In my last Hot Cocoa column I posed the question, "How 'human' is the Mac OS X interface?." It is clear from the number of observer comments that were posted on the article that the Mac OS UI (user interface) is something that many of us feel very strongly about.
That column focussed on the general paradigm shift that has come with the Mac OS X UI, as exemplified by the controversial Dock. This time I'm zooming in on specifics rather than paradigms, and the object under discussion is the Finder.
Unlike other aspects of Mac OS X, opinions on the Finder tend to be much less controversial. Agreement is almost universal: it sucks. But rather than just list the flaws of the Finder one by one, I want to take a different angle this time. Although I will mention the problems and issues with the current Finder build, my main goal here will be to talk about how Apple can address these faults and how they can turn the Finder into something that is a revolutionary leap beyond that which came before. My argument will be that the Finder has a "fantastic future" ahead of it despite its current status as a "foundering flop"...
Numerous critics -- among them Bruce Tognazzini (founder of Apple's Human Interface Group), John Siracusa (whose series of Mac OS X reviews are widely regarded as the best on the web), and Tony Smith (of Register fame) -- have argued that the Finder has serious shortcomings. Sometimes the problems could be considered "lacks," at others "paradigm problems," at others "bugs," and at others still, "missed opportunities." I'll look at each of these in turn and offer some counter-arguments for why these issues are likely to be remedied in the future.
What's lacking in the Finder
This is the area in which critics of the Finder have been most vocal, not only Mac media commentators, but legions of users on various forums around the Web. The claims are that the Finder is too slow (though last night's update to 10.0.4 has gone some way to improving this); it lacks optimization (though again, the new update seems to have helped); it appears to have been inefficiently coded; it uses the less mature Carbon API instead of the highly-evolved Cocoa API; it is inadequately threaded, often resulting in "blocking" (the spinning rainbow busy cursor); and it loses key functionality that was available under the Mac OS 9 Finder (spring loaded folders, pop-up windows, comprehensive and customizable contextual menus, Finder labels).
Despite all these causes for complaint, there is reason to hope. Even Steve Jobs must concede that the Mac OS X Finder is inferior to its Mac OS 9 counterpart. The only areas in which it clearly holds an advantage are its ability to finally exceed the 31 character filename limit and its useful multi-column view. Apple is working long and hard on developing the Finder to remedy its faults and ensure that it can claim to be the best file-browser in the world. Check out our forum thread on the new update to see how this has already been improved.
Derived from the NeXT file browser, a Cocoa version of Apple's Finder has been developed in tandem with the released Carbon version at every step of the way. In fact, the earliest developer previews of Mac OS X included a Cocoa Finder. In the Rhapsody days, Carbon didn't even exist. When Apple is ready, we will all get our hands on the much-sought after Cocoa Finder.
In much the same way Apple is working hard on replacing missing functionality such as spring-loaded folders, Finder labels and better contextual menus. Pop-up windows as we knew them in Mac OS 9 will never return, largely due to the fact that the Dock now occupies the bottom edge of the screen; however, expect to see improvements to the Dock which will enable users to drag items into pop-up folder menus much as one would've used pop-up windows in OS 9.
More abstraction: away from spatial navigation
Mac OS X's Finder represents a key departure because it breaks with the old Finder's spatial navigation model. In OS 9 each window had its own, unique place on the screen at a precise location and with a specific size. It was not possible to have two views of the same folder at once. Mac OS X differs here because the model is no longer spatial but more like that of the World Wide Web. Users can open multiple windows each viewing the same folder. It is even possible to use a Finder window to look at the items on the desktop at the same time as them residing in plain view on the desktop itself.
The strength of the old model was that it was concrete and simple. It was one of the things that made the Mac OS easier for beginning users to understand. The new model, however, brings with it more flexibility and more ways of working. Those who pine for the old way of working can have it back by hiding the Finder toolbar (a flawed implementation of the old system at the moment, but expect to see it improved in future Finder releases).
The most interesting aspect of the new browsing model, however, is the logistic difficulties that it introduces. For example, we all want the Finder to remember our settings for icon size, position and arrangement. But what does the Finder do when we have two views of the same folder open and we arrange each one differently. Which one should the Finder remember?
This is a definite problem that Apple must resolve if it is to lift the Finder above allegations of being confusing and inconsistent. (Any observer comments or suggestions about this one would be greatly appreciated!) At present the options for managing multiple-views of the same folder appear to be as follows, listed in order from least- to most-intuitive. In each case it must be determined which window's settings are to be considered the authoritative view settings for a given folder:
- Always honour the view settings of the window that was opened first
- Always honour the most recently closed window's view
- Always honour the most recently altered window's view settings
- Any changes that are made to one window are automatically "copied" to the other window(s) by the Finder on the fly
- If you double click on a folder for which a view is already displayed elsewhere in another window, bring that window to the front instead of opening a new window or creating a duplicated view in the current window
There are some issues with the new Finder that can be classified as nothing other than bugs. There is very little that we can do about these other than waiting patiently for Apple to rectify them, and occasionally pestering them with bug reports and feedback.
Here I'm speaking about issues such as window sizes and positions being remembered inconsistently or not at all; the states of disclosure triangles in list view and the position of the scroll thumb in the scroll bar not being remembered when you use the "back" button or re-open a previously-visited folder; the Finder allowing you to drag locked files to the trash and later displaying an ambiguous error message when you try to empty it; and flaws in the system for selecting icons by typing the first letters of their names. Perhaps worst of all is the Finder's maddening habit of truncating long filenames in the middle even when there is an abundance of space to display them in wide open windows or on the desktop. I believe that we can all feel confident that Apple will remedy these kinds of problems as soon as possible. I am still checking out the new update, 10.0.4, to see if any of these issues have been resolved.
Before I conclude I'd like to talk about a couple of missed opportunities that Apple could've implemented in the new Finder and has not as of yet. One is the lack of functionality of the Finder shelf, which unlike its NeXT counterpart cannot be used for file-copying operations. In the meantime utilities like Dock-It are stepping in to fill the gap. Apple would do well to follow the lead of third party developers like this who are responding to user demand.
Speaking more generally, Apple may have missed a key opportunity here to make some paradigm changes with the file browsing interface. When they introduced their desktop model with the first Macintosh in 1984 the changes they brought in were quite revolutionary. Icons, the desktop metaphor, and files and folders. With Mac OS X we are still left with the same metaphors, even if they have been extended in some ways and "prettied up." It seems that interface designers are reluctant to think outside the square and come up with something really new. Apple's 128x128 photographic icons are a good step forward, but while the icon metaphor remains the same their potential will be limited.
For example, consider Tog's argument that Apple should be developing the "pile" metaphor for which it holds patents.
A pile is a loose grouping of documents. Its visual representation is an overlay of all the documents within the pile, one on top of the other, rotated to varying degrees. In other words, a pile on the desktop looked just like a pile on your real desktop.Piles, unlike today's folders, gave you a lot of hints as to their contents. You could judge the number of documents in the pile by its height. You could judge its composition very rapidly by pulling through it.
This is a line of argument that I am inclined to agree with and I hope that Apple moves in this direction in the future. This does not mean that we should throw away the familiar metaphor of the folder. Rather, we should look at ways of extending and improving it. If Apple doesn't do this it runs the risk of being left behind...
The way forward
It is doubtless that Apple will address many of the concerns mentioned in this column. Apple has worked very hard on Mac OS X -- some would say it has achieved the impossible by unifying the power of Unix with the simplicity of the Mac -- and it will continue to work hard improving it in the future.
In addition to the fixes required by the current problems in the Finder we can hope and even expect to see a range of enhancements and refinements which return the Mac to its place at the top of the UI pile. Some of these touches will be technical in nature -- like the ability to manually resize columns individually in multi-column view; or the implementation of intelligent auto-resizing in that view; or the option to show information in addition to file and folder names in the same view; and the ability to specify multiple sorting criteria in list views (in other words, an order of precedence such as "date then size" and so forth).
Other changes will be more subtle, yet more profound. We can hope to see greater iconic feedback in the future: folders "bulging" according to how much they contain; icons looking dusty and cobweb-covered to indicate they last modification time; animated icons that hint at folder contents or which provide detailed previews of file content; folder items that move (vibrate) or pulse (in color) to indicate that they are being written to, and so on.
We can be confident that such innovations will take place in the future. It's just a question of when. The important thing is that Mac OS X provides the kind of architecture that can accommodate this kind of variation and modification. Furthermore, Apple prides itself on its user interfaces, so we can expect them to be riding the crest of the wave. As I often say, it's an exciting time to be a Mac user.
Wincent Colaiuta - firstname.lastname@example.org