Acceptable Shrinkage

Amidst several news items/rumors this morning that Google was implementing their own URL shrinker I began to think again about some issues that have always bugged me. Like, how exactly does a URL shrinker (like tinyurl) work? And what is the persistence of the usability of that shrunken URL? Most URLs are vaguely human-readable. If I have a link or address that doesn’t work, I can typically figure out if the text got cut off when I copied it or if it somehow got some extra text added to the end or somewhere else. But just the other day I clicked on a shrunken link in a Tweet and it led to a broken links page from the shrink app. The page didn’t tell me what the actual full link was, or how I could fix the break, it just informed me the shrunken link did not exist and maybe I had some extra text in it.

I guess this means the companies that distribute these apps possibly have databases where the original link is associated with the shrunken link, or the shrunken link goes through some sort of decoding process during some routing through their servers. Whatever the case, we are essentially relying on a proprietary encoding format when we use one of these shrinkers. tinyurl claims their encoded URLs are permanent access routes, but should we trust that any more than we trust that file formats, codex and operating systems will persist over time? There is a Firefox add-on that decodes tinyurl-encoded addresses (Greasemonkey), but that is only one encoding system among a growing number.

This may seem like a tiny issue, but shrunken URLs are an ingrained part of the functionality of the shortened forms of communication we are using more and more nowadays (texts, Tweets, instant messaging, short emails). If they do not persist, then the fullness of our communication does not persist. Some people may question if these forms of communication are worth saving, but this is an issue beyond just what belongs accessioned into an archive; this is about maintaining our personal histories. The image of pulling out a stack of old letters tied up in a ribbon or an old photo album is something that will be increasingly rare for one’s own memorabilia. Opening files on a thumb drive or from a storage system will be what we are doing (unless you really want to print out all of those Tweets and texts…).

Just as we must be with all of the image/text/video/audio file formats and codexes out there, we must also be aware of the encoding/decoding/tagging cycle that occurs in all aspects of materials we produce and receive. It seems tedious or uptight or unimportant, especially because we have such easy access to our stuff right now and can’t imagine when we won’t. But then again, it would seem difficult to misplace 22 million e-mails until it actually happens. And that’s no small problem.

— Joshua Ranger