I have a messy desk.

When my administrative assistant drops off snail mail, she stares at the stacks that cover every flat surface and wonders where she can put the mail so that I might see it. And don't get my wife started on what it takes to get me to clean my home office! If the words "do we really need this?" sound familiar, you know what I mean.

I just counted 119 icons on my laptop desktop. Maybe the theory that the computer desktop mimics the real-world desktop is true after all. Only seven of these icons are shortcuts to executables. The rest are documents or folders of documents.

If you like the model of "filers and pilers," I am a piler. And I work in data warehousing. Coincidence? I think not.

Many years ago I took a Myers-Briggs survey on cognitive style. One of the dimensions of cognitive style is described by a range from "judging" to "perceiving." Neither is better than the other. Some people like to consume facts as they learn them.

They extract information content, add it to their base and move on to the next fact or decision. These are the "judgers." Judgers are a delight to work with when you are trying to keep a project moving. They are typically thoughtful yet brisk in decision-making. At the opposite end of the spectrum are the "perceivers." Perceivers like to assemble an array of facts and inputs before using the collective information it contains. They always think there is one more thing this fact or input can add and generally don't like rushing to a decision. These are fun people to work with when creating new ventures. As my marriage attests, P's and J's can live together very well. (If you have not taken the Myers-Briggs test, you might want to take a look. You'll learn about yourself. If you take it with your team, you will all see new ways to cooperate and understand each other. I'm sure there are other models that work very well also.)

With absolutely no science behind me at all, I posit type-J people are more likely to be online transactional processing (OLTP) people and type-P people are more likely to be data warehousing people. I also suspect type-J people keep relatively clean desks and desktops. Type-P people will be the opposite, of course. I know that my boss of many years in SQL Server has the neatest desk I've every seen. He reviews and files or tosses every piece of paper as it arrives in his office. His inbox never runs more than a few dozen items. Mine has almost 4,900 pieces of mail that I will get to one of these days. The OLTP application consumes transactions and dispatches them as soon as possible. It parses the order, sends a new transaction to manufacturing, one to finance, etc. The data warehouse is quite the opposite. If someone bought string beans in Des Moines five years ago on a specific Tuesday, we know about it in the data warehouse. Data warehouse types are the pack rats of the corporation. When I first started in decision support systems (DSS), we were straining the hardware platforms of the day just by keeping aggregated data. We might have held sum of sales by product, market and time period. With today's hardware and networking platforms, we can keep all of the transactions. With ever-growing hardware, bandwidth and technologies like RFID, we will soon store all the pretransactions. It is a type-P playground!

We keep realizing dramatic growth in storage capacity and speed, CPU power, memory capacity and network bandwidth.

At the same time, management costs are dropping, albeit slower than capacity is growing. Will we have data warehouses in five or 10 years? Twenty? After all, data warehousing is really just an optimization. In theory, with enough computing oomph behind you, you could keep every transaction (and pretransaction) and simply comb that store to ask any question you like. We would still want a dimensional model to aid user understanding, but why would we want to materialize it if we didn't have to? We're not there yet in terms of computing speed and capacity. We are growing as a set of interrelated and Internet-fueled economies.

At some time in the future, the beams will cross. I don't know if that's anytime soon. But when they do, the type-Ps and the type-Js are going to need to come to an understanding.

A while back I had a notion to scan every piece of paper that came into my office. Once scanned, I would tag each file with enough metadata to allow searching in the future. I would not need any other storage hierarchy. Forget file and folder names. I could just dump all my files in one folder and find anything quickly. Then I thought about how long it takes to scan a page, much less a 20-page document. Ugh. However, very soon we will have cost-effective adhesive RFID tags. What if you could type a few keywords into an applet and have it print an RFID tag? You stick the tag on the document and toss it on the ever-growing pile. Now you can truly say, "Yeah, it looks messy, but I know where everything is!"

I'm not really advocating this solution. I know that after a few days, I'd just forget about it and go back to piling up paper, untagged. But I do enjoy a laugh thinking about the angst a perfectly tagged, perfectly searchable 10-foot high stack of documents would cause the type-J people in my life!

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access