SQL Developer Data Modeller and Data Dictionary question

I've Community Wiki'd this

Oracle

• Storage can hold multiple databases
• Databases can be opened by multiple instances
• Server can run run multiple instances
• A server can only run a single instace for any given database
• Data is stored in tablespaces
• A tablespace may be made up of multiple data files
• Tables/indexes/Materialized Views can be partitioned across multiple segments
• A segment lives on a tablespace
• A database object (table, index, stored program unit...) is owned by a user
• A table may be be referenced (through a foreign key constraint) by another table, including a table owned by another user
• A table may have at most one primary key constraint
• A table may have zero or more unique constraints
• A primary key or unique constraint is logically enforced by exactly one index (though you could have a multi-column unique constraint which is backed up by multiple unique indexes with the columns in a differing order)
• An index may be unique or non-unique. A unique constraint may be enforced by a unique or non-unique index.
• A table must have at least one column
• Each column has a datatype (and optionally a length or precision/scale)
• A trigger may relate to at most a single table. (Some triggers, such as SERVERERROR, don't relate to a table).

In my experience it varies from one company/situation to the next.

As a dba I loved toad and sql developer because of the ease of use.

However there are times where you're required to script things using the lowest common denominator. For example writing installation or upgrade scripts that may be delivered to a number of customers. Some may have sophisticated tools others may not. But they all have SQL or PL/SQL.

My advice is do whatever you feel is right for you. Keep focused on your objective. Is the end result more important than how you get there? Or is the quality of your code and learning experience more important?