Persistence of Data
Introduction, page 1
OLE structured storage is one of the most exciting technologies to come along in quite a while. It will dramatically change the way you work with files. It is......
Wait. Isn't this the way most articles begin when describing a cool technology? They focus on the features and capabilities of the tool or technique being described. But is this really the information that most programmers need?
You see, most programmers (myself included) tend to fall in love with technology. We hear about a cool new tool or technique, and it becomes, for a while, the greatest thing that we've ever heard of. When I was studying computer science I remember being introduced to a new language every week or two, and each one instantly became the "best" language in my repertoire - the language that I could surely use to perform any task more efficiently and quickly than any other. And it would remain the "best" language until the next one came along.
This is something that I think is common among many programmers and engineers. We become passionate about technology. We tend to love our work. Our preferences of languages, tools and machines are almost religious in their intensity (in fact, I dare say that there are programmers who are considerably more passionate about their programming tools than their religion).
So before discussing this "cool new tool" called OLE structured storage, I think it is important that we pause for a moment and place it in perspective.
You see, as programmers our real job is not to fall in love with technology. It is to solve problems using software. And as such, it is our professional responsibility to choose the tools that will best help us solve the problem at hand - not the one that we happen to be excited about at the moment, or the one that the industry is busy hyping.
That said, here is the truth about OLE structured storage: It is a powerful technology that is highly suited to solving particular problems, and totally unsuitable for others. So, how does one determine which problems are right for this technology? Let us start with considering the entire class of problem at hand: that of persistence of data.
