Brief History Of Databases

A society like ours relies, to a large extent, on mechanisms to securely store, store and retrieve information. From the banking operations of millions of companies and individuals, to your seat preferences when traveling by plane, everything is stored in gigantic databases.

In 1956 IBM introduced the RAMAC , which you can see in the image that opens this article, the first commercial magnetic storage hard drive. In a closet just over two meters high and only one ton in weight, it was possible to store about 5 MB, equivalent according to IBM itself to about 64,000 punch cards.

The greatest innovation of the magnetic disk was the random access to the data, something impossible to do with the tape storage units of the time, which we can see in the background of any James Bond movie.

But there is a very long way from the first tape storage systems, in the mid-1960s, to modern NoSQL databases, which can store the flood of data that is on the Internet, unclassified, and then find the needle. you are looking for in that gigantic binary haystack. We are going to see a brief history of how these systems have evolved.

1960. The CODASYL group develops COBOL

With more than 200 large systems in operation and another 175 on the way, the US Department of Defense had spent more than $ 200 million in the late 1950s to make all that electronics work, at increasing expense. To facilitate application development, a working group was constituted between manufacturers, academics and user groups,

which had to define a common language for all architectures, aimed at solving business problems, with the aim of facilitating portability and that a only code could be run on any machine. Among the most important characteristics, the new language had to be easy to understand, with a syntax close to that of spoken English.

One year after the CODASYL consortium was established, the first COBOL (COmmon Business Oriented Language) specification is published , based largely on the FLOW-MATIC language developed by (then) Navy Commander Grace Hopper .

1970. Edgar F. Codd defines relational systems

While COBOL had solved the problem of how to simplify and make the development of business applications cheaper, far removed from the first computer programs more focused on purely mathematical problems, tape storage systems had a serious problem: they did not allow easy access randomized to the data and generated an enormous duplication of information.

To solve these problems, a mathematician employed by IBM, Edgar Frank Codd , devised a system of organization in rows and columns grouped in tables, that made use of the recent advances in mass storage in magnetic disks.

These works were reflected in an article that is already historical, « A Relational Model of Data for Large Shared Data Banks«. Unfortunately, IBM itself spurned Codd's ideas in favor of other products that were already on the market, and for years, they remained a mathematical curiosity.

Also Visit: https://www.norient-beta.com

1974. Chamberlin and Boyce create SQL, the universal data management language

Codd refused to sit in a quiet corner, occasionally discussing with clients the virtues of the system he had invented, creating pressure within IBM to launch a product that applied those ideas.

Tentatively, the blue giant (as the company has been called for decades), launched an internal project to apply the relational model called System R, but commissioned it to a group of programmers who had no direct relationship with the work of Codd. As a result,

Donald Chamberlin and Raymond Boyce created the SEQUEL language., with a high-level approach to the problem (using a syntax close to English) similar to that proposed by Grace Hopper almost 20 years ago. After the name change to SQL, due to a trademark conflict,

it had an immediate acceptance, proving to be an excellent language for data management (not for programming, since it lacks flow control structures) and during the last 50 years it has been one of the most popular languages ​​in the world, always among the 10 most used.

1979. Larry Elison launches Oracle, the first relational system

Drawing on the works of Codd, Larry Ellisonfounded a small company called Software Development Laboratories in 1977 together with Bob Miner and Ed Oates, convinced of the commercial viability of a relational data management system (RDMS) oriented to the medium-sized market,

which made use of modern «minis 'Computers that fit in a small room and didn't need an army of gown engineers to work. In June 1979, they presented the first version of Oracle, called "V2" because Elison thought that no one would want to buy a product without references,

which would be the first RDBMS in the world. It was so successful that in just three years it changed the company name twice to Oracle, as the product had become synonymous with "relational databases," such as tissue Kleenex.

1978. Wayne Ratliff launches dBASE II, the first database for microphones

If Oracle was the gateway for mini-computers to the world of data management, dBase II was for micro-computers, small electronic devices that were mostly sold in kit form, so that the user himself could use them. assemble with a soldering iron and screwdriver set.

Wayne Ratliff , a programmer at the Jet Propulsion Laboratory, wanted to take advantage of the capabilities of his IMSAI 8080 to calculate baseball statistics and improve his sports betting results, so he relied on work by Fred Thomson to create first base. of personal use data, called Vulcan in reference to the character Spok from Star Trek.

Shortly after, he signed an agreement with George Tate and Hal Lashee, distribution experts, and they formedAshton-Tate , whose flagship product for years was dBASE , first for the CP / M operating system and later for Apple II and MS-DOS .

1996. Oracle launches the first object-relational database

For almost two decades, the only way to program applications that made use of relational databases was to include calls to SQL in another language, such as C, or to make small functions stored in proprietary extensions, such as PL / SQL from Oracle or Transact- Microsoft SQL. In the mid-90s, the object-oriented development paradigm,

which defines data structures that are capable of performing all their internal operations, was gaining strength and ended up being applied to RDBMS. Oracle Database version 7.2 incorporates this functionality and soon the company selects Java, a purely object-oriented language, as its preferred development environment.

2004. Facebook develops BigTable, almost unlimited storage

A new generation of Internet services grew with the turn of the century and the arrival of Web 3.0, in which users are the ones who generate the content of Web pages with their contributions. The avalanche of photos, comments and videos of kittens collapses the capabilities of any relational system, due to the volume and speed at which they are generated.

In order to deal with the problem, Facebook launches an internal project, Bigtable , which is not a database in the traditional sense, but a gigantic storage system using data pairs (key or name and value), which can later be used to search with other systems such as HBase or MapReduce.

Bigtable is the basis on which the HDFS distributed storage system would later be developed , one of the basic components of Hadoop, which would open the doors to massive data analysis applications, or BigData , and a new generation of non-databases. relational, called NoSQL (Not Only SQL).

Also Read: where does vanilla flavoring come from