How To Use the Oracle CDD/Repository Interactive Demo
-----------------------------------------------------

1./ Start demo by typing: @SYS$COMMON:[SYSHLP.EXAMPLES.CDD]CDD_DEMO

2./ Answer the question: "NAME OF THIS WORKSTATION?" with the nodename of
    the workstation on which you are running the demo.  

3./ When a list of features is shown, read it aloud and call attention to the
    changing screen colors (CDO=blue, SQL=yellow, VMS=green).  Each time the 
    screen (and simultaneously the screen color) changes, it indicates that 
    you are now switching from a Repository (CDO) environment to a Database 
    (SQL) environment, and vice versa.  This should help the audience to 
    orient themselves as the demo switches between the different activities. 

4./ At logical intervals the demo will ask you, if you wish to continue. If 
    you type "N", the demo will finish and delete all traces of work, reposi-
    tories, databases, files included.  Pressing any other key will tell the
    demo to continue.

5./ At different points during the Demo, you should call attention to the
    following:
                                                                   
    a./  DEFINE AND POPULATE FIRST REPOSITORY 

         We are creating the first repository, which is a simplified copy of
         the "PERSONNEL" database, used often in RDB training.  It takes a 
	 few minutes to create a repository.  The messages should be ignored 
	 at this point.  The demo is defining (= creating) the repository, 
	 and will then, in the empty repository, define fields and records.
	 The records are equivalent with the tables of the database, while 
	 the fields are equivalent to the columns in the table.

    b./  SHOW THAT DIRECTORIES CAN BE VIEWED FROM VMS: 

         At the end of this section, we create a few subdirectories to demon-
         strate how these directories that were created in the repository by
         CDO, can be viewed from VMS.

    c./  CREATE DATABASE (NO AUTOMATIC INTEGRATION FOR THIS ONE): 

         In this section, we'll create the first database, using SQL.  Note
         that we only use this brief SQL command:

                  CREATE DATABASE FILENAME CDD$DATABASE1:B1;

         This database was not created for automatic integration.  All future 
         attempts to integrate data between the repository and this database 
         will have to be done "manually", using the SQL INTEGRATE command.
         We create two tables: EMPLOYEES and SALARY_HISTORY (equivalent to 
         the record of the repository) and several columns (equivalent to the 
         fields of the repository) and since this is an actual database that 
         contains data, instead of metadata, we insert rows of data into one 
         of the tables.  We create two additional tables: JOB_HISTORY and 
         WORK_STATUS.

    e./  SHOW REPOSITORY BEFORE MANUAL INTEGRATION (EXPECT ERRORS):

         We'll return to our repository: PERSONNEL and take a look at the 
	 tables, (records in repository terms).  While we have no difficulty 
	 viewing EMPLOYEES, that we created in this repository, we are unable 
	 to view the two new tables that were created in the database. 
	 Errors are reported.

    f./  MANUALLY INTEGRATE DATABASE CHANGE WITH REPOSITORY:

         Back to SQL, we will integrate the database with the repository with
         this command:

	 INTEGRATE DATABASE FILENAME CDD$DATABASE1:B1 CREATE PATHNAME 
							CDD$REPOSITORY1:B1;

    g./  SHOW RESULT OF MANUAL INTEGRATION:

         Once again we try to view the records in the repository.  We had 
	 difficulties earlier.  We have no problems this time!  After manual
	 integration, the repository shows the records (tables): JOB_HISTORY, 
	 WORK_STATUS that were created in the database.

    h./  CREATE AND POPULATE SECOND REPOSITORY

         We are creating a second repository, whose values will be based on
         the first repository: PERSONNEL.  This repository is called: 
         CREDIT_UNION.  Note the dependency of the second repository as it's 
	 shown by the "SHOW REPOSITORY/FULL" command.  Note the difference 
	 between the "SHOW REPOSITORY/FULL" and the "SHOW ALL" commands.

    i./  CREATE DATABASE (FOR AUTOMATIC INTEGRATION): 

         Now we are creating another database, but at creation time we indicate
         that this database is connected to a certain repository.  Note the
         difference between this database creation and the previous one.

         CREATE DATABASE
         FILENAME CDD$DATABASE2:B2
         PATHNAME CDD$REPOSITORY2:B2
         DICTIONARY IS REQUIRED;

         The fact that this database is linked to the repository will be clear
         as soon we create the table: CREDIT_RECORD.  Without any specification
         in SQL, the table will be the exact copy of the record CREDIT_RECORD 
         in the (second) repository: CREDIT_UNION, we created earlier. 

         We also create a new column: CREDIT_RECORD in the table: 
         ELIGIBILITY_DATE, show the altered table with the "SHOW TABLE 
	 CREDIT_RECORD" command and exit from SQL.

    j./  CHECK INTEGRATION IN SECOND REPOSITORY

         Switched back to the [second] repository: CREDIT_UNION, we now check 
         if automatic integration works.  The new column, id. new field (in 
         repository terms): ELIGIBILITY_DATE should show up in the repository. 

    k./  SHOW RESULTS OF CHANGE:

	 Note that the datatypes in the second repository, CDD$REPOSITORY2, 
         are all longwords.  These elements were all based on a field: 
         SALARY_AMOUNT, that located in the first repository: CDD$REPOSITORY1. 

         Now we set default to CDD$REPOSITORY1 and change the datatype of
         SALARY_AMOUNT from longword to byte.  This should effect the datatype
         of the fields in CDD$REPOSITORY2 since those fields were based on
         SALARY_AMOUNT. 

         Set default to CDD$REPOSITORY2 and note that the datatypes of those 
	 fields changed.  They are now all bytes instead of longwords.  

    l./  DEFINE, RESERVE AND REPLACE COLLECTIONS:

         We are creating a partition and a context within that partition.
         We open the context ("SET CONTEXT...") and create some collections
         within the context.  We reserve the collection and create some fields
         in the collection.  Replace the collections reserved earlier.  Now, 
	 do a reserve and replace by using the "CLOSURE=TO_BOTTOM" option. 
	 Note how each replace creates a new version of the replaced element.

    m./  CREATE DEPENDENCY WITH COBOL PROGRAM (EXPECT ERROR!): 

         The COBOL compiler creates a dependency.  The record: EMPLOYEES can 
         not be deleted. 

    n./  WOULD YOU LIKE TO SEE OUR Oracle CDD/Administrator DEMO?

         This part of the demo is optional.  Any key, but "N" will bring
         up Oracle CDD/Administrator.  It will point to CDD$REPOSITORY1, which
	 is the the PERSONNEL repository.  With the "SET DEFAULT" capability
         however, you can change the default to CDD$REPOSITORY2 (CREDIT_UNION).
         In either repository, you can view the contents, edit, show graphs,
         etc.

         Orientation:

         On your right is the CONTROL PANEL, on your left is the NAVIGATOR.

         Suggested steps:

         Let's take a look at the NAVIGATOR.  At this point it should show the 
         contents of the Repository: PERSONNEL and the role: Standard should be
         in the heading.  Choose an element, like: EMPLOYEES, click on it once,
         then from the View menu, select: Expand.  Note how the fields of the
         EMPLOYEES Record will show up under the record heading.  Double click 
         on the icon (of EMPLOYEES), it should collapse.  Double click again.
         It should expand. 

         On the CONTROL PANEL, pull down the Preferences menu.  Choose: Set 
	 Role.  In the Set Role window, select: Repository Administrator. 
	 Click on OK.  Note that the CONTROL PANEL's Menu Bar is extended to
         include one more option (Maintenance).  In addition, other options 
	 are added to the Menu-list of Tools.
         
         At this point, the Navigator Panel should also show the change and 
         say: "Repository Administrator".  In the Navigator panel, click (1x) 
         on D1, then choose "Expand Recursive" from the View menu.

         Now let's see the same thing on a graph.  Click on EMPLOYEES again 
         and pull down the Tools menu on the CONTROL PANEL.  Select the
         "Dependency Graph".  Enlarge the resulting Dependency Graph Display, 
         move the top element boxes around (to the top).  Most of the fields, 
	 with their version numbers will be visible.  Close this window by
         selecting: "Close" from the File menu.

         Now, let's take a look at: SALARY_AMOUNT.  So many other fields were 
	 based on this one.  In the NAVIGATOR window, select SALARY_AMOUNT and
	 click on it twice.  This will bring up the Editor.  Change the 
	 datatype of this element again.  By using the datatype window, select 
	 "signed quadword", then pull down the File menu (in the same window) 
	 and select: "Save/Close".  The Editor window will disappear.

         To see the content of another repository, from the Preferences Menu 
         of the CONTROL PANEL, select: "SET DEFAULT".  In the resulting 
         Set Default Panel, write in the selection: CDD$REPOSITORY2.  In the 
         CONTROL PANEL, pull down the Tools menu and select: "Repository 
         Contents".  A second NAVIGATOR window should appear, showing the 
         contents of CREDIT_UNION. 

         Click on the second NAVIGATOR window (CDD$REPOSITORY2).  In the 
         CONTROL PANEL pull down the File menu and select: "Create Element".
         In the "Create New Element" window, select: CDD$DATA_ELEMENT and 
         click on "OK".  In the resulting Editor window, fill out the name: 
         "CDD_DATE" (be sure to keep the cursor on the name-field while 
	 typing), then change the datatype to "DATE".  From the File menu, 
	 select: "Save/Close".  If you select: "Redisplay" from the View 
	 menu, you'll see the newly created element showing up on the list.
                  
         To exit CDDA, in the CONTROL PANEL pull down the File menu and 
         select: "Quit".

6./ If everything ran correctly, the demo should clean up after itself. In
    other words: nothing should remain in the directory.