A Document Database - 1960s Style!

I was working with a younger colleague at work the other day. We made an interesting discovery that just goes to show how some older technology does the same thing modern software does - except back then perhaps it was a little slower.

In this case, it was an example of how content databases operated decades ago, versus how they're handled today using software and computers.

Documentation Project

My colleague and I had been selected by a manager to work together on a project involving sorting through loads of paper documents. I was chosen because, as a tech writer, I would know what documents would be useful, versus those that would be obvious junk. My colleague was a pure geek: a database administrator, and perfect for the job at hand.

Stone-Age Artifacts

Our manager had tasked us to work together on a project requiring us to sort through loads of paper documents, convert them into an electronic format, and then render this massive load of content into some kind of organized fashion.

Most of these were typewritten. They had no electronic source files back then. So what we saw was what we got - in physical form. The documents needed to be scanned, and then run through optical character recognition (OCR) software - software which reads the typewritten text first as an image and converts that image into plain-text documents.

Decades ago these paper documents had been hastily stuffed into boxes and squirreled away into some dusty hole in a warehouse, where they were recently rediscovered and dumped upon us for sifting. They were ancient relics, dating back to the 1960s, and written in a style that few people use now.

Luckily there were only a few boxes of these docs, but they had enough content to keep my colleague and me busy for at least a few days.

Way Before Software Was Invented...

My colleague - I'll call him Jim - was anxious to get started. Jim grabbed a small pile of the documents and pored through them. Then he grabbed a bottle of white out and began blanking out information that ran across the top of the first page of each document.

"Hey there, Jim. What are you doing?" I asked.

Jim's answer was that he was blanking out some obviously useless garble on the document cover pages - information he didn't want scanned into text and then have loaded to his database. While his notion seemed to be well-intentioned, I convinced him to stop. Later on, he thanked me.

Rather than grabbing documents haphazardly and beginning to scan them, I suggested we pull out all of the documents and look through them to see whether they were organized in some fashion. At first glimpse, they seemed to have no order. But the fact that they had a more-or-less consistent number at the top of each cover page seemed to suggest some order.

Old Typewritten Documents

In taking all the documents out of their boxes, we found some index cards strewn seemingly haphazardly across the bottom of one box. These cards shed some light on the numbering on the cover pages. After a few moments' further inspection, I turned to Jim and declared, "Hey, these cards are a 1960s style database!"

"How could they be? There's no software?" asked Jim, completely dumbfounded.

I explained that not all databases needed software, especially those that existed before the invention of computers and the software that drove them. As it was with the cards dating back to the days of yore, folks devised more primitive means of keeping record of inventory.

To back up this claim, I explained to Jim how the earliest of business databases came to be.

Tax Audit Databases

In the earliest of times, people used their fingers or small stones to track the things they kept stock of. Later as society grew and became more complex, tax collectors took "audits" from the local merchants they taxed. The meaning of the word "audit" is of course directly related to the word "auditory" - having to do with sound and hearing.

In the olden days an audit took place when a tax collector visited a merchant for a tally of the goods he or she had on-hand. Because many of the common folk were illiterate, there was no written record of the merchant's stock. To make up for this, merchants would call upon readily available "wet ware" technology. They recited the tally of their goods by singing their lists aloud to the tax collector, usually to the tune of a popular jingle.

Because the earliest forms of recorded history came to us as songs passed down over the generations, singing and the auditory hearing of song has been the database memory tool of choice for humans since the beginning of time.

Thus, an audit was perhaps the earliest form of a database there ever was.

Index Card Database

Getting back to our index cards, I explained to Jim that with the advent of written communication, the index cards were just that: an index database in physical form. Hardware, not software.

I pointed out the various hand-written entries on each card signified certain metadata - attributes or metatags attached to the document or piece of information that we would find on a modern software-driven database.

One entry stood for the document number - corresponding to the document's unique identification number (UID) found on many modern databases. Another entry signified the box in which the document could be found. Yet another was the last name of the author who had written the document, or perhaps who had control over its content. The number entries went on and on, and Jim and I were at a loss as to what they all meant.

What was clear: the index cards were a database.

Success!

At the end of our venture, Jim and I were able to successfully capture the data from those documents. It took some time to scan them all, and later even more time to attach metatags to each item as an entry in Jim's software-driven database.

In retrospect, both Jim and I marveled at how much thought an planning had been put into setting the index cards up in the first place. Primitive as the index cards seemed to be at first, they certainly must have done the job back in their time.

Given that there were so many documents, it must have taken quite some time to search through the cards to find the document one may have been searching for. Just think: how sore one's fingers must have gotten if searching through thousands of entries!

Thank goodness for the software DBs we enjoy today!