Database Management Systems (DBMS) Technical Papers
Database Mangement systems (DBMS) have dramaticaly evolved over these many years. These were first started in the late 1950s and continue to evolve even now in the middle 2020s.
Initially, databases were on punch cards (look it up if you need to). The database was the card tray. Then there were tapes that held data, then disc drives starting with 3 million bytes. Today data exists on solid state drives in Terabyte size disks the size of an thickness of 10 -15 playing cards.
That said, has anything fundamentally changed? Essentially no. There still are the same four overall components: Logical, Physical, Interrogation, and System Control.
The same can be said of DBMS data models that are defined to support the database's data. Originally, data models were "flat files," wherein a record was a single set of values for a linear set of fields.."
Record types that defined the structures of data records were unable to handle and/or simulate data hierarchies or data networks.
As technology advanced, that is, as the hardware necessary to accomplish a unit of data-base work became smaller and smaller, both network and hierarchical data models arose.
Record types grew in their definitional complexity. Relationships were able to be formally defined and embedded ponters existed to traverse interrecord relationships.
Initially, databases were on punch cards (look it up if you need to). The database was the card tray. Then there were tapes that held data, then disc drives starting with 3 million bytes. Today data exists on solid state drives in Terabyte size disks the size of an thickness of 10 -15 playing cards.
That said, has anything fundamentally changed? Essentially no. There still are the same four overall components: Logical, Physical, Interrogation, and System Control.
The same can be said of DBMS data models that are defined to support the database's data. Originally, data models were "flat files," wherein a record was a single set of values for a linear set of fields.."
Record types that defined the structures of data records were unable to handle and/or simulate data hierarchies or data networks.
As technology advanced, that is, as the hardware necessary to accomplish a unit of data-base work became smaller and smaller, both network and hierarchical data models arose.
Record types grew in their definitional complexity. Relationships were able to be formally defined and embedded ponters existed to traverse interrecord relationships.
Starting in the early 1970s, however, there arose a demand for interrecord relationships that were based on "real-data values" rather than DBMS manufactured pointers. Replacing Computer generated pointers with real-data values dramatically increased the durability of relationships because the relationship mechanisms were "commonly" known values that had common everyday meanings.
These improvements caused the creation of data models that consist of record types that are now called tables, and fields that are now called columns. Relationships between tables are defined within the DBMS's data modeling language. Relationships are procesed through the use of matched-value pairs. For example, an invoice's number contained in the invoice's header table that matches the same value within the invoice detail table's column, invoice number.
While value was increased because of these "simplification" improvements, processing speeds decreased. Today, however, those speed decreases have been overcome due to dramaticlaly increased computer hardware speed.
During the timeframe of the 1960s through the 1980s, alsmost all access to DBMS databases was through programmer-based computer languages. New languages were created that made it easier for non-programmers to create and run reports and to create natural-language command streams. The most notable of these natural-languages the Structured Query Language (SQL). It was stared at IBM and was dramaitically expanded and brougt to market through U.S's Database Language Standard's committee.
All that said, there still are the four major DBMS components, Logical, Physical, Interrogation and System Control, and ironically, data models that were once complex to handle hierachies and networks, and then become simple because to hadle the needs of the SQL language are again starting again to become more complex via DBMS generated pointers so that both hierarchies and networks can be efficiently processed.
These improvements caused the creation of data models that consist of record types that are now called tables, and fields that are now called columns. Relationships between tables are defined within the DBMS's data modeling language. Relationships are procesed through the use of matched-value pairs. For example, an invoice's number contained in the invoice's header table that matches the same value within the invoice detail table's column, invoice number.
While value was increased because of these "simplification" improvements, processing speeds decreased. Today, however, those speed decreases have been overcome due to dramaticlaly increased computer hardware speed.
During the timeframe of the 1960s through the 1980s, alsmost all access to DBMS databases was through programmer-based computer languages. New languages were created that made it easier for non-programmers to create and run reports and to create natural-language command streams. The most notable of these natural-languages the Structured Query Language (SQL). It was stared at IBM and was dramaitically expanded and brougt to market through U.S's Database Language Standard's committee.
All that said, there still are the four major DBMS components, Logical, Physical, Interrogation and System Control, and ironically, data models that were once complex to handle hierachies and networks, and then become simple because to hadle the needs of the SQL language are again starting again to become more complex via DBMS generated pointers so that both hierarchies and networks can be efficiently processed.