Some questions about K2

I hope the following helps:

Q) Can a large process map be broken into smaller process maps and then embedded into the parent process?

A) Yes, by using IPC. IPC Concepts explained here: http://help.k2.com/en/k2blackpointhelp.aspx?page=IPC_Concepts.html. More on the IPC event wizard can be found here: http://help.k2.com/en/k2blackpearluserguide.aspx?page=Reference_-_Event_Wizard_-_IPC_Event_Wizard_-_Orientation.html.

Q) Can a process be reused in other process maps?

A) Again, IPC is your answer. Have a look at the URLs listed above.

Q) Can we extract the K2 code as a Workflow Foundation (can we create the code quicker using K2,  what happens to all the k2 reference

A) You can to an extent (right click on an event and click on View Code), but keep in mind that K2 has its own “workflow engine” sitting behind it, and doesn’t rely solely on the .NET framework and WF code when doing execution. So I won’t recommend you use K2 as a WF code generation tool.

Q) Can we set Preconditions and Events

A) You can set up preceeding and start rules... Not sure I understand the question.

Q) Can variables be used to determine paths of execution?

A) Yes – you can use datafields or SmartObject values to determine the validity of a line rule, or your can add your own custom code behind a line. Either way, this is how you’ll determine the path of execution. You can also look at Outcomes – more info

Q) Can you initialise the variables before the process starts?

A) If you’re speaking about variables contained in the process, such as process data fields... Mmm, no – the process instance doens’t exist before you create it, so the objects doesn’t exist... You can set the values of the objects as you create the containing object, such as the process istance, SmartObject or Activity instance (if you use activity data fields).

Q) When is it correct to use Smart objects or Data fields

A)This question really has a million answers. Rule of thumb is to store as little information inside a K2 process as possible, and then store all your data in an external system (like SQL, SPS, ERP system etc). You can then put a K2 SmartObject service on your external data source and expose that data to K2 using SmartObjects. (http://help.k2.com/helppages/k2blackpearl/Reference_-_K2_SmartObject_Principles.html)

Q) How does it enable users to easily call a stored procedure on a relational database using a Data Access Service? How can the returned values be used to drive and populate process variables?

A) Use the added Dynamic SQL SmartObject service. (http://help.k2.com/helppages/k2blackpearl/Reference-WS_MCon-SO_Services.html)

Q) What interfaces can K2 integrate to?

A) K2 can use anything with a .NET interface.

Q) How does it enable users to easily access COM and .NET EnterprisesAssemblies (within the process map)? How can the returned values be used to drive and populate process variables?

A) Since you can integrate to anything with a .NET interface, you can use the outcomes from a .NET function to populate data into a SmartObject, use it directly in the process definition, or to set a value of a data field. I recommend you explore the developer reference for more examples on this (http://help.k2.com).

Q) What interfaces can external systems call to integrate with the K2 process?

A) Have a look at the K2 developer reference on help.k2.com. There are multiple APIs that you can interface with, but mainly assemblies ending with Client - these are used for working with runtime aspects of K2 entities, such as processes and SmartObjects

Q) Out of the box, what activities are available to integrate with Microsoft products?

A) K2 integrates smoothly with Microsoft Office SharePoint Services (MOSS), InfoPath, SSRS and SQL.

Q) Does it provide ways to address unexpected exceptions, other than manual intervention or process termination (ie Web service failure, Process Stalled, Duration Overrun)?

A) There is exception handling built in on every level of the process – from Process level to event level. You can add your own custom code behind the exception handlers and have it execute what you need it to do.