An ADAM application is a program which uses ADAM subroutines. From the user’s point of view, the most important of these is the Parameter system since this controls (in conjunction with the user language) the interaction between the program and the user. The Parameter system uses a file called an Interface File which stores information about a program’s parameters.
Many users will only use applications which have been written by someone else. In this case, everything necessary to enable them to be used should already have been carried out. This includes the preparation of the interface files and the definition of commands. All you need do is carry out a specified startup procedure. However, if you have written your own application, you will need to do some preparation of your own before using it.
The process of running an ADAM application can be summarized as follows
I know it’s hard, but it often really does help to read about the program you intend to use. Program documentation often tells you what the program does, what data it uses, and what its parameters are and how to specify them. Why not try it?
Don’t forget to enter ADAMSTART — people do.
The options are DCL, ICL, and SMS at the moment. ICL is the one we recommend. SMS is usually used only for data acquisition and is not covered in this guide. Some packages won’t run under some languages — Chapter 4 tells you which ones.
Some packages need to have various definitions and logical names defined before you can use them. For example, before you can use ASTERIX you must enter the DCL symbol ASTSTART. Once again, Chapter 4 tells you what to do for specific applications.
Most applications contain many programs. You must select the programs you want to use. For applications provided by other people, this will normally involve just typing the name of the program. However, if you have written your own application, you will need to define the appropriate commands required to run your programs. Applications which have large numbers of programs can be executed more efficiently if they are grouped together into a Monolith. These are considered below.
Once you start running applications you will need to control them. This is done by specifying parameter values — names of data files, values, options, and so on. This is the most complex part of the process and is discussed in Chapter 8.
You normally select programs just by typing their names. However, for this to work, the program name must have been defined as a DCL symbol or an ICL command. Also, there might be a delay the first time you run a program in an application. This is usually because a large Monolith is being loaded.
To run programs under VMS, their names are defined as DCL symbols so that typing the name runs the program. The symbol definitions are executed by running a command procedure.
Similarly, to run programs under ICL, their names are defined as ICL commands using the DEFINE command described in Chapter 6. Usually, the command definitions are executed by running an ICL procedure which contains them. Before a program can be run under ICL, a command to run it must be defined.
Programs are sometimes stored together in structures called monoliths which are loaded for execution as a single entity. In this case a command must be defined for each program in the monolith. When you use an application package, the commands needed to execute its programs are normally defined for you when you start up the package. Thus:
will cause all the commands required to run the KAPPA programs to be defined. What happens is that a file called KAPPA_DIR:KAPPA.PRC1 containing the required DEFINE commands is automatically input to ICL for execution. The commands look like this:
There are also commands to display the initial messages output by the package. KAPPA.PRC is an example of an ICL Command File which is described in more detail in Chapter 9. Notice that all the commands refer to the same executable image (the monolith), and that the directory in which this is to be found is specified explicitly so ADAM_EXE is not used.
Monoliths are useful because they enable lots of programs to be loaded within a single process creation. Since process creation is quite slow under VMS, you save time by doing it only once. The advantage of monoliths is that they reduce the time required to create processes. The disadvantage is the time required for the loading a large monolith instead of a small program; however, you only have to do it once.
1The file type ‘.PRC’ is now obsolete. The normal file type for command files is ‘.ICL’.