waffle

The new Fireworks CS4 beta produces a user interface that looks thoroughly custom and unlike anything offered by any Adobe application I’ve seen going back the entire CS era, including Lightroom — sorry, Adobe Photoshop Illustrator Soundbooth Lightroom. ™.

This kind of stuff intrigues me and deserves a deeper look before I pose any sort of opinion. But among my first reflections is this idea: what if Adobe is entirely bulldozing their UI problem? Adobe’s problem right now is that they have Mac products that look like Mac products, except for those weird Adobe deviations which also vary from product to product, as anyone who has witnessed the steppers-who-look-like-popup-buttons in Illustrator can confirm. (Technically, they also have Windows products that don’t entirely look like Windows products, but very few Windows products do these days, including most of the ones bundled with Windows.)

By switching from what already has to be pretty custom UI to entirely custom UI, and providing platform-specific skews of this custom UI, Adobe could come out looking better. From “they can’t even get a standard modern Mac interface going” to “they have a custom interface that still works like a Mac app!”.

I recently had the privilege of setting up an HP laptop out of the box for someone. I’ve done this a few months ago with a similar model, and the out-of-the-box experience was so abysmal that I braced myself this time around. It would turn out that I was right to do so.

Here is an abridged list of what I had to endure:

• I’m asked to select one of three nordic language versions of Vista, followed by lengthy hard drive activity and restarts (probably copying of this setup to another partition). Note that they already know what I’m going to choose here: Swedish, Norwegian and Danish all have different keyboard layouts, and the laptop came with the correct keyboard. This is time-and-money-saving laziness on HP’s part.
• Vista installation ensues. I actually like Vista’s installation, and this step was relatively fast.
• Vista configuration ensues.
• Vista “hardware optimization” ensues — this is the driver installation/Windows Experience Index assessment stage.
• Now an HP tool tells me that it’s creating a backup partition. This takes several minutes, has a completely unhelpful progress bar (during the last stage, it sat at 0% for five solid minutes) and not only not a remaining time estimate, but in its place instead an “elapsed time” counter. Way to make this slow step feel slower.
• After all this, here comes the droves of HP crapware, installed with, “helpfully”, no interaction whatsoever. Uninstall programs, here we come.
• Finally, registration in a really weird window, taking up the full width, but only about 60% of the window (aligned to the top, not centered, natch). And of course, it’s not mentioned how to skip this step. (Or maybe it is: at this point I’m so winded from the barrage of weird, inconsistently styled HP dialog boxes that I skimmed this one, but I sure didn’t see anything like it.) Our old friend Alt+F4 works to exit.

Total elapsed time: 55 minutes. In contrast, all Macs I’ve set up have been usable within 10 minutes, discounting transferring information from an old computer, an option which isn’t even offered here. But this isn’t about Macs, it’s about respecting your customers.

The battery wasn’t in its slot already but helpfully attached as part of the “random crap” box; the “BTO” RAM wasn’t already installed (although this might simply be a problem with the dealer); I still haven’t gotten the Intel stickers off (the “warning: using a keyboard without the proper ergonomic training may cause your genitals to fall off in spectacular ways” sticker came off, for what it’s worth) and of course the installation took about 45 minutes too long because the geniuses at HP can’t prepare a fucking Ghost image.

This is HP’s fault. I quite enjoy most of the Vista setup which is way better optimized than XP setup in asking the right questions in one go. Most of what was weird, took time, felt unnerving, lacked in usability and was completely useless from a technical standpoint clearly came from HP.

Team HP is betting on you taking your day off when you want a new computer. And it’s true that it already as a rule takes a while to transfer everything over from your old computer to your new computer. But that’s an impetus to make the out-of-box experience as short and as smooth as possible, not an excuse to waste another hour of your time. That’s just being a complete, if perhaps slightly more profitable, ass.

Long-time readers will know that despite some initial worries, I quite like the Office Ribbon.

When Microsoft went about remaking the user interface of one of only two bread-and-butter products, they did their homework. They tried numerous mockups and threw pretty much anything at the wall to see what would stick. They based contentious decisions on observable user metrics. User interface benefits abound in the product — the Office button, while a victim of a hyperactive marketing department, adheres to Fitt’s law by swallowing top-left corner clicks; the status bar has received a major redesign and is filled with useful, customizable information and a standard page zoom control; standard tooltips have been replaced with meaty “this-is-X-it-is-used-for-Y-press-F1-for-help-on-this” explanatory tooltips that are shown beyond the Ribbon and are guaranteed to not obscure any other tools; “Galleries” show style changes in-line and previews live as were they already applied to the selection.

No, the benefits of the Ribbon as a concept have been proven in usability testing. The thing that gets everyone riled up is twofold: the theory and the reality of abrupt change.

If you were to go to people after Office 2003 had just come out, they would have told you to bank on two things: the interface sucking right now, and the interface not changing in a billion years, no way, because we knew what Microsoft’s method of change was, and was slapping new paint on the same buttons or making the menu items you used disappear because you were some sort of bell curve-outlying freak.

That they did change the interface was a testament to some sort of change in the Microsoft culture, and I personally approve of it. But everyone involved knew that rearranging everything in the world’s most commonly used applications that aren’t web browsers or media players would require an enormous retraining effort. People didn’t expect the dramatic shift. Half of getting frustrated with the Ribbon is in just being frustrated with the change itself, and despite being real, legitimate criticism, I suppose that this will go away with time.

But the other half is a different kind of user interface criticism, which would still exist in a vacuum of “just Office 2007 and knowledge of a previous version”, regardless of Microsoft’s update habits. Finding rearranged commands is arduous. Going from n custom toolbars with arbitrary positioning to 1 custom toolbar with two positions is hard to stomach. Most sources of pain can be directly derived from one core decision: “a feature should only be exposed in one place”. This rule is broken in two places: the mini-toolbar (which appears in context, or in Office UI engineer parlance, on object, near selected text) and the customizable toolbar, but otherwise is maintained, as far as I can tell, across every window that uses the Ribbon, for every command in the Ribbon.

Which brings us to Office 2008, the first Mac Office version to come out since the Ribbon was invented.

Office 2008 has gone off in a completely different direction. (Contrary to some mainstream reports, I affix no “whence the Ribbon” astonishment to this sentence; it is simply a statement of facts.) The floating toolbars have been integrated into something that, if it isn’t, at least looks remarkably close to a standard Mac OS X toolbar. This is a bit confusing, since the problem the Ribbon was supposed to solve in Office 2007 was to deal better with the wealth of commands available. A good part of the Ribbon concept is just a fleet of toolbars, and Office 2008 instead crams many commands into a single toolbar. For two versions of the same suite, no matter how “light” one version is, these are very different ways of dealing with the same problem.

Most of Office’s commands, then, are outsourced to a mighty-morphin’ Inspector; a model that has worked well for Apple’s own apps, like Keynote. The interesting thing is that Microsoft bakes font and color selection into the Inspector, where most of Apple’s formatting choices are in the Format Bar or system standard panels, inconvenient as they might sometimes be. Overall, Microsoft’s Inspector gives the appearance of being very messy, having not only multiple top tabs, but multiple collapsible sections within each tab, each causing the Inspector to resize.

The part that has without a doubt led the Ribbon confusion in media coverage has been the Element Gallery; a set of tabs that hangs below the toolbar and which opens up galleries with templates. This works about as well as the in-Ribbon galleries in Word (for styles) and PowerPoint (for slides and transitions), but it is a brutal slap in the face in Excel, where the formula bar has been outsourced to a panel but the Element Gallery — which is of minimal importance in any spreadsheet, even a good-looking one — gets to hang around just below the toolbar. The Element Gallery may superficially slightly resemble one, but it is no Ribbon.

The really sad conclusion is that I don’t know what sort of conclusion to draw from all this. Office 2004 was reasonably un-Mac-like, and would have represented the old guard in Office design had it been left untouched. Office 2008 is instead some sort of franken-Office, being slightly more Mac-like at the cost of being, arguably, more complex and less customizable, while looking nothing like any Office version ever, including the new Office 2007.

Porting the Ribbon right off the bat is hard, because the Ribbon depends on being the main interface. The Mac OS X menu bar would have to go, or the Ribbon would have to be duplicated on all documents. The best compromise I have been able to reach is porting the spirit of the Ribbon tabs, including contextual ones for manipulating images, to a standard Mac OS X toolbar, and using a custom space below to host the Ribbon groups (whose usability remains excellent).

The Mac Business Unit is between a rock and a hard place here, and I would suggest that they too consult Jensen Harris for help. One of the best user interfaces in the world is currently shipping inside boxes for the Windows version of Office while the Mac version is plagued with inconsistency and indirection. Think about that for a while, before you go ask Bizarro for directions home.

There’s so much to speculate about this that I can’t even decide where to start. So let’s recap what we know that might be useful.

We know that the iPhone 2.0 software will arrive in late June (slide, iPhone Roadmap event). It seems like a profound waste of effort to launch anything earlier with, say, 1.2 in the meantime, so that brings the earliest iPhone shipment date to June 20th; not only is it far enough into June that it could be considered “late”, it’s also the first Friday of late June, and Apple is fond of big product launches on Fridays. The likely date remains June 27th, which is equally Friday-ish and equally June-ish but even more late-ish.

We know that the new iPhone will have 3G. We even know that it’ll have high-speed 3G; AT&T has been upgrading their network to support HSUPA, and Telstra, a confirmed Australian carrier, committed to the new iPhone supporting 42 mbps speeds [next-generation, "Evolved HSPA"] by Christmas — where “by Christmas” couldn’t logically resolve to anything other than network build-outs instead of a new new iPhone, released in the intermediate term. And we know that, like any half-brained 3G device, you can temporarily turn 3G support off and fall back to GSM-level performance… and battery life.

We know the level of “enterprise support” (read: ActiveSync, Cisco VPN and remote wipe) that we’ll be able to expect. We know, roughly, the proposition for creating iPhone apps. We even know the shape of things to come.

What don’t we know?

Precisely what will 3G support entail? Will we get video calls? (On the one hand, Apple’s well-positioned to turn this from a dud to something moderately interesting due to all their technical work and cultural heritage in audio and video. On the other hand, isn’t mobile video calls fundamentally just a bad idea?) Will surfing and network activity really speed up that much? Loading freeze-dried sites from bookmarklets using the current iPhone software takes almost as long as loading the site itself, which suggests an efficiency problem in the browser and rendering software, not the network hardware.

Precisely what else will the new iPhone get? With 3G speeds and supposedly improved hardware, maybe we can fill the last square in the 2×2 grid and get video recording; I know Apple could build a better mobile video editor than anyone else. With that in there, is MMS still off the table? There’s been sightings of — cringe — “iControl“, which would be the well-duh way of remotely controlling iTunes playback we’ve all been hoping for since AirPort Express was launched, and even before that.

How about some functionality that, in the same vein as MMS, is already in every other mobile phone? Sending and receiving contacts, photos and perhaps music over bluetooth or email. Maybe a fifth of what the jailbreak software cottage industry produces exists solely to serve these needs. How about being able to edit and create ringtones? We already know that the iTunes Store approach is complete, money-milking bullshit, and that using 30 second snippets to annoy people on the bus is well into fair use. And besides, Apple already supports it… provided that you own a Mac and haven’t uninstalled GarageBand.

We will see at WWDC.