'US' and generates a couple of empty objects
>     along with real ones. However, it seems that EX in a condition actually
>     checks the object instance of the first element in the multliline - so
>     even though there are a few objects in the multline tab it still fails
>     because of the empty first object. Another 'Hmmm'... should EX fail
>     because of a blank first element in a table?
> 
>     Cheers
>     Mike
> 
>     Mike Pokraka wrote:
>      > Hi all,
>      > Someone please tell me I'm not going insane:
>      >
>      > &MULTILINEOBJ& EX           --> True
>      > &OBJECT.MULTILINEOBJ& EX    --> False
>      >
>      > In both cases multilineobj contains data, the only difference is
>     that it's
>      > a container element in one case and an attribute in another. Quite
>      > straightforward to test in the condition editor using the test
>     feature.
>      >
>      > Anyone seen this?
>      > Cheers
>      > Mike Pokraka
>      > Senior Consultant
>      > Workflow Connections Ltd.
>      > Mobile: +44 (0)7786 910 855
>      > _______________________________________________
>      > SAP-WUG mailing list
>      > SAP-WUG at mit.edu <mailto:SAP-WUG at mit.edu>
>      > http://mailman.mit.edu/mailman/listinfo/sap-wug
>      >
> 
> 
>     Mike Pokraka
>     Senior Consultant
>     Workflow Connections Ltd.
>     Mobile: +44 (0)7786 910 855
>     _______________________________________________
>     SAP-WUG mailing list
>     SAP-WUG at mit.edu <mailto:SAP-WUG at mit.edu>
>     http://mailman.mit.edu/mailman/listinfo/sap-wug
> 
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> SAP-WUG mailing list
> SAP-WUG at mit.edu
> http://mailman.mit.edu/mailman/listinfo/sap-wug