[Logo] Function Point Analysis And Software Estimation Forum
  [Search] Search   [Recent Topics] Recent Topics   [Hottest Topics] Hottest Topics   [Members]  Member Listing   [Groups] Back to home page 
[Register] Register / 
[Login] Login 
Messages posted by: Steve Neuendorf
Forum Index » Profile for Steve Neuendorf » Messages posted by Steve Neuendorf
Author Message
I think there is a simple and logical answer. There is a real simple answer and it is to the question of why are you using adjusted function points when it fell from grace almost a decade ago (about 20 years too late IMNSHO)? That notwithstanding, as Carol points out you are constructing your UFP from a set of discrete numbers that are not even continuous. Assuming you have counted correctly so that accuracy is not an issue, the best possible precision for a typical count would be no less than + or - 5 FP. I would not be the least uncomfortable, and consider it an added service to my customers, (not to confuse them as the most recent post notes) to round everything up/down to the nearest 10. Don't fall into the "measure it with a micrometer; mark it with a grease pencil (soap); and cut it with an axe (torch) trap.
Hello mpalmier,

Welcome to a great community. I am assuming your class was on IFPUG Function Points. The Challenge is always to ignore the physical view and implementation knowledge of your application and use the logical user view. Hopefully you have access to the IFPUG Counting Practices Manual (CPM) and you can always access the networks of knowledgeable and experienced FP counters and users through this group and several others, including the message board hosted at ifpug.org.

The key concept for answering your question is that of (in)dependence, as explained in the CPM. From your explanation, you have 3 distinct entities: Device, Attributes and Applications. The primary subject of your application is device management, So I would start with Device as being the logical group of data from an user view. A device can have any set of attributes and any number (or no) applications. From an end user view, a device with no attributes would not be meaningful, as nor would any attribute(s) not associated with a device. The way I think, the cardinality is many to many (no zeros) so the two entities are dependent. Device is an ILF and Attributes is a mandatory RET. The cardinality is also true for applications; however, a device with no applications is meaningful, since it can have applications installed and it may be useful (meaningful) with none of the applications of record installed. Also, an application not installed on any device may be meaningful, since it can be installed on any device (think of an application acquired and recorded for an anticipated shipment of devices), so Application is independent and an ILF also.

Let me as an aside recommend a cult classic movie: "Harry In Your Pocket." "Harry doesn't hold" and "Steve doesn't do DETs" (at least not on the boards or outside of "Metrics Views". HTH Steve
 
Forum Index » Profile for Steve Neuendorf » Messages posted by Steve Neuendorf
Go to:   
Powered by Function Point Modeler © FPM Team