Home

Stephen Hackett's weblog about Apple, Apple history, technology, journalism and design.

Progress Shouldn't Mean Regression When It Comes to Apple Software

There’s been lots of noise over the last couple of weeks about the new version of iWork, and the fact that Apple stripped many features from the desktop applications in the name of inter-operability with iOS and iCloud.com.

Earlier this week, Apple announced that it will be adding several features back to iWork for Mac over the coming months. The article is an interesting one, but this particular paragraph is worrisome:

In rewriting these applications, some features from iWork ’09 were not available for the initial release. We plan to reintroduce some of these features in the next few releases and will continue to add brand new features on an ongoing basis.

While the Mac App Store leaves existing iWork applications in place, the new versions require a file format change. Files can be exported back to the older format, but this leaves things in the hands of users that probably shouldn't be. For most users, the new iWork is a one-way street.

While it’s easy to speculate what happened with iWork, I can’t help but think it is a man-power issue.

As we discussed on The Prompt this week, I fear that the recent update’s issues stem from mismanagement and the inability to set (or meet) reasonable deadlines.

If this is indeed the case, it wouldn’t be the first time Apple’s struggled to meet deadlines. Here’s a footnote from my OS X Mavericks review:

Apple pulled engineers to get the iPhone out the door. While developing Leopard in 2007, The same thing appears to have happened this year: John Gruber reported that OS X engineers were pulled to work on iOS 7 leading up to WWDC. While I can’t speak to the validity of the report, having run both betas for some time, I believe it to be true.

Some have suggested that Apple’s software divisions are short-staffed, and that perhaps the company should just bulk up on engineering talent.

Outlined in his book The Mythical Man-Month, Frederick Brooks suggests that adding manpower to projects makes them even later. While starting with more software developers can be beneficial, adding them mid-stream is usually a bad choice.

While the root of the problem may never been made clear by Apple, the pattern the company has set is easy to follow. Let’s take a trip to the recent past.

In 2007, Apple launched iMovie ’08 as part of an iLife overhaul. Here’s a paragraph from that press release:

With iMovie ’08, Apple’s movie-making software has been completely reinvented to let users rediscover and enjoy their video library, make movies in minutes instead of hours, and share their movies with family, friends or the entire world in a snap. iMovie ’08 can import video from the latest AVCHD, HDV and DV camcorders, as well as from digital cameras, and displays a user’s entire video library whether it’s stored on internal or external drives. Users can preview any of their video clips by simply moving their mouse over the clip to “skim” through it forward or backward at any speed, including faster than real time. Users can select video as easily as selecting text, make a movie by simply dragging the selected video into a project, then easily add a soundtrack from iTunes®, voiceovers, and elegant effects and cinematic titles. iMovie ’08 makes it as easy as a few clicks to enjoy your movies on an iPod, iPhone or Apple TV, or share with the entire world on YouTube.

iMovie ’08 sported a drastically updated UI and a slimmed down set of features from its predecessor (iMovie HD), and users were not happy. The new app lacked the sophisticated audio controls and plug-in architecture users had come to rely on.

David Pogue wasn’t impressed, either:

The enhancements in iPhoto, iWeb and GarageBand are great. But iMovie ’08 is an utter bafflement.

Most people are used to a product cycle that goes like this: Release a new version every year or two, each more capable than the last. Ensure that it’s backward-compatible with your existing documents.

iMovie ’08, on the other hand, has been totally misnamed. It’s not iMovie at all. In fact, it’s nothing like its predecessor and contains none of the same code or design. It’s designed for an utterly different task, and a lot of people are screaming bloody murder.

The new iMovie was, as Apple admits, designed primarily for throwing together movies quickly. It lets you scan through a clip to see what’s in it, isolate the good parts, and rapidly drop them into a sequence.

Apple ended up putting iMovie HD back up for download on its website, and while later versions of iMovie improved things, the application’s reputation suffered for years.

As if the video application team in Cupertino hadn’t learned its lesson, in 2011, the company released Final Cut Pro X:

At the heart of Final Cut Pro X is the Magnetic Timeline, a trackless approach to editing your footage that lets you add and arrange clips wherever you want them, while other clips instantly slide out of the way. You can use Clip Connections to link primary story clips to other elements like titles and sound effects, so they stay in perfect sync when you move them. You can even combine related story elements into a Compound Clip that can be edited as a single clip. The groundbreaking new Auditions feature lets you swap between a collection of clips to instantly compare alternate takes.

The update was a big one, and Macworld suggested users would need to re-think some things:

Most of the features introduced in FCP X are welcome and badly needed. Some are long overdue. Still, others are positively jarring and require a change in mindset to appreciate.

[…]

Much of FCP X version 1.0 is staggeringly impressive. There’s no doubt that Apple’s under-the-hood engineering will make life very fluid for some editors. On the other hand, this product at launch is nowhere near perfect, and is sure to defy expectations and disappoint many longtime working video pros who spent the last year looking forward to an upgrade of Final Cut Studio.

The video editing community wasn’t hip to many of the changes, and a petition with over 600 signatures quickly surfaced. Conan even joined in the debate.

Apple’s Chief Architect of Video Applications Randy Ubillos claimed that “Final Cut Pro X 1.0 is the beginning of a road, not the end,” promising updates that would restore features that had been stripped from the initial release.

That’s just what the company did. Each of the subsequent releases has packed in tons of features, but to this day, the Final Cut Pro X name is garbage to many in the industry, despite Apple’s PR efforts.

Of course, Apple isn’t the only company to do this sort of thing, and has shipped firmware updates to updated machines post-launch for years. But that doesn’t mean it’s okay.

Apple isn’t a software or hardware company, but an experience company, and installing new software that does fewer things than its predecessor sucks. Being worried about pushing the "Update" button is a bad user experience.

That’s not to say Apple shouldn’t ever remove features. Some things must be sacrificed to move platforms forward. What if Apple hadn’t switched to Intel to preserve OS X’s Classic Environment? What if the company had pursued 64-bit Carbon further?

Progress is not the same thing as regression, and the latter keeps being an issue with Apple’s non-system software. Updates to applications shouldn’t drop features without good reason. Apple double-back to re-add features is clear proof that these regressions aren’t as intentional as some would believe. (Not to mention the fact that the company will keep old copies of software on users' disks or provide old installers.) Apple seems to be in over its head (or in a rush) in some areas, and releases like Final Cut Pro X and iWork set a trend that worries me, and one that I hope Apple can put to an end.