Tuesday, April 23, 2019 [Tweets] [Favorites]

Collision Course

Jeff Johnson:

You could argue that Apple is wasting resources trying to put “Pro” features into iOS. They’ll be mostly wasted on people buying the cheapest, smallest iPads. The idea of iPad as a Mac replacement seems to be ideologically-driven by a vocal few.

Unfortunately, Tim Cook is one.

Mac unit sales increased 75% between 2009 and 2018. Empirically, iPad has not replaced Mac. Not at all. And given that iPad ISP is $422, have to conclude:

1) iPad is a supplement to Mac, and/or

2) iPad is a device for people who wouldn’t otherwise buy a Mac (maybe cheap PCs?)

Too bad Apple is no longer reporting unit sales.

Dan Masters:

Interesting point – the Apple community bubble assumes much about how the general public uses particular products, but how much of it is true?

It’s hard to know. My mother’s 2017 Mac is newer than her iPad Air, so her wallet is voting for the Mac, but she spends far more time on the iPad. Yet if she had to give up one or the other, she’d keep the Mac. My father is similarly non-technical and only uses a Mac and phone.

Michael Potuck:

Apple’s VP of software engineering, Craig Federighi, answered the question of whether the company was working on merging iOS and macOS with a huge “No.” on stage at WWDC yesterday.

Federighi’s “No” is to UIKit/AppKit/Marzipan as the word “modular” is to the new Mac Pro. It was intended as reassurance but isn’t, really, because it can mean just about anything. Everyone is reading their own hopes and fears into it. I tend to think he meant only that there won’t literally be a single OS that runs across Apple’s hardware lineup. But that’s not saying a whole lot if it ends up that he thinks the future of Mac apps is the same UIKit code running on a slightly different substrate. If the majority of apps end up being designed for touch and limited by what UIKit can do, what difference does it make that it’s called macOS?

Apple has definitely been making iOS more like the Mac and macOS more like iOS. In some ways this is good, but both platforms also seem to be losing aspects of what made them great. iOS was inevitably going to lose its simplicity due to competitive pressure. The Mac situation seems unnecessary, and I think stems from a combination of vision (apply Apple’s favorite iOS ideas to the Mac) and neglect (no time/budget for Mac-only stuff).

Steve Troughton-Smith:

So the ‘Marzipan SDK’ is just the iOS 13 SDK all along — a checkbox in your iOS project settings. I theorized it might be, months ago. What will that mean for the separation between UIKit and AppKit world? The iOS SDK doesn’t include Mac frameworks like AppKit, AppleScript…

Colin Cornaby:

Was pointed out to me that if Marzipan and AppKit are partitioned, that not only would Mac apps not get Shortcuts, but Marzipan apps wouldn’t get AppleScript.

AppleScript doesn’t get a lot of attention these days but it’s invaluable to a lot of power user workflows. What a mess.

IF I take the Siri Shortcuts/Mac rumors at face value: Siri Shortcuts only being available to Marzipan apps implies that Marzipan still segmented from the Mac. That’s not great for people who want to mix Marzipan and AppKit, or transition one way or the other.

Brent Simmons:

For UIKit to become an app framework for Macs that makes great Mac apps, it will have to become AppKit. That is, it will have to adopt so much (perhaps with revision, of course) that it does pretty much everything AppKit does.

Brian Webster:

I had this same thought myself. Makes me wonder whether ultimately it will be easier to provide a way to combine/embed UIKit and AppKit controls/windows, much the way Apple did with Carbon and Cocoa (over the course of a few years).

Jim Rea:

When this was announced last year, I thought that was what they were doing. Very disappointing that this apparently isn’t happening. But we’ll see for sure in 40 days.

Currently, it doesn’t even seem like they are unifying the types.

Michael Love:

And my theory that AppKit = Carbon seems to be if anything understating what’s about to happen; Apple’s going to simply bring the interesting AppKit UI pieces to UIKit and deprecate the rest, by 2020 nobody will be writing AppKit apps anymore.

Drew McCormack:

In any case, in my discussions with people at Apple, it seems UIKit is not designed to replace AppKit, or even be used by skilled Mac developers. They simply don’t have that goal in mind.

However, Marizpan will surely have consequences that were not a goal but that nonetheless follow inevitably from its introduction.

Nicolas Zinovieff:

For sure, but what I’m mostly concerned about is the “hybrid side-effect”: so you fear AppKit and you do your app in UIKit because “easymode”. Then you want or need to add a feature that’s not supported. You will just drop that feature rather than rewriting the UI.

Steve Troughton-Smith:

If you missed last week’s @_connectedfm, @viticci had a pretty interesting scoop that he’d been sitting on re mouse support coming to iPad as an accessibility feature. As far as I’m aware, that is indeed in the works. I feel like every pro user will turn that on, day one 😂 with UIKit adopting all kinds of API for mouse support for the Mac, it makes a ton of sense to let that be leveraged on iPad too if somebody has a mouse connected. Not changing the OS for anybody using touch, but adding functionality for those who choose, or need it

iKyle:

Unsigned apps, full system access, full file system access, Steam and 3rd party stores all exist on the mac.

These are the mac’s strengths, they are what makes it so powerful and useful.

Apple seems to view all of that as a weakness, a failing to be corrected and eliminated.

They think anything less than the lock down and control they have over iOS is a security and strategic vulnerability to be removed.

Locking down and controlling the mac, ends it being a real computing platform and kills innovation on the platform.

One could argue that the most significant Mac changes in recent years have been putting up roadblocks to development rather than opening up new avenues of functionality like at the beginning of the OS X era.

Previously:

Update (2019-04-24): Colin Cornaby:

Without speaking to the accuracy of rumors/speculation: I worry about a Windows 8 style Pyrrhic victory. You can succeed in unifying all your applications around the lowest common denominator and filling the store with apps, at the cost of the platform you were trying to save.

The iPad and Mac have been on a collision course since the very beginning. But it feels like Apple is scared of the iPad becoming a Mac. So they continue to sit on their position of “we’re not merging the platforms” and we get situations like (possibly) this.

Take macOS, tweak it, and make it ProOS. Run it on the iPad and Mac. Have AppKit and UIKit and AppleScript and Terminal and Xcode and eGPU and drivers for everyone. This all feels like a giant exercise in keeping the iPad from becoming more open and locking down the Mac.

I love my iPad but I hate that I’m constantly having to grab my MacBook Pro to do things. I want my MacBook Pro on iPad hardware, not an iPad on my MacBook Pro hardware.

Brent Simmons (tweet):

Maybe because I lived through this — maybe because I’m a certain age — I believe that that freedom to use my computer exactly how I want to, to make it do any crazy thing I can think of — is the thing about computers.

That’s not the thing about iOS devices. They’re great for a whole bunch of other reasons: convenience, mobility, ease-of-use.

[…]

In a way, it feels like iOS devices are rented, not owned.

[…]

Macs carry the flame for the revolution. They’re the computers we own, right? They’re the astounding, powerful machines that we get to master.

Except that lately, it feels more and more like we’re just renting Macs too, and they’re really Apple’s machines, not ours.

Drunken Dogcow:

Agreed, except giving Apple a pass on iOS for their nanny authoritarianism while simultaneously arguing against the Mac heading the same way. This makes no sense to me. iPhones and iPads are computers and could be so much more if they weren’t locked down. Let’s not make excuses.

Ken Kocienda:

Over time, I’ve become and less less interested in software that doesn’t have access to its hardware. I want the option to “drop down a level” and then another and another, as far as necessary to get the job done right.

Steve Troughton-Smith:

Apple seemed to need a lot of reminding that the Mac was important. Now, it’s important to remind them why the Mac is important, so that they get any coming transitions right. The future is still being written

Will Cosgrove:

We’re always talking about how Apple created a Pro team to simply research how pros use their computers. They don’t understand why the Mac is important anymore.

Damien Petrilli:

It’s funny to see a claim that the Mac App Store is involving without developers because they didn’t embrace it.

When you known 100% of developers were forced to "embrace" the App Store on iOS and Apple still didn’t give a shit about them.

See also: Marco Arment.

Michael Love:

On this, a) he’s right but also b) the war between consumers and power users for the soul of personal computing has been raging for decades; this is just the latest front.

Scott:

I feel as though Apple with the Mac has become like the proverbial snake eating its own tail… while—muffled, mouth full—extolling how they were so prescient to “see” that the iPad was to become the natural successor to the Mac.

It’s sad.

#SquanderedAdvantage

Mark Bernstein:

The end of personal computing?

iKyle quoting Mark Pilgrim:

“Once upon a time, Apple made the machines that made me who I am. I became who I am by tinkering. Now it seems they’re doing everything in their power to stop my kids from finding that sense of wonder.”

Steve Troughton-Smith:

It was described to me recently by somebody in the know that AppKit will stick around as long as Photoshop needs it. I’m not sure the source was aware that Photoshop’s new cross-platform (‘iPad’) version was rumored to be destined for the desktop, like all of their revamps…

I very much got the impression that Apple was not planning to maintain two dueling UI frameworks on the Mac. If that is the case, developers and users really need to make it clear to Apple what parts of legacy macOS are worth bringing to UIKit and iOS, and soon

Jason Snell:

I fell in love with the Mac nearly 30 years ago, in the fall of 1989. It’s been the center of my tech world ever since, and I’ve been writing about it professionally for 25 years. And yet these past months, I’ve noticed something strange creeping into my thoughts occasionally while I sit at my desk working on my iMac Pro: iOS does this better.

That’s how I feel much of the time using my iPhone.

Update (2019-04-28): Kyle Howells:

I think most of the danger associated with the Marzipan transition is just Apple touching apps in any way. Because Apple can’t touch a mac app without turning it into a terrible iPad app lately, regardless of the framework used.

Eric Schwarz:

While I can sympathize with those who are concerned of the direction of the Mac and what it means for their usage, the vast majority of Mac users probably won’t care because they’re using Macs like oversized iOS devices anyway. Outside of iOS development, Apple has basically said that you have a choice: pick a Mac that suits your needs and live with the weird macOS purgatory that we’re in right now or feel free to buy literally any other computer out there.

See also: Accidental Tech Podcast and The Talk Show.

Update (2019-05-13): Mike Zornek:

For a meeting who’s goal was to clear things up, they sure did leave us speculating [about the Mac Pro].

Update (2019-05-14): Steve Troughton-Smith:

Apple’s dual operating system strategy is ill-equipped for foldable tablets; what happens when a MacBook is an iPad and an iPad is a MacBook? Does Apple sell two identical devices with different OSes? Or does the Mac stay with legacy form-factors only?

Update (2019-05-24): Steve Troughton-Smith:

A really interesting question we probably won’t have an answer to for years to come is whether UIKit is the ‘Carbon’ or the ‘Cocoa’ of this transition. I think the only appropriate answer is ‘yes’. UIKit is the present, and the developer ecosystem it will bring with it is incredibly important. AppKit is also the present, and it provides and powers the Mac as we know it.

I’m sure we will have great, genre-defining apps from both UIKit and AppKit on the Mac. With Carbon, we had iTunes, Photoshop, Microsoft Office and Final Cut Pro. Eighteen years on, Carbon is finally reaching its end date, and the transition of all these apps to Cocoa/AppKit is complete. If AppKit still has eighteen years left ahead of it, I think the Mac will be just fine.

Both classic Mac OS and NEXTSTEP came to an end; the Mac did not. I think everybody can agree the unified whole was much greater than the sum of its parts, but this was not clear at all in 1997. The future is still being written, and we each, ‘Mac developers’ and ‘iOS developers’ alike, will get to be there to help shape it.

Brent Simmons:

So, knowing how this has worked out in the past, why do I fear the reaper?

Because bringing UIKit brings no new power. If anything, it subtracts power. UIKit apps — at least so far — are all sandboxed and available only via the App Store. They don’t offer everything AppKit offers.

And, to make things worse, it’s reasonable to be somewhat skeptical of Apple leadership’s understanding of the platform. Daring Fireball quotes a source at Apple as saying they had “taken their eye off the ball on Mac.”

46 Comments

I'll quote myself from last year (https://mjtsai.com/blog/2018/06/20/webview-and-uiwebview-deprecated-in-favor-of-wkwebview/):

"Soon:
I was expecting to see massive improvements to [Swift] before [Objective-C] was deprecated...
I was expecting to see massive improvements to [UIKit/Marzipan] before [AppKit] was deprecated...
I was expecting to see massive improvements to [iOS] before [macOS] was deprecated..."

Gird your loins.

Also, I hate Troughton-Smith, but that doesn't mean I think he's wrong. To the contrary, I think he has a very accurate read on Apple's attitude toward the Mac and Mac users/developers, right down to the constant smirking trollface emojis he uses while mocking them (https://twitter.com/stroughtonsmith/status/1120683777766952960). I think his idea of where Apple is headed is correct, and that makes me hate it and him all the more.

@Michael I really hope porting UIKit to the Mac is the first step in making a more modern UI toolkit that will eventually be better than AppKit for Mac development. That's the best outcome, the worst is what you I and both fear - that the Mac just becomes a dumping ground for fairly underwhelming iOS apps that are optimized for touch and out of place on the Mac. But I have to ask this - right so much app development means mobile first (iOS, Android) web second, and desktop distant third (MacOS, Windows, pretty much never Linux). Also, a lot of "cross platform" apps are shipping as web stacks shoved inside a native container.

What should Apple do to make the Mac (and honestly the iPad) a platform that developers will target with native apps? Porting UIKit apps is a sort of moat that at least keeps the native container web apps somewhat at bay. It would deliver a lot of (un-maclike and perhaps bad) software to the Mac where we'd otherwise get a web app if anything. If iPhone-iPad-Mac becomes closer to a single target, you'd have to think the outcomes is more apps on the iPad and Mac.

I guess my fear is I've lost a lot of faith in Apple. I don't think they have a coherent vision of what the iPad should be or how to reinvigorate the Mac's software ecosystem. The best outcome is migrating UIKit to be the reusable core foundation for any Apple platform and to support specializations for phone, tablet, and desktop that makes each experience feel tailored to that experience. I think the Apple of today will likely give us touch first apps bereft of powerful features and crippled by sandboxing and walled garden app stores.

I really hope I'm wrong because I dunno where else to go if Apple wrecks the Mac.

Lanny Heidbreder

I fantasize about some of these announcements drawing loud “Boo”s from the audience at WWDC. But then I realize, what fraction of WWDC attendees these days cares about the Mac at all?

"What should Apple do to make the Mac (and honestly the iPad) a platform that developers will target with native apps? "

In 2001, when Mac OS X was released, if you wanted to release a hello world app, you could just open Xcode (which was provided with each Mac), build the executable and distribute the app on a disk image or a tar.gz on a FTP server.
In 2019, you need to get an Apple ID to download Xcode on the Mac App Store. You need to pay an additional $99 to get a Developer Membership and the required certificates. You need to codesign and notarize your app and your zip or disk image. Or if you want to distribute it through the Mac App Store, you need to submit it and then be told that your app is worth nothing, that there are already a million of hello world apps on the Mac App Store.

I would not recommend developing on the Mac or any Apple platform to a new developer.

"One could argue that the most significant Mac changes in recent years have been putting up roadblocks to development rather than opening up new avenues of functionality like at the beginning of the OS X era."

No need to argue here - I think this is just a fact.
I find it literally depressing. When did I last want to check out and play with a new framework that was released as part of a new OS version at WWDC? Instead I now only worry what the next thing will be that will cripple one of my Apps or distribution methods.

Mac aside, I personally find the iPad the most interesting platform as a developer. Since the release of the latest iPad Pro (which I think in terms of hardware is a great piece of technology and apart from the Mac the only other thing that Apple makes that I am interested in as a developer) iOS to me now drags even more behind with regards to what could be possible on this platform.

But even if Apple would magically fix all of these limitations on the iPad, there is still this other big problem that I see as an Indie: The marketplace.
Unless you make something completely mainstream (and even then), it feels there is almost no chance that you can sell an iPad App in the iOS App Store and make more than a bit of pocket money (subscription, freemium or whatever model you use).

The Mac still has more options and is more open, but I fear that one day not too far from now it will be some version of "App Store Only", therefore "Sandbox Only" and not to mention "-30% of Revenue Only". And then I fear the marketplace problem will be the same there too, with no way out.

"What should Apple do to make the Mac (and honestly the iPad) a platform that developers will target with native apps?"

That's a hugely complicated problem, that I think ultimately boils down to something simple: make the platform somewhere developers want to be. As "someone" mentioned above, it has gotten incredibly difficult to publish software on the Mac. That's fine, IF there's commensurate profit to be made and trust between the developer and the platform maker. However, Apple has been squandering that coin like there's no tomorrow, and profits for desktop software are rapidly getting eroded by the web (in many instances by "native" web apps). That puts desktop developers in a really tough spot. There need to be powerful--or at least well-supported--desktop frameworks that offer a compelling reason to avoid the "good enough" web wrapper.

Right now, the frameworks are shrouded in uncertainty, the development process is a living hell with a huge barrier to entry, and it's arguable whether ANYONE has really figured out how to deal with Electron apps and similar. It seems to me that Apple has no idea what developers even want/need, whereas Microsoft literally owns GitHub (Microsoft does a lot of things wrong, but their decisions in recent years have made it clear they're dedicated to at least understanding developer needs enough to try and exploit them).

Tim Cook: “Wouldn’t it be amazing if we rearranged these deck chairs?”

@Fred

What should Apple do to make the Mac (and honestly the iPad) a platform that developers will target with native apps?

Apple needs to show that they care about the Mac. They can’t convince developers if they don’t believe it themselves. They should make it easier to develop and publish apps (fewer hurdles, more reliable frameworks, better documentation) and monetize them (the App Store makes it hard to find/browse apps and makes trials and subscriptions needlessly difficult, fees are high). And then, for users and developers, add power features to differentiate from iOS. They should probably also develop their own Mac-focused apps, both to inspire and to dog-food.

But instead it looks like we’re getting attention diverted to a second native framework and wheel spinning and API loss with another architecture transition.

@Ian Yep, “squander” seems like the operative word for the last 10 years or so.

I remember the days when I would be delighted and surprised at the things that MacOS or the built-in apps could do. I'd think "I wonder if I can do XYZ?" and the answer was usually "Of course!" because somebody at Apple was thinking about all of the myriad ways that their software could be used. This ethic made its way to 3rd party apps too -- keep it simple and obvious on the surface, but anticipate and provide for what power users might want or need to do.

But in the past few years, many features have been removed from the built-in apps (such that it's good enough to kill competition, but still not A+, leaving us with gaps in functionality). Many things that previously "just worked" have had glaringly obvious bugs introduced (and never fixed). Or we get new features that are half-baked or don't work as expected (i.e. there are no extra "power user" affordances).

Given this, how are we supposed to have ANY hope that gluing iOS onto MacOS could possibly make it better?

For me personally, I can't think of a single app that I have on my iPhone or iPad that I desperately want or need on my Mac. Most iOS "apps" that I use are really just a fancy wrapper for a website (Kayak, Google Maps, my bank...) or are just not appropriate to have on a non-mobile device anyway (e.g. Uber, Delta Airlines, Overcast...). And the best iOS apps that would also be great on the Mac (Pixelmator, Fantastical...) already have native Mac apps.

Is this the case for anyone else? Like... WTF is the upside to making it easy for iOS developers to port their apps to the Mac? So we can have even more shitty apps in the Mac App Store? So Apple can cite phony numbers about how many apps are available on the Mac, with zero regard to the quality of those apps?

Am I missing something here? Why are some people apparently so excited about this? I still don't understand. It seems like an impending disaster.

@Ben The only one that comes to mind for me is Overcast. I probably wouldn’t use it to play podcasts, but I’d like to use the big Mac screen and keyboard to manage them.

@Michael -- Yeah I almost noted Overcast, only because I know that Marco would make a killer version of it for the Mac. But I don't have the same faith in most other developers. And frankly, even having used Overcast since late 2014 I've only ever used the website version of it on my Mac <5 times.

(I mean, no faith in most other *iOS* developers... because on the whole, I find the majority of iOS apps -- even? especially? from big name developers like Google, Spotify, my bank -- to be fairly terrible UI and bug-ridden, despite iOS apps being "simpler")

@Michael — That's a very good and succinct list of things Apple could do to reassure Mac developers. Unfortunately, it doesn't seem like Apple will do any of it.

One thing I would add to that list: Apple need to start shipping a stable OS that doesn't have concerning regressions mid-cycle. In addition to the Core Animation bug you highlighted previously, looks like Brent Simmons has run into a notable WKWebView latency regression in 10.14.4 as well.

@Ben I don’t use the web version much, either, but that’s only because it’s so limited.

@remmah Yeah, 10.14.4 seems to have introduced a whole bunch of regressions. One of my apps is currently in App Store limbo because of an undocumented code signing requirements change that Apple has yet to be able to explain to me. Xcode 10.2.1 also added a crasher to the Swift compiler that forced me to rewrite a whole bunch of code. Remember when they introduced APFS (the initial version, with the Unicode problems) in a spring maintenance update?

"Is this the case for anyone else? Like... WTF is the upside to making it easy for iOS developers to port their apps to the Mac? So we can have even more shitty apps in the Mac App Store? "

Main upside is for games and GPS oriented apps which tend to not exist for the Mac. (In part due to the lack of a laptop with cellular) There's other apps that'd be nice to have such as HomeKit configuration apps, health oriented apps (assuming the HealthKit stuff is accessible). So I'm rather surprised by people saying this lacks utility.

I do think there's reasons to worry, mind you. But they mainly come from the quality of recent Apple applications. Apple's mostly abandoned the pro market and has tended to strip features from its other apps. Even the consumer apps like Photos are worse than the consumer apps from the prior generation let alone there being an Aperature type app with the Photos backend.

"The iOS SDK doesn’t include Mac frameworks like AppKit, AppleScript…"

It's hard to see them making it so say the AppleScript framework couldn't be accessed. Unfortunately the reality is that Apple's never really cared much about AppleScript or the fantastic AppleEvents underlying subsystem since the days they weren't beholden to print for survival. Even apps that support AppleScript like Adobe typically offer most scriptability support in their own scripting systems. AppleScript itself is a horrible language that Apple should really have replace in a robust fashion years ago. (That poorly designed Javascript "replacement" doesn't really count since there are plenty of things you can't do in it.) More significantly Apple didn't ever really push the support and evangelism to get developers on board. One can hope that for all its limits and flaws that Shortcuts might expand in that direction. Although it's hard to see how it will. There's a lot of short term thinking here rather than solving the more general problem that AppleEvents did. (Although the APIs are themselves still in many ways emblematic of the tools and methods of the early 90's rather than offering a solid Swift/ObjC way of doing things)

“Main upside is for games and GPS oriented apps which tend to not exist for the Mac. (In part due to the lack of a laptop with cellular) There's other apps that'd be nice to have such as HomeKit configuration apps, health oriented apps (assuming the HealthKit stuff is accessible). So I'm rather surprised by people saying this lacks utility.”

I swear I’m not trying to be crass... but iOS style games, really? Because everyone is dying to play Candy Crush on their Mac? If people want iOS games on a bigger screen, isn’t that what the iPad and ATV are for?

HomeKit, I’ll give you that, I guess since I don’t use it and can’t have an intelligent opinion. But again, aren’t most HomeKit products and apps from huge multimillion dollar companies? Why aren’t we holding their feet to the fire to develop native Mac apps? Surely they have the resources.

And what would you do with these apps on the Mac that wouldn’t be more convenient to just pull your iPhone out of your pocket anyway?

What GPS apps do you have in mind, given that Macs don’t have GPS nor accelerometer chips? I’m curious. I’m still not seeing overwhelming use cases for iOS apps on the Mac... certainly not on the scale that would justify butchering the Mac in the process.

I couldn't agree more with Brent Simmons' "Freedom" essay!

If all our WWDC fears come true, where do us types go? Career change? Retirement? Open/FreeBSD? Linux? They feel like going 15 years in the past. :(

Jason snell: “I’m often opening multiple Finder windows to drag things around and view projects I’m working on. Sometimes they overlap and hide one another, so I open a new Finder window... only to later discover I’ve got five windows viewing the same folder scattered across my desktop.”

1: that’s a feature not a bug. 2: has he heard of tabs in Finder windows?

Seriously this seems like a minor annoyance at best, not a major problem.

I find that iOS not having the ability to drag and drop multiple files to be a hindrance, not a feature that’s better. For example, if I want to attach several types of documents to an email, or even just several photos, it TAKES FOREVER to do what is a 2 second operation on my Mac because on iOS you have to add the files one at a time and navigate to many different parts of the Files app (or Photo app) to add them, versus on Mac where the different types of files can easily exist together wherever the hell I want them to (e.g. a designated project folder).

More...

“The Mac approach of an infinite number of overlapping windows of arbitrary sizes is a classic—but might a different approach to full-screen, split-view, and side-snapped windows be easier, cleaner, and more efficient?”

Sure, maybe for your workflow. And guess what? YOU CAN ALREADY DO THAT. That’s the power of the Mac. Just install the “BetterSnapTool” app.

What a stupid article. The headline says “iOS does many things better” but then he goes on to list only 3 things — which IMO aren’t actually great reasons that iOS is better. BTW I can easily list dozens of things that the Mac does better, which iOS either can’t do, or is so tedious (like adding attachments to emails) that I almost never do it on any of my iOS devices.

If bringing more iOS features to the Mac is so great, why can’t anyone seem to make a coherent and irrefutable statement about how it will make the Mac better? And how many of these wishes are actually completely new concepts from iOS that really ARE better, versus things that iOS has — like Shortcuts — simply because Apple has pulled development resources from the Mac? e.g. has let Automator / AppleScript wither and die.

It seems like Snell just really ought to be using an iPad?

>> I’ve got five windows viewing the same folder scattered across my desktop
> that’s a feature not a bug

This is clearly not a feature. Originally, Macs intentionally prevented this situation. If you opened a folder that was already open, it would force the original folder to appear, or to close (if you expanded a list view folder).

I think the underlying problem here is that the desktop paradigm worked well in 1984, when people had a a few megabytes of storage space at most, and so couldn't have more than a few dozen documents in the Finder at once. Nowadays, the whole fundamental concept of how the Finder works just isn't adequate anymore, and the way Apple has tried to work around this issue has created as many issues as it has solved.

Unfortunately, Apple managed to also screw this up on iOS, despite not having 30 years of UI baggage to contend with, so there's that.

> This is clearly not a feature. Originally, Macs intentionally prevented this situation. If you opened a folder that was already open, it would force the original folder to appear, or to close (if you expanded a list view folder).

I do think it is a feature, sometimes it is useful. But I still miss the spatial orientation in classic Finder, to be able to organize windows visually and trusting they stay where you put them.

How’s it not a feature? Sometimes I want the same folder open in more than one window, more than one set of Finder tabs, more than one Space, etc for various valid reasons. If you don’t need a folder open more than once, then simply don’t do it — it’s not that hard to NOT use a feature, if you don’t like it, right? I don’t see why Snell sees this as a negative, it sounds to me like he’s (surprisingly!) bad at navigating Finder windows and possibly doesn’t even know about the Window menu!!

> If you don’t need a folder open more than once, then simply don’t do it —
> it’s not that hard to NOT use a feature, if you don’t like it, right?

This is precisely why it is not a feature. In literally all cases where I open the same folder twice, I do it against my actual intention. I do not want to open the Downloads folder twice or three times or four, yet for some reason, I always end up with that folder open multiple times. Thus, it *is* hard to not do this even if you want to avoid doing it.

To be perfectly clear, I'm not saying that, as a general concept, the idea of opening the same folder multiple times can't be a feature. I'm saying that in the context of the Mac Finder, it is not a feature, since the Finder was originally not designed to be used in that way, is still largely based on that same design philosophy, but no longer enforces this rule, and thus regularly creates situations where users end up with a lot of extraneous, unwanted, duplicated windows.

> it sounds to me like he’s (surprisingly!) bad at navigating Finder windows

It used to be that Mac users prided themselves on working with the operating system with the single best user interface. I think the shift towards blaming people for just not being smart enough to use Macs properly coincided with the degradation of Apple's user interface design.

The problem here is not that Snell is too dumb to use Macs. The problem is that the Finder is broken.

In fact, much of Mac OS X's user interface is broken. It somehow manages to combine the worst aspects of traditional Mac user interface design with the worst aspects of iOS UI design, and then liberally sprinkles some fresh new terrible ideas on top of everything. As a Mac user, to me, even some of the more popular Linux distros nowadays feel like giving a glass of ice water to somebody in hell.

I don't understand. If I open the same Finder folder again, I want a new window. If I don't want a new window, I just go to the window that's already open. This is natural and works very well.

Also, I like Finder precisely because I have terabytes of storage that I use. Trying to organize tens of large projects on the iPad is beyond my abilities.

> If I don't want a new window, I just go to the window that's already open

How? Genuine question, how do you do that?

You probably can't just click on it, because it's hidden behind other windows.
You could Cmd-tab to the Finder and see some of its windows, but some are probably still hidden or on different spaces.
You can use Exposé, except that this only works if you have less than maybe half a dozen open windows, and still doesn't work with spaces.
You can switch to the Finder and use the Window menu (finally a list of all windows on all spaces!), but now we're talking about a seriously inconvenient UI that also kind of stops working once you go over a small number of windows.
You could right-click on the Finder and find the window you're looking for, but now you're forced to use the mouse and click around, and again, this only works for small numbers of windows.

Worst of all, once you start using tabs, all of this completely breaks down, because inactive tabs don't even show up in the Window menu or the Finder menu! They're just invisible. So if you open a folder in a tab, and then switch to a different tab, that folder's open window has just become invisible.

(By the way, how is it possible that Apple still can't deal with tabs properly? Apple, if you want tabs, make them a first-level UI concept, not something every app has to poorly implement on its own.)

I don't know, I'm probably missing something, but as far as I can tell, all of the ways the Finder offers to retrieve an existing window are just way less convenient than Cmd-tab, Cmd-N, navigate to folder I want. But if that's the most convenient way of opening a folder, even one that's already open, then the natural outcome of this is the Snell situation. So there's something fundamentally broken about the Finder's UI design.

This is one of these (increasingly common) situations where even Windows is so much more convenient than Mac OS.

>Also, I like Finder precisely because I have terabytes of storage that I use. Trying to organize tens of large projects on the iPad is beyond my abilities.

That's like praising a bike for its speed because it's better than a skateboard. Meanwhile, everybody else is driving a Miata, and if a good UX designer gave this a little bit of thought, they could probably build a Ferrari. And that's what I want Apple to do. Stop adding skateboard wheels to my bike because you're now into skateboarding, that's not making the bike better! Build a fricken Ferrari instead!

Sorry, that analogy went a bit off the rails.

Vlad Ghitulescu

>> If I don't want a new window, I just go to the window that's already open

> How? Genuine question, how do you do that?

I'm using Witch exactly for this. And Path Finder for everything Finder.
Hmm... That's saying something about the Finder I guess.

@Lukas "Apple, if you want tabs, make them a first-level UI concept, not something every app has to poorly implement on its own.)"

Actually, we are somehow in the opposite case. If you have a Cocoa document-based app, you got tabs support by default when this feature was introduced in AppKit.

Because, of course, nobody at Apple, at that time, thought this should have been an opt-in feature for developers. And of course, it was quite buggy (and it was super easy to reproduce issues in system applications).

So instead of enabling 3rd party developers to spend time to check whether it worked well in their apps and, if they wanted to support to it, to optimize their app for it and enable ot, they had to speed up to turn this feature off and release a new version.

Luckily, the same stupid idea was not repeated with Dark Mode. Which is a good thing considering all the bugs in Dark Mode.

@Lukas Yes, it’s a mystery why tabs don’t show up in the Window menu. It’s one reason I rarely use them in Finder and why it’s hard to find things in Safari.

@someone I agree that the tabbed window feature was rolled out poorly. It was also incompatible with a popular third-party split view class I was using, which necessitated a lot of code changes.

@Bob — I think the hard part of going somewhere else is... for all of the problems with MacOS and its likely future, it's hard to find a viable substitute for AppKit and the various creature comforts that come with the Mac's window manager. The good parts of these would have to be reimplemented elsewhere.

And yeah, I think Brent's post resonated with me as well. And I don't think it's a 'generational thing'... I'm a good bit younger than Brent, but I've read and understood enough history to know that the intentions of the folks who started the 'personal computer revolution' were for personal computers to be far more useful and significant tools than provided for by iOS.

@Lukas

"Worst of all, once you start using tabs, all of this completely breaks down, because inactive tabs don't even show up in the Window menu or the Finder menu! They're just invisible."

Not on my Mac running Mojave. I currently have 3 Finder windows open and each window has multiple tabs. Some of those tabs are even the same folder (in different windows). They ALL show in the Finder's "Window" menu. Nothing is 'invisible'.

This IS a problem in Safari (it doesn't show tabs at all in the Window menu), but it's NOT the same behavior in the Finder.

@Ben G, Lukas: the spatial mode of one window per folder still exists in Finder. All you have to do is to disable the tool bar ... for all windows. The downside is of course lack of toolbar, sidebar, tabs and preview. But it works perfectly, I'm using it day to day and dread every new release of MacOS when someday it will be the newest thing Apple takes away.

What goes perfectly with opening or re-activating a Finder window is a launcher like Alfred. Spotlight works too, of course, but …

@Tim That’s not the case. Even with the toolbar disabled, you can easily get into situations (e.g. with list or column view) where the same folder is shown in more than one window. Whereas, the classic Finder would prevent this.

"I swear I’m not trying to be crass... but iOS style games, really? Because everyone is dying to play Candy Crush on their Mac? If people want iOS games on a bigger screen, isn’t that what the iPad and ATV are for?"

That's about on par with the answer to poor Mac games saying, "isn't that what a PS4 is for?" Lots of people don't want to carry multiple devices with them. The idea is that a MacBook could be the iPad replacement. Right now there's simply lots of things the MacBook can't. The mere fact you said, "isn't that what the iPad and ATV are for" kind of demonstrates the point, doesn't it? Maybe you don't want to play Grimvalor or the like on your Mac. But clearly Apple wants to make it more possible. Their new gaming service appears to require the games work on the Mac, aTV, and iOS. Marzipan is an obvious way to enable that quickly.

"But again, aren’t most HomeKit products and apps from huge multimillion dollar companies? Why aren’t we holding their feet to the fire to develop native Mac apps? Surely they have the resources."

No. Most of the alternatives to the Home app are from small third party developers. Most of them do many things Apple's app won't. And while it's fine to say the large companies should also have Mac apps guess how likely that is? Heck it's sometimes hard to get large companies to even support the iPad let alone imagining first class Mac support. With Marzipan suddenly it's far more doable. It used to be I had to install hardware using either a PC or at least Parallels. These days you have to do everything on your iPhone.

"And what would you do with these apps on the Mac that wouldn’t be more convenient to just pull your iPhone out of your pocket anyway?

What GPS apps do you have in mind, given that Macs don’t have GPS nor accelerometer chips? I’m curious. I’m still not seeing overwhelming use cases for iOS apps on the Mac... certainly not on the scale that would justify butchering the Mac in the process."

Apple easily could (and long ago should have) added cellular + GPS as an option to Macs. However there are tons of bluetooth GPS units that'll sync to your Mac. The APIs are there so your Mac will start using it with Maps. As for what mapping apps, any honestly. A few have web apps for doing routes or the like. If you're setting up way points for a hike or the like though it's vastly easier with a mouse than with touch. So pretty much every mapping app I have I'd love on my Mac.

I'd add in weather apps too. There are a ton of good ones for iOS and usually the equivalent web page is filled with ads and much more of a pain to use. Few of them have Mac versions. Surprisingly video apps are also typically crappy on the Mac. Whereas nearly every camera vendor has an excellent app for iOS. Health apps are an other major category since again it's much easier doing data entry on a Mac rather than an iPhone. (I know, you're answer will be "get an iPad and an Apple keyboard and carry that with your Mac.")

I still don't understand why some people here have a problem with a folder being open in more than one Finder window at a time... and why the "old" Finder was apparently better to restrict you to one folder = one window. How is this hindering anyone at all? I don't get it. It's a nice feature, especially if I want to have a folder open in one view as a List and another view as Icons, or if I have Finder windows arranged by project (containing tabs of folders relevant to that project) and certain folders need to be open in both sets of views.

> Not on my Mac running Mojave

You're right, it's fixed in Mojave, but only one of my Macs actually runs it (another one is too old, and my work Mac still hasn't been allowed to upgrade), and it's *only* fixed in the Finder. Everything else remains broken.

But still, good on Apple for fixing it.

> How is this hindering anyone at all?

It's not hindering anyone. It's just bad design. It creates a situation where windows accumulate until you just have to do ye olde Cmd-Opt-W and start fresh.

And, again, the basic problem is not that you can open the same folder twice. That, in itself, could be a useful feature in a well-designed file browser. Unfortunately, the Finder isn't a well-designed file browser anymore.

The problem is that the friction for finding an already open folder is much higher than the friction for just opening it again. So naturally, people tend towards just opening folders again, instead of navigating to existing windows. Which then makes using the Finder less pleasant the longer you use it, and progressively makes it harder and harder to find open windows.

Even when I do the same basic task repeatedly (e.g. copy files to my NAS), I somehow end up creating tons of duplicated windows, which is unpleasant and confusing.

> and why the "old" Finder was apparently better to restrict you to one folder = one window

The old Finder wasn't just better at it, it was literally perfect at it. The idea that a folder's open window was representative of that folder and therefore could only exist once was a fundamental design principle. In the old Finder, it wasn't possible to just open a new window, you always opened a window by opening a folder. This meant that the Finder could always resurface an already opened window if you opened a folder that was already open.

The fundamental way the Finder was designed made it inherently impossible to open the same folder twice, and it did so in a way that made this behavior natural and intuitive in the best sense of the word.

To be clear, I don't think that the OS X Finder should work like that. But I do think that the way it works now combines some of the worst aspects of the old Finder and of the NeXT file browser, and adds some new bad ideas on top of everything. It can't quite decide what it wants to be, and there's clearly nobody at Apple who has any kind of coherent vision for where it should go, so it ends up being bad at everything, accumulating random features (and, along with them, weird bugs) without a clear end goal.

"This meant that the Finder could always resurface an already opened window if you opened a folder that was already open."

Sounds like this could easily be added as an option to the current Finder if Apple wanted to do so. Or even as a 3rd-party utility. Maybe it would be as 'simple' as "If the folder window already exists, bring it to the front; otherwise open a new window". Could AppleScript do it? If it's possible to get a list of all open windows / tabs in the Finder, and match them based on the path name of the currently selected folder vs all open folders. Yeah, it would be annoying to have to script such a frequent and necessary action, but it'd be an interesting proof of concept.

"It can't quite decide what it wants to be, and there's clearly nobody at Apple who has any kind of coherent vision for where it should go, so it ends up being bad at everything, accumulating random features (and, along with them, weird bugs) without a clear end goal."

Agreed.

I don't think the Finder is perfect by any means, but when I try to think of anything significant that it's missing I can't think of anything off the top of my head...

> You're right, it's fixed in Mojave, but only one of my Macs actually runs it (another one is too old, and my work Mac still hasn't been allowed to upgrade), and it's *only* fixed in the Finder. Everything else remains broken.

Actually it's fixed in all the apps the use the standard AppKit tabs implementation.

My sweet solution for Finder windows in macOS X is to always have exactly two open, all the time. They are almost the full width of the screen, one above the other and usually in column view. Then inside these two windows I use tabs extensively for various current projects and tasks. So I never open and closes windows, only tabs. This has worked surprisingly well over the years. The main operations in Finder is either finding files/folders (to open/use/delete etc) or moving files/folders from one place to another. So at any time there's almost never a need for more than two windows. (The fact that Finder in macOS X still can't restore these two windows after a restart bothers me a lot and that I miss from the old Finder.)

> If the folder window already exists, bring it to the front; otherwise open a new window

The basic issue is this: let's say I have the Downloads folder open. Now I open a new folder, and click on "Downloads" in the sidebar. At this point, what could the Finder do to avoid having Downloads open twice?

1. It could close the active window, and bring the original Downloads window to the foreground. This would probably be a little bit confusing to users, since it would look to them as if selecting "Downloads" had just randomly moved the active window to a different position and size.
2. It could close the original Downloads window. This would probably be a good option to have, but I think it would have to be optional behavior, since it might not always be what the user intends - maybe I clicked on "Downloads" not because I wanted to open the Downloads folder, but because I wanted to open a subfolder of Downloads? Since by default, everything now happens in the same window, and opening a folder doesn't open a new window, it's not clear what exactly my intention was when I clicked on "Downloads".

But let's say Apple implements one of these options. The problem is that neither of these two options solves the problem completely, because you also have list view and column view. Just browsing through column view might mean that, in that single window, multiple different folders are visible and open at the same time. Does that mean that you want to close other instances of these folders in other open windows? Even if those other open windows are also showing column views? This creates very strange side-effects very quickly.

There is probably no good solution to this problem.

> My sweet solution for Finder windows in macOS X is to always have exactly two open, all the time

I think embracing this kind of interaction design for the Finder might be a good idea. Instead of going back to the old spatial Finder, the path ahead might be to go to something that's closer to a tile-based windowing system, at least for the Finder, where things are self-organizing not because folders can only be open once, but because the arrangement of windows itself is much cleaner. This would probably mean getting rid of things like the desktop, and fully embracing the idea that file management in 2019 is not really like managing a bunch of pieces of paper on a desktop, but is closer to interacting with a database. After all, it would be pretty weird if your podcast app or your music app (or EagleFiler!) used a desktop metaphor, so why do we think it makes sense for the Finder?

I'm pretty sure anything like this would create some kind of violent rebellion against Apple, though ¯\_(ツ)_/¯

"I still don't understand why some people here have a problem with a folder being open in more than one Finder window at a time... and why the "old" Finder was apparently better to restrict you to one folder = one window."

The idea was to utilize the cognitive processes of spatiality. People instinctively associate things with spatial memory. This has a long history. The "Art of Memory" used to memorize staggeringly long texts in the medieval and renaissance era often made use of this as well. Humans simply evolved over time in situations where deep spatial memory was extremely valuable.

What people are forgetting is that by the Sys9 days the spatial Finder was already running aground in problems. The main problems people point out are (1) it doesn't work well with shared folders (2) it doesn't work well with large numbers of files. It really was designed for a primary icon view, although in the later days of the classic Mac most power users quickly switched between icon and list views.

The problem was that for dealing with network shares or large numbers of files (common on Unix) it completely broke down. It worked great for typical users who had maybe a dozen or so oft used directories without a lot stored in them. By Sys8 there were already NeXT inspired file access apps. While the main reason OSX used the NeXT styled browser were it's NeXT roots, Apple (and likely Steve Jobs) tried to merge the two paradigms. It just didn't work for the reasons people have complained about above. It's gotten a bit better yet fundamentally a browser like view and a spatial Finder prioritizing icons are incompatible paradigms. So what you really have on macOS now is a primary browser type view and a few semi-spatial views. Yet windows, while they retain their spatial location most of the time, don't really guarantee it. The best solution is an alternative spatial only Finder. Yet while there have been plenty of Finder alternatives over the years no one has made a spatial oriented one.

> Yet while there have been plenty of Finder alternatives over the years no one has made a spatial oriented one.

This is probably not what people have in mind when they ask for the spatial Finder back, but Raskin is a radically spatial way of organizing files on your Mac:
http://www.raskinformac.com/

It doesn't seem to be under active development anymore, as far as I can tell.

I guess I’m weird but the only time I get a bunch of surplus Finder windows on my Mac it’s because my cat was on the keyboard.

Stay up-to-date by subscribing to the Comments RSS Feed for this post.

Leave a Comment