Process Flow Version Number

Hi Jeff

FYI - For this type of question, you'd probably be better off in the ProcessFlow forum rather than the S3 Customization/Development forum.

In any case, are you using "classic" processflow (PFI/PFA), or the new Infor Process Automation (IPA)? 

I don't believe there is a delivered variable for this in either product, but you can query to find it.  If you are using pflow classic, then you should be able to query {logan prodline}.WFPROC.VERSION_NBR by passing in the process name (there is a variable for this.  If you are using IPA, you can query for PFIFlowDefinition.CurrentFlowVersion the same way.

If you are using IPA, you can also query the current workunit to determine the flow version for that workunit.  I don't think this is available in "classic".

I hope this helps.

I am sure that there are differences for the versions (PFI and IPA) but I bet this query would work in either version:

PROD=LOGAN&FILE=WFWORKUNIT&INDEX=WFWSET1&KEY=XX&FIELD=PROCESS-NAME;WORKUNIT&OUT=CSV&DELIM=~

XX = oWorkUnit

Shane is correct that his query should work in PFI/PFA. 

However, it won't work in IPA, which is the new Landmark version of pflow.  Landmark doesn't have a "logan" and each app prodline has it's own set of pflow tables.  You'd need to use a Landmark node with this query: 

_dataArea="{prodline}" & _module="pfi" & _objectName="PfiFlowDefinition" & _actionName="Find" & _actionOperator="NONE" & _actionType="SingleRecordQuery" & _runAsUser="" & _pageSize=30 & PfiFlowDefinition="{ProcessName}" & CurrentFlowVersion

We are converting our flows to V10 next week.... Looks like i might need to rewrite some of my queries. Woozy - do you know if the conversion program will create these in landmark for Me? I am guessing I will need to write them from scratch but hoping there will be some conversion help. Is the landmark node pretty easy to build queries in?

Posted By Woozy on 05/23/2013 02:28 PM
FYI - For this type of question, you'd probably be better off in the ProcessFlow forum rather than the S3 Customization/Development forum.

Moving to ProcessFlow forum

Posted By Shane Jones on 05/23/2013 10:30 PM
We are converting our flows to V10 next week.... Looks like i might need to rewrite some of my queries. Woozy - do you know if the conversion program will create these in landmark for Me? I am guessing I will need to write them from scratch but hoping there will be some conversion help. Is the landmark node pretty easy to build queries in?

Hi Shane,

The conversion program is slick, but it isn't very smart.  All it really does is change the file format - the logic is unchanged.  It changes the PFI nodes to the matching node types in the IPA designer and converts the flow to the IPA xml format.  If you are using an S3 Query node in PFI then it will remain an S3 Query node (with the same call) in IPA.  If you are moving to IPA then you will have to replace any queries that reference the logan pflow tables with new Landmark nodes querying the Landmark IPA tables.

The Landmark query builder is pretty simple to use, but it isn't anywhere near as robust as the S3 query builder.  For example, there isn't a wizard for adding data selection or filtering, so this has to be built by hand - and it is extremely finicky and filtered performance tends to be terrible.  Also, you can't just add fields from related tables but you have to do a "relationship query" instead.

If you run into issues or questions during your conversion, please post the question and I'd be happy to help if I can.  We haven't done much with the conversion program since we aren't live on TM10 yet, but we use IPA for everything other than the few flows (less than 5) that we continue to run on PFI.

Good Luck!

The migration utility gets you about 95% of the way, but I have had to modify some nodes to get them to work correctly. It's mostly little stuff, but it's time-consuming since you have to really test out each flow to make sure it still works as desired. Examples:
- display exec part of the user action node isn't migrated correctly (I have reported this and it should get fixed).
- flows that call WF servlets (e.g. to add/update a workunit variable) are moved from Webrun node to Landmark transaction nodes, and didn't work without changes.

Regardless, you might want to take this as an opportunity to revise your flows anyway to take advantage of new functionality, for example:
- user action node can now send an email notification when work is put in the inbasket; previously required putting an email node in front of the user action
- the email notification associated with the user action can also include the actions; so a user can read email notification, click desired action, all from email -- never even logging into Lawson and going to inbasket. (yes, there are still some issues related to network connectivity and security; I can share details if you want).
- error handling in the flows is now handled much better

Just with that you can reduce number of nodes in a flow by about 20%.

I would recommend if you are interested, that we start up a new topic.

One clarification about the Landmark query builder.  Data selection by key fields is simple.  It is filtering on non-key fields, or querying multiple values for a field, that is more difficult.  I find that I use SQL queries when the logic gets complex as I have much better control and performance.

Posted By Woozy on 05/23/2013 05:05 PM
Shane is correct that his query should work in PFI/PFA. 

However, it won't work in IPA, which is the new Landmark version of pflow.  Landmark doesn't have a "logan" and each app prodline has it's own set of pflow tables.  You'd need to use a Landmark node with this query: 

_dataArea="{prodline}" & _module="pfi" & _objectName="PfiFlowDefinition" & _actionName="Find" & _actionOperator="NONE" & _actionType="SingleRecordQuery" & _runAsUser="" & _pageSize=30 & PfiFlowDefinition="{ProcessName}" & CurrentFlowVersion

For LPA/IPA... the above returns the current flow version. For actual flow version used for a given workunit, get it from the workunit itself (since it might be using an older version of the flow than the current version): > _dataArea="{LPSDataArea}" & _module="pfi" & _objectName="PfiWorkunit" & _actionName="Find" & _actionOperator="NONE" & _actionType="SingleRecordQuery" & _runAsUser="" & _pageSize=1 & PfiWorkunit="{WorkUnit}" & FlowVersion