Bit Literacy - Mark Hurst [33]
[year] -> [month-event]
Thus the top level of the photo collection is divided into years, as far back as the user’s photos go. For example, I began taking digital photos in 2001, so the top level of my photo library is a series of folders named for years: 2001, 2002, 2003, and so on up to the present year. (I use Apple’s iPhoto to create a library for each year and a separate program called iPhoto Library Manager to navigate between the years.19) Of course, I could just as easily move all my photos into Picasa, or any other photo program, as long as it allowed for two-level storage.
It’s important to create the year folders manually so that you have full control over how you file your photos. Some tools such as iPhoto have a feature that automatically organizes photos by year, based on the creation date of the picture file. It’s best to ignore this feature, since it’s not always accurate. For example, if you scan a year-old photo print into the computer, iPhoto assigns it the date it was scanned, not the date the original photo was taken. Similarly, if someone e-mails you a photo, the creation date of the picture file is not necessarily accurate. The lesson here is not to allow computer-generated metadata to organize your photos. Only you, the user, should have say over where and how your photos are organized.
Within each year-library, I store photo sets in chronological order. The name of each photo set starts with its two-digit month number—01 through 12—followed by a keyword or phrase describing the set. In my photo library, there are two kinds of photo sets:
Photos in ordinary, everyday circumstances at home or around New York City, where I live: these I put in folders with the descriptor “nyc” following the month number.
Photos associated with a particular trip to a place outside New York, or a special occasion: these I put in folders with a descriptor of the place (“maine”) or occasion (“kelly’s upstate wedding”) after the month and date range.
Thus a given year’s photo library may start with these folders:
01 nyc
02 nyc
0215-18 chicago
03 nyc
0308-9 kelly’s upstate wedding
0331-0402 maine
04 nyc
05 nyc
...and they go on from there, ending with “12 nyc”. Each of the folders contains one set of consistently good photos that I took during that month, in that place. (And they are all good photos, since I saved only the best shots from the many I took.)
The simplicity of two-level storage has two benefits:
It doesn’t require much from the user. Perhaps once or twice a month, the user creates a new folder, names it with a month number and a descriptor, and drags the filtered photos into it. Once a year the user creates a new year folder. That’s all.
It’s easy to find photos later, because there aren’t many places a given photo could be. Reducing the possible places where bits can reside is just as important as any other aspect of storing them.
It’s very easy to go back to find, say, the vacation photos from last summer, or the photos of the trip to Denmark two years ago. Even if I don’t remember the month of a given event, I can quickly scan the folders from that year—there are usually not many more than twenty—to find the right one. And this is why it’s important to have two levels of storage. A one-level storage system would mean that every photo set I’ve ever taken would sit in the same list. This wouldn’t scale, because after a couple of years the list of folders would stretch far down the page, making it difficult to scan. A two-level system, on the other hand, scales nicely. Each year contains a modest list of folders, and the list of years won’t get very long for twenty years or more.
To be fair, two-level storage won’t work for everyone, under all circumstances. For example, a professional photographer with dozens of photo sets a month might find it too lightweight to be practical. A high-tech user with an obsessive need to input lots of metadata would