Other than some pretty sketchy interface ideas, I really don’t know what I want to see in 10.5.
A single rumour I saw after WWDC speculated about Core Animation and the Finder as the prime suspects behind Jobs’s “Top Secret” gambit — and that just happened to dovetail with a whimsical wish I’ve had ever since Dashboard came out: I’d like to see Finder windows revolve in 3D space like the Widgets do — why not put all the prefs on the virtual backsides of Finder windows? But there’s a giant caveat: other than the coolness factor, what ease-of-use advantage does this give the Mac? Doesn’t it just duplicate existing functionality, albeit in a more eye-catching way? Yup. Is that enough? Maybe it is.
There’s been a lot of talk about a tabbed Finder, but as a long-time PathFinder user I found the utility of tabs to be pretty marginal after using them for just a few minutes — and I’d been very eager to see them implemented in PFv4. You live and learn. Tabs don’t automatically simplify file management; in fact, they can complicate it. In PF4, browsing up and down the folder hierarchy changes the name of the tab, which I find highly undesirable, since for me the tab name is a reference point. In my opinion It’s got to remain fixed for it to be of any use (I think the single best feature of PF is the drop stack, and they’ve had that at least since v3, when I discovered PF).
So if Apple’s planning a tabbed Finder, they need to do it the right way (which as a rule Apple tends to do, but they’re not perfect — look at all the grumbling about UI inconsistency in OS X; or the first brushed metal QuickTime Player and its stupid thumb wheel volume control, which was justly and righteously ridiculed into oblivion). I think a better place for tabs would be in Keynote or Pages, as a way to more gracefully handle multiple open documents (as Dreamweaver does).
So, back to Leopard. What do we want? The same thing we’ve always wanted — more power to manage the ever-increasing volume and variety and complexity of the data we store on our Macs. Enough eye candy to differentiate the Mac in the OS marketplace and keep the user base from getting bored. And enough (but not too much) corporate imagination from Cupertino to surprise us every once in awhile because sometimes we don’t know what the hell we want until we see it.
To be really practical for a moment — I keep hoping for a drastically-improved Help system. For all its renowned ease-of-use and elegance, there are still far too many times when the Mac’s behavior is inexplicable. For example, I don’t think a user should ever have to resort to a program like Printer Setup Repair to replace a missing or corrupted CUPS directory, as I’ve had to do twice in four years after my Mac simply refused to (a) print and (b) tell me why it wasn’t printing.
I also generally think more priority should be given by the system to user input (keyed or moused) over background processes. I think error codes should, to the extent possible, be replaced (or at the very least supplemented) by plain English explanations, and if there are several possible explanations for system misbehavior, they should be presented to the user — in plain English — in a brief, bulleted list, starting with the most likely explanations at the top, along with links to the appropriate local or online sources should additional information be needed. And those links, when followed, should not point to information that merely repeats what’s already been said in the error message.
Granted, this kind of improvement may not be the most exciting work for programmers or marketers, just as anyone who runs a city would (understandably) much rather build a fancy new skyscraper than fill in all the pot holes — but work of this sort, properly done and vigorously promoted, would be a massive selling point for the Mac, and a genuine realization of the ease-of-use promise it has always held out (and so often fulfilled) to users.
Of course, don’t forget those revolving Finder windows. :-)