Unlike "run", certain procesures like proc sql, datasets use "quit" statement to end their step. Every document specifies that they are "interactive" procedures which ic different from other proc's. So, my question is "in what way are they different?". Can we visualize this by an example code?
Any of these PROCs have commands that can be executed depending on the specific requirement, such as PROC DATASETS or PROC CATALOG - you can review the SAS Procedures Guide for the commands related to the procedure. And so the given procedure remains in a "running" state (depending on the OS platform and whether executing in batch or interactive) until the QUIT; command is issued. Also, for some PROCs, an allocated "libref" or "fileref" remains open until the QUIT; command is issued, which is another consideration for the command.
Further to Scott's reply, another example of a command is a SQL statement in proc sql.
You can start proc sql, and then run SQL statements as you go. Each SQL statement is standalone within the procedure, and doesn't need anything else (like a run to be executed.
A third example is proc gplot: You start it, and then do as many plots as you you want on the table the procedure uses. In this case, because a graph might need several statements, each graph is created when the run; statement is encountered.