The quickest way to introduce yourself to NH's Function Point Analyzer is to use this program. This quick start shows the general functioning of the program. Some knowledge about the function point method is assumed.
At first it is usefull to define some project classes in the menu "Options -> Project classes", for example host, client/server etc. New counting projects will be assigned to project classes, to allow better effort estimations.
At next create a new counting project in the menu "Project -> New". You should count a development project or a application, because release projects are based on this counting types.
Choose one of the previously defined project classes and if you are counting a application, type in the effort to create this application. Press "OK" and use the menu "Database" and "Function", toolbar, context menu or short cuts to type in databases and functions.
Create some simple databases and functions with default values. The different options will be discussed later.
If all databases and functions are created, save your project.
Now create another project, but now create a release project. In the listbox "Based on" you can choose the project, you created first. When "OK" is pressed, all databases and functions of the "Based on" project are shown. Function points are set to zero. The state of the base project is set to closed, to ensure that no unintentional modifications will be made later.
You can create new databases and functions or modify or delete existing databases and functions. Depending on the action, function points will be calculated. If done, save your release project.
To calculate all databases, functions and function points for the modified application, when the release project is done, create a new base in the project menu. This base is the new starting point for following releases.
After this quick start a few special functions will be discussed.
NH's Function Point Analyzer can automatically save the actual data of a project in work. The last saved data can be restored by "Project -> Open Autosave", for example when the system has crashed. Autosave is activated or deactivated at "Options -> Autosave". The intervall for saves can be defined here although.
New projects are always created with version 1. "Projects -> Copy project" creats a new project with the same data, but the version number is increased by 1. This way it is possible to create and maintain different variants of a counting.
Data can be stored in different folders for different units or companies. The folder in use can be defined with "Options -> Define data directory". NH's Function Point Analyzer uses only data stored in this defined folder. The definitions for function point adjustment factors, lines of code per function point, COCOMO parameters, variable informations and project phases will be created and administrated in each folder. The program will work with the data in this folder.
Here the adjustment factors for the opened project are defined. Tooltipps with short descriptions for the adjustment factors help during data entry.
When a new release project or a new base is created, the "previous" (based on) project will be closed. Data of a closed project can be viewed only, not modified. If you have to modify something later on, you can set the status to open again in meu "Project -> Set status".
In menu "Interfaces -> Import" XML data of a project stored in any folder can be imported.
If a database is marked as release relevant in a release count, function points are calculated for this database. If a database is marked as not release relevant, no function points are calculated for this database. This functionality helps, if in a release count it is noticed, that in a previous counting of a development project or application, a database was not counted, a database too much was counted, or if a group of databases must be splitted. Without counting incorrect function points, this can be corrected.
If a database is marked as conversion database, function points will be calculated, but the database will not occur when a new base is created.
If a function is marked as release relevant in a release count, function points are calculated for this function. If a function is marked as not release relevant, no function points are calculated for this function. This functionality helps, if in a release count it is noticed, that in a previous counting of a development project or application, a function was not counted, a function too much was counted, or if a group of functions must be splitted. Without counting incorrect function points, this can be corrected.
If a function is marked as conversion function, function points will be calculated, but the function will not occur when a new base is created.
Databases can be assigned to functions, which uses these databases. Not assigned databases are listed in the "Not used databases" report. Those databases are not used by the application.
These reports are available:
The different print reports can be combined to one print.
The average person days per function point for the corresponding project class is calculated. Based on this, a estimation is made, how many person days the opened project needs. Only projects with effort will be used.
COCOMO calculates how many person days are needed in a project. With COCOMO Basic based on function points,
adjustment factors and programming language the amount of lines of code (LOC) is estimated. Depending on the complexity
of the project (organic, semi-detached or embedded) COCOMO Basic calculates an effort estimation.
With COCOMO Post Architecture unadjusted function points are weighted with COCOMO parameters. Based on these adjusted
function points the programming language the lines of code and project costs are estimated.
It is assumed that one month has 20 person days.
Actual project costs will be shown. Lines of code must be defined before in Estimation -> COCOMO Post Architecture and in Options -> Project Phases costs must be defined.
Data for each project are stored in one XML file for each project. Additionally some important values for each project (project name, function points, effort etc.) are stored in a file called projects.txt. For each folder defined in "Options -> Define data directory" a projects.txt is stored in this folder.
Menu "Open project" shows all projects from projects.txt. You can work on these projects and these projects are used in estimations and statistics. When a new folder is defined in "Options -> Define data directory", the projects.txt of this folder is used. If there is no projects.txt in a folder, a new projects.txt is created. Function "Reoranize database" checks if an XML files exists for each entry in projects.txt. If no such XML file exists, the entry in projects.txt is deleted. This function is used to clear inconsistencys, for example after system crashes.
If you want to store more informations than are stored by default (project manager, names of function point counters etc.), you can define your own informations that will be stored in the project data file.