Sherlock and iTunes
I was thinking today about interface design and specifically the work of the Mac OS (X) Human Interface Guidelines. I have grown to love the Mac OS immensely over the years, and value all the effort that goes into creating a consistent and usable operating system. However, I feel that in a few specific areas, Apple has either done the wrong thing either through active decisions or through passivity (i.e. not doing what they ought to do).
The specific case I would refer to (more cases will become evident later) is file searching. At first, Sherlock provided a well intentioned search tool for the Mac OS. Setting aside the recent decision to turn Sherlock into a Watson look-alike (web-services tool) while moving the file search functionality back into the Finder, the history of file searching on the Mac OS reads like a list of things not to do if you value the principles laid out in the HIG.
To demonstrate this point, let’s consider what I consider to be what a Mac file search tool should have been: The BeOS Find Panel. Let’s start by placing a value on consistency of interface - when we see a window that looks like a folder window with files in it, it should always act in the same way. This construct immediately breaks when using Sherlock, or it’s Finder-based Jaguar-equivalent. While the results window in Jaguar finally looks like a regular Finder window, it has both a “refresh” button, and an info pane showing a selections actual location on disk.
The results window isn’t a folder, despite the fact that Finder windows usually represent folders. Consistency should dictate that a Finder window is the graphical representation of a folder. Be did the right thing here by declaring that search results were folders. You could place the folder on the desktop, open it and see the files in it, and use it in any other way that you might expect a folder window to act. Even better, since BFS was designed with this functionality in mind, search results were always real-time; If a file changed in a relevant way to a search, the search folder would be updated instantly.
Moving on, Be developers again did the right thing with media players based on the features of BFS. Media file attributes (bit rate, play length, artist, title, album, etc.) were stored as file system attributes. Combined with search result folders, this allowed dynamic playlists (folders) to be used in applications such as mp3 players which updated in real time. iTunes’ tries to pull off this same feature, using Smart Playlists, but is hindered by the need to contain all mp3s in the Master playlist prior to showing them in a smart playlist.
Once the idea of file attributes is expanded to contain more than just file creation and modification dates, ownership and permissions, things like ID3 tags are an obvious use of attributes. It is a real shame that Apple continues not to see the extreme benefit this functionality would bring to users. iTunes would be greatly served by using the Finder for playlists instead of its current awkward interface. Not only would this benefit iTunes explicitly (which gets a consistent interface in the form of the Finder, and can make use of any Finder extensions a user might have installed), the Finder benefits with the ability to see and use a wealth of file meta data (ID3 tags, Email headers, etc.). I for one would greatly value the ability to use ID3 tags in the Finder to replace actual file names.
At this point, I’m only re-hashing old gripes about non-Be systems. Yet I feel my original point stands: the Mac OS (X) would be greatly benefited by taking the principles of the HIG to their logical conclusions (or perhaps their logical next-step). In my opinion, Apple is staying a step ahead of other operating systems with the technology of OS X, but only just. If they were to follow that next-stop in the HIG process, they would immediately find themselves with an OS that far surpassed most others technologically, while having a large enough user base to avoid the problems Be faced - the technology isn’t quite enough for success, but it must be seen as needing these revolutionary yet rewarding advances.