For several thousand years mankind transmitted information in the form of writing, and as long as a copy of something remained intact we can still read it today. Obviously over the years many documents were destroyed in some manner or another and much of the written history of older times has been lost forever.

But starting early last century we started recording things in new formats based upon new technologies. We’ve already seen large swaths of our historical record start disappearing. For example, early films recorded on celluloid begin to rot over time until they are unusable. There have been initiatives to copy many older films over to digital format before it’s too late, but many are already lost to history.

Today we have largely moved to digital records. A lot of our written world today is totally online, from news sites to blogs to emails. I have always been interested in looking deeper at historic figures and have read a number of books of letters between famous people, such as the letters between John and Abigail Adams. Since those letters were preserved we can still get a peek today at how John Adams thought about things and how his world functioned.

But are we going to have the same ability in the future? I think sadly that the answer might be no. Already in our short digital history we have changed and upgraded hardware and software continuously. In doing so, documents in older formats become very hard to read. If you don’t think so, then try to read something that was written using an old word processor like Omniwriter on an old computer like the Commodore 64 and stored on a floppy disk. If you can put together all of the hardware and software from that time you can read it, but otherwise you are out of luck. There are still some services available today where, for enough money, somebody will read such things for you. But the Commodore 64 is only 30 or so years behind us and already the number of people in the world who can read these files is dwindling. It’s hard to imagine that in a century or two anybody will be able to put together the right combination of hardware and software to decipher these kinds of older digital records, and that is assuming that the data on floppy disks won’t deteriorate (which it does).

One might think that since we now have everything on the web things are safer from becoming obsolete. But this is not necessarily true. While things may be stored somewhere on servers, there are a number of reasons why digital records of today might eventually disappear. First, things might disappear in the same manner that historic books disappeared—destruction. Servers might break or be decommissioned without backing up what is on them. Or the companies that own such data might fail and fade away (the shelf life of most tech companies doesn’t seem to be really long).

But we also face the same software issues that we did back with the Commodore 64 example. We are constantly upgrading our browsers and software and not everything we do is backwards compatible. Over time we make it harder and harder to read older digital material, even when it’s available on the web.

As an example, consider something called ‘blink’. You will remember this best as a feature on early web sites where certain text would flash to draw your attention. But blink was not just the function of flashing text, it was part of the operating system in early browsers. If you want to be nostalgic, here is an article that talks about this history of browsers and of blink. As is described in the article, as we have updated our browsers we have eliminated the need to read blink and today most browsers won’t read older pages written with that format. This includes a lot of our early web history and includes really large piles of information such as the whole of GeoCities and other popular web sites during the early start of the web.

I am sure today that there are a ton of smart people who could pull up an older browser and get it to work on an older computer and who can still read the early web history. But will anybody be able to puzzle together all of the right pieces in a century to see this older data?

The history of blink is not unique. Our technology and software will continue to grow and morph and along the way we are not always going to decide to make new things backwards compatible with older digital formats. Over time a lot of our digital history is going to be lost or just grow indecipherable. And that assumes that a century from now that a lot of it will even be available on a working server somewhere. It’s somewhat ironic that in a few hundred years we might know more about the Civil War era than about this century.