Skip Navigation

Sort of

Track sorting in iTunes was one of the first things I noticed about Apple’s software engineering: they actually thought about it. Sorting by song name (obviously) sorts the songs alphabetically. Sorting by artist name (obviously) sorts the artists alphabetically.

But it’s the secondary and tertiary sorts that are important.

Sort by artist name and you aren’t sorting it by alphabetical artist and by alphabetical song… you’re sorting it by artist, then by album, then by track number. Albums are indeed supposed to be in order —track order, nor alphabetical order— and it’s that kind of thinking that makes me happy to own a Mac.

But they missed something.

Sorting by album only, the secondary and tertiary sorts should be by artist then by track number. Having the tracks of Crowded House’s Afterglow intertwined with Sarah McLachlan’s Afterglow has taught me this cruel lesson… and it’s no coin toss: it needs fixing.


Tamara Crowe writes to point out that sorting by album and then by artist would break compilation discs, split discs, and several other cases. There are plenty of ways around this, like checking the “is a compilation” bit, checking for different dates, different album lengths, et cetera.

I don’t much care how it’s fixed. I just care that it is.

Update deux

iTunes 7 fixes this problem in a roundabout way. It requires human intervention, sadly, but it works.

Amazon and the Monster

I’ve played a little more with Delicious Library over the last day or so, read some reviews on other blogs (Michael, Erik, Sven) and on Ars, and came away with the conclusion that maybe I just wasn’t hard enough on Delicious Monster… maybe their software really did suck and I didn’t notice because (I’ll say it again) it’s just not my kind of app. No matter how well it’s done, it’s hard to be picky about something you view as fairly trivial; faults in a FreeCell clone are more likely to draw criticism from me than faults in a library manager (and have) simply because I require a decent game of FreeCell from time to time and don’t want to have to fire up the PC.

Delicious Library’s icon

But again, after tooling around scanning in an assortment of CDs, books, and games I still can’t see that much wrong with it — a lack of secondary sort criteria won’t kill me (even if it does kill me in iTunes), the diminutive size of peoples’ mug shots in the Borrowers list doesn’t really yank my chain. But I’m not here to defend Delicious Monster —everybody’s complaints are warranted in a bizarre and thankless way— I’m here to discuss one of the biggest barbs aimed at Library: the Amazon butt–boy claim.

The application ties into amazon too heavily for my taste. And that’s the friendly way of making my point. Frankly, some menus look like amazon should give us the app for free.

Sven’s comment above isn’t the only example, and I am by no means picking on the man, but the general gist of the complaint is that Delicious Library is too integrated with Amazon… and that if we’re going to get this many ads and links to Amazon products, maybe we shouldn’t be paying so much to organize our CD collection.

It’s true, Delicious Library relies almost entirely (and when I say “almost entirely” it should perhaps be read “entirely entirely”) on Amazon’s web services API to function. Without that API, Library would be that much less fun —and all the more effort— to use. The problem appears to be four–sided:

  1. There are too many links back to Amazon
  2. The cover art is oftentimes absent
  3. The cover art is sometimes screwy
  4. Amazon’s selection isn’t as impressive as you thought

Yup, there are a lot of links to Amazon, but those links are a by–product of the web services API agreement. There’s a fine balance of give and take that Amazon oversees, and they’ll pull the plug on your developer account if you don’t send enough referrer sales their way. It happened to Wincent Colaiuta with Synergy, it’d happen to Delicious Monster if they weren’t careful. Sure, it’s a bitch, but if you want access to the world’s biggest source of product metadata you just have to grin and bear it.

As for the cover art issues, Library can only deal with the art it receives. Though I will express bewilderment at why Delicious chooses to skip over some art (my screenshot in yesterday’s post shows missing art for Mad CaddiesDuck & Cover and Lynne TrussEats, Shoots and Leaves… both of which are perfectly artful at, screwy art and wholly–absent art is something of a problem on Amazon’s side and nothing of Delicious’ concern. I really don’t think Google Image Search is a good answer to the question —it’s too unfocussed— like using Googlism for a product description.

And as for the selection, the World’s Biggest Selection™ does lack a lot of rare, out–of–print stuff… duh. Amazon is a store, not a library — it only stocks items that are in supply and in demand, which is exactly its downfall in the kind of application whose purpose is to identify all those things you already own. Again, not really Delicious’ problem, but in using any service except Amazon (say… eBay, or Google) the signal–to–noise ratio is much too low, and it’s hardly worth the risk. We need accurate, structured data that doesn’t require fuzzy parsing or screen scraping. We obviously need something with a decent API, but where do we get it?

This is the part where I put forth a solution to everyone’s problems, not taking into account the legality of my proposal, nor the difficulty and cost. It’s the part where I explicitly deal with what’s wrong with Amazon from Delicious Library’s perspective, what’s wrong with the CDDB from iTunes’ perspective, and what’s wrong with all of them from the developer’s perspective of any number of future apps that deal with everyday media: I want an international product information wiki.

Displace Amazon as the leading source of product metadata and cover art. Displace the CDDB as the programmer’s first choice for CD track listings. Build a new database as expansive and as ambitious as Wikipedia, where people the world over can contribute accurate information and keep it up to date; so those stupid mixups you get with track names and artist names on compilation CDs don’t happen, and you can actually get accurate metadata on that out–of–print edition of James Joyce’s Dubliners. Do it so iTunes can fetch all the info it takes to fill up its ID3 tags, including the cover art, and so Delicious Library can operate without a billion links to Amazon. Do it for the betterment of the online consumer community. Who’s with me?

Screw it, I’m bored already. Let’s go make a Homestar Runner wiki instead.