Last Week on My Mac: Homage to Catalina

Some will have come away from WWDC with the impression that it’s all gone a bit flat. The only hardware announcement, that of the new Mac Pro, isn’t what many had expected, and won’t ship until the autumn/fall anyway, as if you could even afford it. There’s no sign of Apple breaking away to its own processors for Mac systems, and the headline changes in macOS 10.15 Catalina look quite underwhelming – pensioning off iTunes at long last, and letting you use an iPad in Sidecar.

As I always write, the devil is in the detail, and this year there was so much new detail that some major improvements to macOS barely got mentioned in the presentations. If you were seeking evidence that Apple still values Macs and macOS, you wouldn’t have to look far to see this torrent of the new and changed. Indeed, those trying to convince us that Apple wants to merge macOS with iOS got a shock, when instead of reducing the number of its operating systems, it has added one with iPadOS, albeit in name only for the moment.

Many of us were hoping for Time Machine 2.0, and, for now at least, we’re disappointed. But look at the detailed accounts of new and improved features in the macOS file system – definitely macOS-only – and two announcements in a single session should be striking.

The first is not just the fact that in Catalina the system itself is moved onto its own read-only volume, but the engineering which has gone into making that as transparent as possible. When you upgrade to Catalina, your current boot volume Macintosh HD or whatever will be divided into two: one data volume for read-write files including Home folders, the other dedicated to storing macOS system files. These form a Volume Group in APFS, in which a new type of bidirectional symbolic link, the Firmlink, joins the two directory hierarchies together seamlessly. Volume Groups and Firmlinks are new with 10.15, and haven’t yet been documented in the APFS reference.

More directly relevant to the future of Time Machine on APFS is what Catalina’s improved Apple Software Restore can do with snapshots. Time Machine started to use APFS snapshots back in High Sierra, and now seems to work with ‘snapshot deltas’, which show what has changed in a volume between snapshots. I’ll explain this more in a separate article, but for now Catalina’s command tool asr looks to be a valuable source of new ways to back up, replicate and restore APFS volumes.

You also have to look very carefully at the detail of new extensions to privacy protection. A lot have seen the headline changes that bring a user’s Desktop and Documents folders within the scope of privacy controls, and immediately assumed that Catalina users are going to spend much of their time responding to TCC’s consent dialogs instead of getting on with their work. That isn’t, of course, what will happen.

Where you, the user, choose to access a file in any of the newly-protected areas, your interaction with macOS will proceed normally, using existing tools such as the Open and Save dialogs, and the Finder. Your intention is clear, and TCC doesn’t need to check that again. However, apps which quietly start trying to access files in those folders without going through such controls will have to be given your permission to do so. This affects many of my apps, which scan folders checking quarantine flags and suchlike. You’ll simply need to add them to a list such as Full Disk Access – which is your decision, not the app’s.

Other oversimplified summaries have led some users to believe that the days of scripting a Mac yourself are over, unless you happen to have a developer ID to hand and can notarize them. Apple was very explicit in explaining that, while it wants to move users away from running unsigned apps and code, even when it makes that more difficult in a successor to Catalina, users will still have that as an option, just not the default that it is at present.

Perhaps the one area which has elicited greatest adverse comment is the prospect of changes to the user interface being driven by an influx of apps ported from iPadOS, and the likes of SwiftUI, a new declarative approach to coding the interface. Naysayers see this as a reduction in the riches of the macOS interface to what is common with that offered by iOS, with its completely different user environment and purposes.

Apple isn’t though replacing the Mac’s AppKit and existing interface with its new SwiftUI. In a few months time, we should be able to run the first apps built using the latter. SwiftUI will undoubtedly evolve, and if it doesn’t turn out as good as Apple expects, users and developers will determine its future irrespective of any ambitions that Apple might have for it.

The macOS user interface can’t stand still, deemed perfect and immutable for all time. Users and apps change, and the ways in which we expect to interact with our Macs change too. It’s not that long ago that standard input devices were a keyboard and one-button mouse. Now many Mac users have replaced their mouse with a trackpad, either built into a laptop, or sat next to the keyboard.

Apple is rightly putting accessibility high in its priorities, and not just for those who are officially deemed as having special requirements. Our needs and tastes change with age even among those deemed to be ‘normal’. Our six year-old granddaughter is far more adept than I am on her iPad, other users like voice control, and old fogies like me want to stick with a trackpad and keyboard. Over time, the population changes, as do users.

The clearest message from WWDC is that Apple isn’t letting its products lie fallow. Its engineers and other staff introduced us to a huge range of the new on all its operating systems, from macOS to tvOS. Undoubtedly not all of this will prove good, but there’s a great deal of future here for everyone.