Technology Standards - Ya Gotta Love 'Em!
JOLAP, XML/A, JDBC, OLE- DB, ODBC, MDAPI, DCOM, CORBA, J2EE, and even dear old MOM is in there somewhere. It sounds like alphabet soup or the refrain of "Old McDonald Had a Farm" (E-I-E-I-O...) in terms of the jumble of acronyms that represent the plethora of standards our industry creates. Standards are a good thing. They make life easier for vendors developing products and for those deploying these products in their environments. They have had a dramatic impact on business intelligence (BI) and its availability and affordability for companies desiring significant decision-making advantages.
|
"Standard something, such as a practice or product that is widely recognized or employed, especially because of its excellence." The American Heritage Dictionary |
In the Beginning, Chaos!
In the beginning days of data warehousing - before we named it business intelligence - the environments were created using a patchwork of tools, software, hardware and human effort. Hand-coded ETL and hand-crafted designs reigned supreme. There were no BI standards because it was a very new and unknown concept and set of processes.
The early pioneers were forging into unknown turf and had to create solutions where none had ever existed. Even online analytical processing (OLAP), the veteran "standard" today for analytical processing, was unheard of. Each environment was unique and contained many of the nascent and innovative technologies that we take for granted today. The environments were also very expensive to build, limited in their functionalities, had few software or hardware products specific to their needs, and ultimately were used by only a few of the intrepid and adventurous souls within the enterprise. Only the largest corporations could afford to create these environments. The cost and difficulty put these functionalities completely out of the reach of their smaller compatriots.
Common Denominators
Over the past decade and a half, we in this industry came to the slow realization that certain processes and technological attributes were common in the environments we implemented. The innovative characteristics that made these early environments work were used over and over for other companies. The BI vendors were also able to reuse what they had learned by incorporating that learning into their products. Their "standards" allowed them to make their products cheaper to build and easier to implement.
As BI became a multibillion dollar industry, it caught the attention of the significant technology players of today such as Microsoft and HP. Acquisitions and mergers began to consolidate the smaller vendors and their products into these big-name vendors. In addition, these large companies began to fund significant internal development projects specifically to support BI. BI moved into the big time; and with that move came a push to create standards so that everyone, not just the largest corporations, could benefit from these analytical environments. BI was coming of age!
Of course, every vendor wanted its internally developed standard to become the standard for the whole industry and therein lies the problem. It is of considerable value to a vendor to have its standard or set of standards become the industry standards for obvious reasons. The identification, documentation and acceptance of technology standards then become a difficult political battleground. Councils, organizations and standards bodies of all sorts are created with each promoting the goodness of their standard over their competitors'.
The Winning Standard(s)
As stated in the definition of a standard at the beginning of this article, a standard is something "widely recognized or employed." There is no law that says it is mandatory for a vendor to comply; however, the industry as a whole recognizes that compliance is a good thing, and you'd better have a good story if you want to sell a non-complying product in the market. That is a testament to how beneficial we have recognized a common set of attributes to be.
Today, the good news is that there are a number of standards specific to the BI environments. In some cases, we have competing BI standards (JOLAP and XML/A are good examples) and a definitive "winner" has not yet emerged. Who decides? Ultimately, you do. The marketplace will decide as it always does - each dollar spent is a vote for which standard will win and which one will fade into history. The Betamax versus VHS battle in the U.S. is a clear example of how the marketplace determines a standard.
Thus, if standards are necessary and vendors are strongly encouraged to comply with them, what are the advantages to the vendor and to you, the customer, of these standards that drive this decision? On the other hand, what is lost when the industry adopts a set of standards?
Advantages to the Vendor Community
For a vendor to adopt an industry standard, the standard must contribute considerable benefit to the company. Let's look at some of these to understand the importance of today's standards:
- For many vendors, a simplistic view of their competitive driver is one in which they strive to create a product for less money than their competitor and then sell that product to the customer for less money than their competitor can. Therefore, the first advantage to adopting a standard must deal with the reduction in development costs. HP's long history of supporting industry standards and their adoption of the standard interface to all hardware devices in their ProLiant and Integrity servers is one example. This interface permits the vendor to create one set of code that can run on virtually any hardware platform at a fraction of the cost of traditional solutions. Traditional (non- standard) solutions had to be developed and supported on multiple platforms, greatly increasing their costs.
- The adoption of database standards means that software applications are portable. The independence created by using a database standard translates into decreased development time and reduced support costs too.
- The hardware and database standards have had a sizable impact upon the extract, transform, load (ETL); enterprise application integration (EAI); and enterprise information integration (EII) vendors as well. Having database and hardware independence directly translates into higher performance for these tools as they perform the heavy lifting of capturing and integrating your data and applications. In addition, their maintenance costs are reduced as is the time to market for enhancements.
- The adoption of standards also promotes the wider availability of software products by making heretofore proprietary hardware and software open and available to all software developers. BI, in particular, benefits from this because of the great breadth of technical diversity needed in a sophisticated BI environment. This diversity includes technology that performs simple querying and reporting to OLAP to data mining to exploration or ad hoc analysis of the data.
