Introduction

OS/390 is composed of two separate operating environments MVS and Unix. Each environment has its own file system, command structure and user interface. The most common user interface to MVS is the system Time Sharing Option (TSO). TSO provides a combination menu and command user interface. The menus provide the user with prompting information that indicates what type of activity can be performed by selection of the menu item. The user enters the menu number or letter onto the command line, presses the enter key and gets presented another menu or a form to fill out. The form may require the user to type information and then press the enter key to have TSO execute a command that performs the indicated action. The form may also have the capability to allow entry of characters or sequences of characters that indicate functions to be performed on the objects (data sets) that they are placed next to. The user goes back and forth through a series of menus and forms to perform data set related tasks such as copying, editing text and JCL files and submitting jobs and reviewing their output.

The following series of snapshots show the sequence of screens needed to copy a member of a library to another library. The sequence starts with the TSO primary options menu.

By entering option 3 on the OPTION line and pressing enter the utilities menu is displayed next as follows:

By entering option 3 on the OPTION line and pressing enter the Move/Copy utilities form is displayed next as follows:

The user then enters the name of the data set that he wants to copy on the Data Set Name line. Types in the letter “c” on the option line and presses enter to get the next input form as follows:

The user then types in the output data set name on the Data Set Name line and presses the enter key to perform the copy operation. He is then returned to the previous panel where he can perform another copy operation. He continues flip-flopping between the two panels to perform this task. A rather clumsy sequence of activities compared to a simple drag and drop operation using a graphical user interface.

The Unix environment can be entered from a TSO session. The user enters the OMVS command from the TSO command prompt or selects Option 6 from the TSO primary options menu. The following panel shows entry from Option 6.

This puts the user into a command prompt user interface. The following panel shows the output of the “ls –l” command (List File and directory names and attributes command). There are no visual prompts in this environment and the output of commands is directed to the same screen as the input commands. Finding previous information displayed by a command can be difficult since there are no commands that allows one to find data previously output from a command. The user must use a function key to scroll up and down the output to find information previously displayed. A very unproductive activity.

This interface is similar to the one provided by the outmoded DOS operating system which was the predecessor to the Windows operating system. This user interface requires the user to know what he wants to do, what commands to input and their respective options and formats. The Unix file and directory structure has to be learned and the commands that operate on these files and directories have to be memorized. There are no visual prompts to aid the user for any task that he may have to perform. The preparation time to perform useful work in this environment is quite lengthy. The chance for making typing mistakes and command formatting mistakes is quit large depending on the memorization capabilities of the user and his typing skills. This is not a user friendly environment.

TSO has been around for over twenty years and provides a user interface that is infinitely better than the command prompt user interface of the Unix environment. However, it is clumsy in regards to the current capabilities of a web or windows type graphical user interface. The user must be knowledgeable about a wide range of capabilities, commands and the TSO menu structure to become proficient in the use of TSO features. The current standard for a user interface is windows and the web browser. They provide graphic objects that easily identify the tasks that can be performed via a simple click, or a drag and drop functionality. All the commands that the interface is capable of executing are easily accessible from the menu bar or a tool bar. The amount of information that the user has to memorize is minimal compared to the MVS TSO or Unix command line user interfaces.

The graphical user interface of ExploreMVS provides a visual guide for DB2, MVS and Unix file systems in a similar style. The names of files and directories do not have to be memorized since they are continuously displayed via a dynamic tree structure. Each files data characteristics (length, owner, DCB, etc.) are always available in a table display that is synchronized with the tree diagram. The ExploreMVS user interface is designed to provide the user a means of focusing on tasks such as copying or editing files rather than on the mechanics of copying or editing a file. The focus is not on how to use TSO to get your tasks performed or what commands one has to issue to get the task performed but rather the task itself. Drag one file to another and the contents of the dragged file replaces the contents of the file where the drag ended. This activity requires no command memorization. No Typing is required. Mistakes are minimized and the time required to perform the task is minimized. The Windows, Macintosh, and Web Browser approach of point and click and drag and drop in a graphical user interface is a proven medium for allowing average users with minimal computing experience to make use of the computer. ExploreMVS brings this type of interface to the OS/390 environment and provides a means for allowing less experienced users to take advantage of OS/390 features and capabilities using a proven user interface. This allows data centers to make productive use of new employees with a minimum amount of training.

College graduates going into the Information Technology area are already conversant with Windows technology and web browsers in particular. Very few graduates are being trained in OS/390 technology. This requires data centers to provide the OS/390 training which is a very expensive and time consuming proposition. It is not unusual for new workers to be unproductive for the first six months to a year and maybe longer from their hire date. ExploreMVS minimizes this training requirement and allows new staff to be productive almost immediately in the OS/390 environment. They should be able to perform most of the basic tasks of file and data manipulation and job submission and review with a few days of training. ExploreMVS allows new workers to perform useful work while they learn OS/390 technology. ExploreMVS puts the focus on what tasks need to be performed rather than how the tasks are to be performed. It is the “how” things work in an OS/390 environment that requires a long time to learn. ExploreMVS provides an immediate return to the data center, via productive work being accomplished, with minimal chance for errors, while this long and tedious training continues.

ExploreMVS provides a single look and feel to all files in the two OS/390 environments. The user does not need to understand the differences between the Unix file system and the MVS data set schemes to perform useful work on these files. He does not require immediate training on the use of TSO or the Unix command prompt and the hundred or so commands that can be entered and the hundreds of options that can be exercised within these commands. The following shows an ExploreMVS display of the files in the Unix directory /u/ibmuser. The left hand side shows the Unix directory structure(HFS stands for the Hierarchical File System). The right hand side of the display shows the files in that directory along with the file attributes.

The next display shows an MVS data set structure that looks and feels like the above Unix file display. The tree diagram on the left is similar to the Unix tree diagram. However the table now displays information that is pertinent to MVS data sets. The user can easily differentiate what type of files he is dealing with and manipulate them in a similar fashion.

By double clicking on “NEWSORC” in the above table display, the user can access the contents of the library “SSGGW.TEST.NEWSORC”. The above table display is replaced by a table containing the names of the files in the library along with the TSO statistics for each member, if TSO statistics are maintained for this library. The user can select a member by clicking on its name, right click to access a popup menu and select the “open” command to get an edit session on the selected member. The following shows the user interface with the member “ASMONLY” ready to be edited. Note that the left hand side of the display still shows where we are in the data set hierarchy.

DB2 Access

The ExploreMVS system also provides a means to explore the mainframe DB2 system just like the exploration of the OS/390 file systems. This feature is implemented by use of a java servlet that is supported by the IBM OS/390 WebSphere product. The servlet uses JDBC connectivity to the DB2 data base to access the tables and information about tables and indexes. The user starts an ExploreDb2 session, which provides the same user interface as for accessing the MVS and Unix file systems. The following two windows shows how the interface displays information about tables and then a display of the contents of a table.

By clicking the table named “EMP” above then right clicking to obtain a pop up menu and selecting the “SELECT” command from the menu, the system displays the contents of the “EMP” table.

The Db2 interface provides for a point and click exploration of DB2 tables and their contents. It provides support for INSERTING new rows into tables, UPDATING rows of a table, and displaying information about tables, views and indexes. It also provides a facility to enter queries to generate user defined output table displays.

Benefits

There are a number of benefits provided by the web based design of ExploreMVS.

· The foremost is the graphical user interface with its “point and click “ and “drag and drop” paradigm. This provides a friendly user interface that requires a relatively small amount of training to gain a fairly high level of proficiency in executing OS/390 tasks.

· The system implements the thin client approach wherein the server is the repository of the code that implements the system. Only a browser is required on the client machine. There is only one copy of the system and thereby it is more easily managed then a distributed client server application.

· Any number of ExploreMVS sessions can be established to concurrently display different facets of the OS/390 environment. ExploreMVS is not a 3270 application and does not require a TSO or telenet session with the host. The same user ID and password can be used in each ExploreMVS window. This feature helps lighten the load on heavily loaded operating systems by not creating a TSO session for each ExploreMVS session. ExploreMVS is only connected to the server when the user requests an OS/390 object that is not already available at the client site.

· ExploreMVS provides the same look and feel and processing capabilities to all files in the OS/390 environment. You do not need a different application for MVS and Unix and for DB2 access. This uniformity minimizes the learning process required to become proficient in executing OS/390 tasks.