Wednesday, November 16

The best date format. And more.


Numerical date formats vary around the globe. Let’s just consider the Gregorian calendrical system, which is what most of us use every day.

The American version is the worst. Month (med), then Day (small), then Year (large). Huh? Not good for sorting, and makes little sense to the human mind unless you're inculcated with it from birth, kinda like Fahrenheit temperatures. It's arbitrary.

Example: 9/11/2001 (September 11th, 2001)

The version most of the rest of the world uses (not counting religious or other ‘traditional’ calendars), is better. Day (small), then Month (medium), then Year (large). It’s in ascending order at least...

Example: 25.12.1972 (Christmas Day, 1972)

However, it still fails when it comes to standard computer sorts, because we need descending order for that. They got it right, except that it's backwards! Well, unless you write in a right-to-left language, I suppose. Tower of Babel indeed.

It turns out the perfect date format would be [year].[month].[day], which is in descending order of magnitude, and thus lends itself perfectly to computer sorting.

Example: 2011.11.16 (today – year: 2011, month: November, day: 16th)

The delimiter (slash, dot, etc.) could be anything, really. Neither of those are probably ideal because forward slashes are used in Web URLs, and dots are used as periods (full stops). Maybe the ‘>’ sign would make sense because the big end is always opening toward the larger concept. Year > Month > Day. Simple. It could even be extended to delve right down into hours, minutes, seconds… the problem with that symbol is that half the people get it confused with its sibling, ‘<’ – ah well, we could find something I’m sure, if we put our collective minds to it. Or rather if our mechanical successors do.

However, two forces remains an obstacle. Inertia (people don’t like to change) and lack of widespread reasoning skills (people seldom do what makes the most sense).

The fact that the proposed system does makes the most sense is the reason it’s probably never caught on anywhere. Still, I’ll bet its moment will come, whether it takes 20 years or 50 years, to achieve some kind of worldwide ISO status, simply because it’s the best ‘sortable’ format and the world’s information, including dates, is being converted to computer data at a rate that has been growing astronomically for some decades now, and that train isn’t about to slow down.

---

I’ve already changed the way I label music, for example, so that it sorts correctly for me: [Artist] - [(Year)] [Album]. Inside of that folder would be the song file itself: [TrackNumber]. [Title]. Note that the “.” after the track number is part of the file’s name construction and that “TrackNumber” is a 2-digit number (XX) so that even that sorts properly. Thus, if I want to listen to the Dan Fogelberg song “Illinois” I would open my music folder and find it in the “D” section (I don’t like using LastName, FirstName mixed with ordinary band names; it’s too confusing and involves commas)…

(list continues above)
Dan Fogelberg - (1972) Home Free
Dan Fogelberg - (1974) Souvenirs
Dan Fogelberg - (1975) Captured Angel
(list continues below)

Aha, there’s the correct album in the middle!

Now before we go on, you’ll notice that the big downside of this is that you have to know which album the song is on, or use Ctrl+F/Cmd+F to find it, which defeats the purpose a bit. This system is for neatness and sorbability, not for finding things most easily. That’s why multi-sort databases like iTunes, using an information tag system such as ID3, will always be superior to a hierarchical folder-and-file system. But I still name and store them this way: It gives me the best of both worlds. I can use iTunes, sure, but if I want to share a few albums from a particular artist with a friend, for example (so they can evaluate them and then buy them themselves, of course), I can do so very easily and the friend will be delighted to see how organized it is even outside of a database.

So, to get back to our example, I open up the Souvenirs album folder with a double-click and see that the song I want is the second track…

01. Part Of The Plan
02. Illinois
03. Changing Horses
(continues)

Some people don’t like the leading zeros in single-digit track numbers, but they are essential if you want to assure accurate sorting on all systems in all situations. I’ve grown used to them and quite like them now! It keeps the names lined up rather well.

That’s just one more example (in addition to the date formatting example above) as to how we’re going to have to pay attention to sort-orders in this increasingly digital world. When an average person may have hundreds or even thousands of files ranging from text and presentations to music and video, it’s vital to have a system to organize it all somehow, just so you can find things. And cleaning your naming system up might even make you enjoy browsing your files again. Sure it’s practical, but I think you’ll find it’s aesthetically nice as well once you’ve done all the hard work of renaming everything.

For example, even my user-content folders (as opposed to System stuff, which I leave alone) are organized similarly. Something like this…

00 Meta-information
01 Downloads
02 Text Documents
03 Music
04 Movies
05 Documentaries
(etc.)

Again, those would be folders inside a main “Justin” or “Documents” folder - the ultimate outside folder tends to be determined by your operative system. I prefer to put them on the root of my boot drive (C: drive for Windows people), but some people may not like doing this, and each new version of all major OS’s seem to keep further encouraging us to use a “(My) Documents” folder, which is usually buried in a user account folder, which is buried… Well, maybe you can see why I like to put everything at the root level so that I can get to it easily even from a command-line interface. When your multimedia/content folder is buried five layers deep, it may not be a nuisance when using the GUI, but it will be if from a command line when you have to try to remember where that sucker is!

I wish OS’s would stop trying to force us to organize things their way, or would at least give us a choice – say, to simplify the directory (folder) structure for single-user systems. If I’m the only one who ever uses my computer, I shouldn’t need a “Justin” account. I would rather see the root of the boot (C: or Main or whatever) hard drive look something like this:

00. Meta-information
01. System Files
02. Drivers
03. Applications
04. Programming
05. Downloads
06. Music
07. Photos
(etc.)

Notice how I changed the formatting slightly in this example from the previous one. This is just to show that there are still stylistic choices to be made. The important thing is to be consistent.

This seems to me not only to sort correctly (or at least according to taste), but to be much, much cleaner than the current paradigms in use. It makes intuitive sense to me and I could navigate it from a command line if I needed to with little effort. Of course, the “System Files” folder may still be a big spaghetti-like hierarchical monster inside, but as users we needn’t really venture in there anyway; that’s where the OS lives and does its thing. The main point is that getting the user’s multimedia data - loosely defined as ‘the stuff you want to back up’ - out of some mess like that and putting it on par with everything else would make things easy to find them where they really reside without having this dated, tangled desktop metaphor in which both files and shortcuts (aka aliases) can be present, but in which the whole “desktop” itself lives in a folder insider the user’s account, yet can reference items outside that account… even the desktop itself. Please don’t think about that last statement too much, or you might grasp how Gödelian it all is and mentally explode or something. And I don’t want that on my conscience, not to mention my monitor.

j.sias 2011.11.16

Archived Posts

Search The Meta-Plane