Riccardo Mori

Ten days with the first-generation iPad

May 2, 2018

IPad 1

Spoiler: I’ve been having more fun than expected.

I can’t resist. When certain devices start being called ‘vintage’ or just plain ‘obsolete’ and ‘useless’, something clicks inside me. Little by little, there’s this urge to take a closer look at such devices: Let’s see how useless they really are. In the case of the first-generation iPad (which I’ll henceforth refer to as “iPad 1”), my interest came mainly from a UI investigation perspective. I’m still doing research for a little project on iOS’s user interface, so I’m looking for iOS devices where I can see past versions of iOS directly in action. I already had a third-generation iPod touch running iOS 5.1.1, but I wanted to see it on a bigger screen, and the iPad 1 can’t go past iOS 5.1.1. So this was the main reason. I was also intrigued by the hardware design: like the first iPhone, the first iPad is different from all the models that came after. It’s thicker, with a slightly curved aluminium back. It has no cameras, not on the front, not on the back. I think it looks beautiful and clean:

IPad 2010 Wi Fi

Well, my excellent mate David was so kind and generous as to send me an iPad 1 about ten days ago. I was excited and really looking forward to exploring this eight-year-old model; I wanted to see whether I could get some use out of it. As I often say, I don’t like collecting stuff just to put it on a shelf and show to friends and acquaintances. I try to give a purpose to (almost) every device I own, and this iPad 1 was no different.

A quick refresher before I proceed, so that you can have an idea of my starting point.

The iPad 1 was announced at the end of January 2010 and discontinued in March 2011.
It is the heaviest of the 9.7″ iPads: it weighs 680 grams. The cellular version is even heavier: 730 grams. For comparison, the current 6th-generation iPad weighs 469 grams. The only iPad that’s heavier than the iPad 1 is the 12.9″ iPad Pro, which weighs 713 grams (Wi-Fi only model) and 723 grams (Wi-Fi & cellular), and both variants are still lighter than the cellular version of the iPad 1.
It has a single core Apple A4 CPU. 256 MB of RAM. Wi-Fi 802.11n. Bluetooth 2.1+EDR. Display resolution is 1024×768. Pixel density is 132 ppi. It came in 16 GB, 32 GB, and 64 GB storage capacities. (Mine is a 16 GB, Wi-Fi unit)
Original system software was iOS 3.2. Maximum is iOS 5.1.1.

Very first impressions

When I took the iPad 1 out of the padded envelope and bubble wrap in which it came (from California to Spain in two days — this iPad is fast), and held it in my hands, I was immediately surprised by how well it handled. I expected something heavier and a bit more cumbersome than my third-generation iPad; instead, the chunkier design of the iPad’s chassis makes for a steadier, stronger grip when you hold it. Also, the iPad 1 feels oddly lighter than the iPad 3. On paper, the weight difference between the two is negligible (according to the specs sheet, the iPad 3 is 28 grams lighter). When handling, the iPad 3 is certainly thinner but has the heaviness of a single block of material, if you know what I mean. It feels denser. On the other hand, the iPad 1 doesn’t necessarily feels hollower, but you get the impression that there’s more breathing room inside its chassis. The result is that the iPad 1 feels better to handle than the iPad 3 when you compare them without any protective case.

After taking the iPad 1 out of the packaging material, I pressed the power switch without really expecting it to turn on, but it still had 25% battery charge left, so after a few moments I was staring at a fresh installation of iOS 5.1.1, very similar to the stock iPad image you saw above. And here’s when I noticed another thing. Accustomed as I am to retina displays, I thought the iPad 1’s low screen resolution and pixel density would bother me. Strangely, that didn’t happen. This, I think, is due to the combination of two things: 1) the pre-iOS 7 user interface, and 2) holding the iPad at a normal reading distance. Because sure, if I move much closer I clearly see pixels everywhere and blurred fonts, etc., but when using the iPad normally, both the text and the interface elements are fine; nothing really different than staring at a non-retina external display when using the Mac. For comparison, when I take my wife’s iPad 2 — which has the exact same display as the iPad 1 — and look at the screen, the UI and text under iOS 9 look blurrier. This is a clear example of how unfit both the San Francisco system font and the flat design of iOS 7 and newer versions are for non-retina devices.

Making it useful: a journey of discovery

After charging the iPad, and doing some basic customisation, I started my adventurous journey in search of third-party apps still working under iOS 5. So far, it has been a journey of unexpectedly positive surprises, alternating with “Oh, I really thought this would work” moments. But overall, I really can’t complain.

One feature that’s been undoubtedly useful comes from the App Store itself, providing you with the ability of installing a previous version of the app you want that’s still compatible with the iOS version your device is running. Of course you’re out of luck in at least two cases:

The app you’re trying to install was first introduced under iOS 6 or later versions (duh);
The app developer decided to remove the old version of the app after a major update. In this case the response you get from the App Store app is something along the lines of This application requires iOS 8.0 or later. You must update to iOS 8.0 in order to download and use this application.

I’ve also run into a funny third case, where I could download a previous version of an app, but although it was listed as ‘compatible with iPad’, when I launched it I found myself staring at the not-iPad-optimised, tiny 1x iPhone interface. My guess is that in this case the app was fully optimised for the iPad’s bigger screen sometime after iOS 6 was out. And believe me, you don’t want to use the 2x magnified version of an app optimised for the old 320×480 iPhone display resolution.

In my app hunt, I’ve also had a bit of luck. For even though some apps don’t have their iOS 5‑compatible versions available anymore through the App Store, I did have the old .ipa file stored locally, so I managed to install it anyway. To get into more detail, for those interested: Back in 2011, when I upgraded from my old iPhone 3G to the iPhone 4, I didn’t want to lose the ability to reinstall apps that worked under iOS 4.2.1 in case I had to restore the iPhone 3G. So I cloned the Mobile Applications folder of my main iTunes Library and transferred it to one of my PowerBook G4s. From then on, I used said PowerBook to manage apps and backups for the iPhone 3G and all the other older iOS devices I’ve been acquiring over time, being very careful not to update the apps from iTunes (though I have a full backup of the Mobile Applications folder with the older .ipa files, just in case).

While having its frustrating moments — the trial-and-error “Let’s see if this app can be installed/does work on the iPad 1” approach can be tedious — the journey has been fun overall. So let me just show you what I’ve managed to install on this iPad in the past ten days:

Screen 1

Screen 2

Screen 3

Screen 4

Some comments related to the apps:

1. Every app you saw in the above screenshots works, except for the iMDB app, which doesn’t fetch contents from the iMDB servers.

2. The apps that do work typically retain a 95–100% functionality. What does that mean? It means they work well for the most part, but may have some non-working features due to their age. For example, if they offer Dropbox sync, it may not work anymore because of deprecated APIs (iA Writer Classic is one of these apps, but local document storage and iCloud sync still work, so the app is still largely functional).

3. As mentioned previously, some of the apps you see come from installing the old .ipa file I still own and that I have kept in my archives. Surprisingly, though, the majority are apps I could install via the App Store thanks to the “Download last compatible version” option.

4. Something beautiful happened as I was searching my local app archive for suitable apps. I happened to unearth some discontinued apps, and apps that I had stopped using after a terrible redesign. What I didn’t expect is that they still worked. For example, in the second screen, you can see an app called Posts. This was a very nice WordPress client which at some point stopped working, maybe around the time iOS 7 came out (I honestly don’t remember), and was removed from the App Store.

I had hoped the official WordPress app would improve (especially after acquiring the excellent Poster client) so I could use that to blog from my iPad, but its interface never clicked with me; I’ve always found it better for just reading my blogs’ WordPress statistics and little else. Left without viable alternatives, I basically stopped using the iPad to write and publish posts on my blogs — in this regard, it was reduced to a note-taker. I would write my posts on the Mac using MarsEdit.

Imagine my surprise when, after installing Posts on the iPad 1 and entering the credentials to access my System Folder blog settings, the app started syncing and working as if time hadn’t passed:

Posts main UI

Now, you may not like the interface guidelines for iOS versions prior to 7, or even the general look of older apps before iOS 7 came and flattened everything. I could write a book about how the pre-iOS 7 user interface was more consistent, better conceived, more user-friendly, but I guess in the end it’s all a matter of personal taste, so I won’t argue. But trust me when I say that Posts was really a nice, functional WordPress client. Now that I realised I can still use it on this iPad 1, I may even go back to writing more on the iPad. That I had to go back to move forward in this regard is, well, pretty incredible.

Another favourite app that stopped working under iOS 7 is Planetary, which is one of the most original takes when it comes to designing a music player interface. But since this is iOS 5, the app works perfectly (and its interface isn’t dated at all, by the way):

Planetary

Another pleasant surprise was having Penultimate’s old design back. Penultimate was my absolute favourite handwriting note-taking + sketching app back then (read this review on AppStorm from 2011 to get a better idea), but then Evernote acquired it, and later launched a rethought version which made the app lose a lot of its usefulness and charm. Again, you may disagree with me, and you may think that all the old, skeuomorphic apps deserve good riddance. I actually think that there are many, many examples of skeuomorphism done right that make the overall experience nicer and more fun. The old design for Penultimate is one of these examples:

Penultimate overview

Penultimate welcome notebook

5. As you can see, I’ve ended up with a good mix of still useful apps, which manage to cover a surprisingly wide range of tasks:

Cloud services — The Dropbox, Box, and Simplenote apps still work and sync correctly. Let’s hope it lasts. ClouDrop is a CloudApp third-party client, and it still works too.
Music apps — The official Spotify client still works, and I even like its UI better than the current version (more on this later). Planetary and Groove are two very nice music players. Groove is one of the rare apps with a good pre-iOS 7 interface, and a good post-iOS 7 interface. But even if you stick with Apple’s own Music app, it’s simpler, cleaner, more pleasant to use than the current iteration. And for making music, while I was disappointed that I couldn’t download an older, suitable version of Apple’s GarageBand, I was pleased to find out that older versions of Novation Launchkey and Propellerhead’s Figure do indeed work.
Podcast apps — So far, I haven’t made many attempts at finding a third-party app that still works under iOS 5. Apple’s own Podcasts does the job, and if you, like me, loved the tape reel interface, it’s back in all its glory. Finding a good app to listen to Internet radio stations has proven to be harder than anticipated. So far, the old Tuner app seems to work, but a lot of streams return errors.
Video-related apps — The official YouTube client still works. There is a reminder that the app is obsolete and ‘may stop working in the near future’, but for now it works. Same for VLC. (The YouTube app Apple used to include in the first iterations of iOS is now useless, instead).
Writing and Productivity — Let’s see, there are iA Writer Classic, TaskPaper (for notes, documents, tasks, etc.), WriteRoom (from the same developer of TaskPaper — I haven’t installed it yet, but it should work), Documents by Readdle (still one of my favourite document viewers), Phraseology and Terminology (a writing app and a dictionary, both from Agile Tortoise, both very nice apps with a clear, friendly interface); the already mentioned Posts and Penultimate; and Dragon Dictation, which works surprisingly well and understands me better than Siri.
RSS feed readers and related apps — Newsify and Slow Feeds (which is now called Web Subscriber) both work great, and now that Feedly has fixed an issue with the authentication page, a lot of older apps with Feedly integration have regained functionality. Instapaper, too, works without problems.
News — Flipboard works, and while the old interface isn’t that much different from the current one, I was surprised at how responsive it is, especially when compared with Flipboard on my iPad 3 under iOS 9. The old AP Mobile app from the Associated Press works as well. There are probably other apps in this category out there that still work, but I haven’t searched much, admittedly.
Weather — Finding working apps that also look decent was unexpectedly difficult. Usual suspects like AccuWeather or The Weather Channel or Weather Underground — apps I knew have been around for a long time — all required more recent iOS versions to work. The three apps you see in the screenshots above — Weather 2x, Weather Motion HD, and Weather HD Classic (both display as “Weather HD”) — are pretty much all I could find after a bit of searching. Weather HD Classic is my favourite of the three, and it also seems to be the most accurate.
Web browsing — Safari is decent, but it’s old. Sometimes it doesn’t handle modern, bloated websites very well. Other two working alternatives are Chrome and Opera Coast. Chrome seems to be the best and most stable option.
Maps — The original Google Maps that came built in iOS 5 still works fine, but doesn’t display transit information (at least in my area): you only get walking and driving directions when you want to investigate how to reach some destination. Maps+ has nicer touches, but offers little more than Google Maps, feature-wise, and doesn’t have transit information either. I’m still searching for a good solution here, but these two apps are a decent starting point to find your way around (no pun intended).
Utilities — For communications, it’s great to see that Scotty and Panic’s Prompt still work. Scotty can be used to wirelessly transfer files between iOS and Mac devices, and between iOS devices (if you’re using an older version of the app; this feature was sadly removed in the last update); it’s simple, fast, and does its job well. Prompt is a SSH client by the good folks at Panic. They have released Prompt 2, so you won’t find the older version on the App Store. I had it stored locally, since I had purchased it a long time ago. Speaking of Panic, I didn’t expect Status Board to work, probably because I mistakenly thought it was a much more recent app. It’s an old iteration of the app, but it appears to be working well. I was also very happy to be able to install an older version of Edovia’s TouchPad, because it’s quite useful to be able to use an iPad as a big Trackpad and as a wireless keyboard/media remote for the Mac if necessary (i.e., when all USB ports are in use, and I can’t resort to other wireless input devices). WolframAlpha is another good addition, though I had to look at a suitable version for the iPad 1 in my locally-stored apps.
Graphics — For sketching, drawing, retouching, editing, I’ve assembled some essentials which should cover basic needs: Penultimate, Skitch, Snapseed, Adobe Photoshop Express, Paintbook, Brushes, and Pix. I was able to install an older version of Diptic, too, which I typically use to create photo compositions or to show screenshots side by side.
Social Networking — This has been, by far, the most disappointing experience; but given how fast social networks and related services move today, it was also expected. I don’t care about Facebook, so I don’t know whether its client and apps work or not on the iPad 1. As for Twitter, its official client doesn’t work (it doesn’t load anything). Using a browser to check the mobile version of Twitter is doable, but for some strange reason Twitter seems to be serving a crappier-looking Web interface when it detects you’re using an old device (or an old browser).
The only viable option for me was to resort to Twitterrific, which still delivers a great experience, but with some limitations: 280-character tweets are not supported in this older version of the app (and when you read the timeline, tweets longer than 140 characters are truncated and displayed with a link to the original tweet); and it appears you cannot add more than one photo when you tweet. Other third-party clients don’t seem to work (e.g. Echofon, as something goes wrong during authorisation). I wish I had the old Tweetbot 2 to try out, but it’s no longer on the App Store, and I don’t have an old copy of the file in my archives. At the moment, Twitterrific looks like the only option for a good Twitter experience on an iPad 1.
Games — I’m such a casual iOS gamer that I still haven’t explored the options in this category, also perhaps because ‘Games’ is by far the biggest category of iOS apps. So far I’ve only installed Drop7, which was one of the first puzzle games I got addicted to back in the day. I’ve been told that Kometen is a beautiful, inexpensive game that only requires iOS 3.0 or later; and that an older, compatible version of Machinarium can still be installed (hat tip to Nicola and Angelo, respectively). Another classic is Deep Green Chess: it now requires iOS 7, but you can install an older, compatible version via the App Store without issues. I’m sure there are many other good games out there: feel free to send suggestions on Twitter if you still use an old iOS device for gaming.

The iPad 1 and iOS 5 experience in 2018 — further observations

Performance

This is where I was most blown away by the iPad 1. Of course, everything is relative. If my daily driver were any iPad model newer than an iPad Air, how this iPad 1 behaves would be nothing to write home about. But since my main iPad is a third-generation iPad with iOS 9.3.5, when it comes to responsiveness, animations, transitions, even switching from an app to another (yes, iOS 5 had multitasking), the iPad 1 felt noticeably snappier than the iPad 3. And where it wasn’t actually snappier, it definitely felt… less overwhelmed.

On paper, this shouldn’t be possible: the iPad 3 has a dual-core A5X processor, while the iPad 1 has a single-core A4. The iPad 3 has four times the amount of RAM (1 GB vs. 256 MB). Evidently it’s all in the software optimisation, and how well the two iPads manage their memory and resources. Once again I was reminded of how wrong Apple was in allowing the iPad 3 to update to iOS 9. Not to mention the iPad 2 which, in this informal comparison, was noticeably the most sluggish of the three and the worst performer.

Sure, given the small amount of RAM, when switching from an app to another the iPad 1 had to reload more content than the iPad 2 and iPad 3, but somehow it was also faster at doing so; while my iPad 3 was quicker at presenting me an app’s state when returning to it, I also had to wait a bit before the app was actually ready to accept taps and interactions. This creates a frustratingly stuttering pace when multitasking. On the iPad 1, such transition felt more organic and homogeneous, for lack of a better term.

Another performance-related detail worth mentioning: this iPad 1 still has a tremendous battery life. I’m certain that how well a device fares in this department depends on how it has been treated, and I guess this iPad has been treated very well. Still, my iPad 3 on a full charge and in normal daily use lasts one day and the following morning. This iPad 1 under the same conditions lasts two full days and then some. One thing I’ve noticed is how well the battery behaves while the iPad is in standby. One day I left it on the table with the battery at 47%. When I picked it up again a few hours later, the battery indicator was at 46%. My iPad 3 loses more percent points of battery life while idling.

Interface considerations

The deep, system-wide UI redesign that came with iOS 7 in 2013 gave a lighter feel to the whole interface. Lots of white space, a thinner variant of the Helvetica system font, and a flatter look for all graphic elements. I remember how polarised pundits and users were at the time. Some hated the iOS 7 look right away. Some couldn’t stand to look at iOS 6 and earlier anymore. I instantly loved the interface refresh, but I also recognised its immaturity in a few places where its usability left a bit to be desired when compared to the iOS 6 æsthetics. In short, I loved iOS 7 at once; but I still loved iOS 6 and earlier.

Again, when it comes to visuals it’s ultimately a matter of personal taste, but one thing iOS’s user interface possessed before iOS 7’s flattening treatment was consistence and more robust, coherent, stricter interface guidelines. Guidelines that were followed by third-party developers more closely, and the result was that under iOS 6 and earlier versions, third-party apps presented a user interface that was cleaner, more predictable, easier to navigate than what came afterwards, update after update. After iOS’s UI got flatter, when it came to designing apps, things got out of hand, in an ‘anything goes’ fashion.

There are apps today with poor discoverability, ambiguous controls, UI elements whose state or function isn’t immediately clear — i.e. you cannot tell whether they’re tappable or not simply by looking at them; whereas before iOS 7, a button looked like a button right away, and you didn’t have to explore an app by tapping blindly here and there. Spotify is the first example coming to mind: its early iOS and Mac clients were more usable and had a better interface. Look at the Playlists view under iOS 5:

Spotify - Playlists

The visual architecture of panes and the element hierarchy are clear. Every control is immediately recognisable and you know where to tap just by glancing at the UI. The colour scheme is more thoughtful. The various panes and sections are more clearly defined, which helps when navigating the app. The typography — while being more ‘boring’ and adherent to the OS default — is also more efficient and legible.

Look at the Search Results page now:

Spotify — Search results

It takes cues from the pre-iOS 7 Spotlight search interface. The various fields — Artists, Albums, Tracks — are well separated by subtle dividers and background colour changes. The artwork is big enough to be easily recognisable. Overall, yes, the interface has a utilitarian character and isn’t quite imaginative or edgy, but that’s the whole point of an app which basically consists of music discovery: it has to be user-friendly and help users find things. Over time, Spotify’s UI has become more complex — in most areas, unnecessarily so.

I could make other examples, but they wouldn’t be much different. During my trip down Interface Memory Lane these days with the iPad 1, I’ve stumbled on many other cases, and the result was always more or less the same: I found the old version of an app to have a more usable interface and a clearer interface language than its current counterpart. Despite all the pre-iOS 7 skeuomorphism, for many app interfaces of that time design was truly ‘how it works’. Today, more and more often (and it’s not only with iOS) I see examples where design is simply ‘how it looks’; attractive apps, but with ambiguous interface controls, poorly-designed UI architecture, and sometimes even with little to no accessibility, disregarding users with disabilities.

Usefulness

I’ll be honest: I didn’t expect much from an eight-year-old iPad in the usefulness department. I expected it to be severely challenged performance-wise. I expected it to be sluggish, laggy, unable to play modern videos, and so on. The iPad 1 completely caught me by surprise on practically every front. I was able to find many more working apps than I could have imagined. The iPad’s responsiveness was unexpected. I’ve watched a few YouTube videos with the official client, and playback was excellent (the only problem was with subtitles, which couldn’t be displayed). Even the iPad’s speaker sounded louder and clearer than my iPad 3’s.

But the best test to prove whether the iPad 1 could still be useful eight year after its introduction, was also the simplest: I’ve brought the iPad 1 with me while out and about, leaving the trusty iPad 3 at home, and tried to accomplish the same things and carry out the same tasks I do on the iPad 3. It turns out that I didn’t miss the iPad 3 or iOS 9 that much. I could check Twitter, read my feeds, handle emails, browse the Web, check the weather, take notes, edit and annotate a few screenshots (Skitch is another example of ‘How great the UI was before’), upload some documents in Box and Dropbox, listen to music… Sure, I missed having an updated Google Maps for transit information, I missed certain specific apps, and a more complete Twitter experience, among other little things; but if I have to quantify this iPad 1’s usefulness at the end of the day, it still gets a solid 70–80% — which is awesome when you don’t expect more than 40%.

Another thing that struck me has been how very little I missed all the stuff Apple has added to complicate iOS’s user interface. Notifications were so few that I checked Notification Centre maybe twice in ten days (I know, push notifications have stopped working in some older apps, but still); Control Centre was missed more, but I quickly remembered how to reach the controls in the old way, and so its absence didn’t bother me that much. Today View… who cares? The uncomfortable truth is that iOS 5 had a great balance of simplicity and efficiency, a degree of straightforwardness in gestures and navigation, that got lost iteration after iteration. Today, you can do more with iOS 11 on an iPad, but the OS has become more complicated, less immediate, with a more complex (and at times opaque) gesture language, and a more stratified user interface. The tradeoffs of progress, I presume.

But for me, the most striking thing is this: if you look at this graph Apple showed when introducing the iPad Pro, you can clearly see how exponentially the CPU and GPU performance has grown in just a few years:

CPU performance

Yet, what I’ve realised these past days while using the iPad 1 as my main iPad, is that if I had to draw a similar graph to indicate how much its usefulness has grown in the same time interval, the curve wouldn’t be equally dramatic. Really, the range of applications (in the sense of use cases) hasn’t expanded that exponentially, despite the increase in raw processing power and the better specifications of every iPad that has come after the first one these last eight years. This iPad 1 is definitely slower and can’t keep up where processor-intensive operations are required, no doubt, but the biggest blow to its usefulness today comes more from the dropped software and services support. And yet, whenever you find an app or service that still works under iOS 5 the same way it works on a modern iPad, then the iPad 1 suddenly gains usefulness points. Maybe the ‘Megahertz myth’ can be applied to iOS devices too?

In short — this iPad 1 has turned out to be more useful than anticipated, also thanks to the careful app setup I’ve assembled. On a personal level, the fact that I can go back and enjoy certain apps I had been missing (such as Posts, the WordPress client) makes the iPad 1 possibly more useful than my iPad 3 in certain areas.

Conclusion — It’s fun

Despite the sheer length of this piece (my apologies, folks), the message I wanted to convey is that it’s been a fun experience so far, and continues to be fun. It’s fun to rediscover the thrill of the first iPad. It’s fun to get back to that rabbit hole that is hunting for apps, quality apps that could still work on this iPad under iOS 5. In fact, I’ve never browsed the App Store this much in recent times. An even better resource has been Appshopper (even its old iOS app still works fine), and I’m also truly grateful to all the people who contributed to the AppStorm network for keeping online the archives of published articles and reviews. It has been a helpful resource for finding older, but still useful apps — if not for the iPad 1, at least for my other vintage iOS devices running iOS 6 and iOS 7. But most importantly, it’s fun being able to put this eight-year-old iPad to good use.

And last but not least, thanks again to David for gifting me the iPad 1 in the first place!

Speculation and dread for the next transition

April 22, 2018

by Riccardo Mori

Tech Life

Andy Ihnatko’s article for Fast Company, The Fifth Age Of Macintosh: What Happens If Apple Dumps Intel?, is perhaps the best I’ve read on this particular subject. If you don’t know what I’m talking about, the first paragraph of Ihnatko’s piece should be enlightening enough:

Apple uses its own purpose-designed CPUs for its iPhones and iPads, built around the ARM architecture. An article reported by Ian King and Mark Gurman, published by Bloomberg yesterday [2 April], says that the company wants to do the same for Macs and could start shipping computers with the new CPUs instead of Intel chips as soon as 2020.

The core section of the article is called “True Disruption”; Andy’s analysis and observations are lucid and I can hear that ‘sound of inevitability’ Agent Smith mentioned to Neo in The Matrix:

See, disruptive change is an opportunity to consolidate a lot of unrelated pain into a single, horrifying ball and get it all over with at once.

Which is why I can easily picture a plan to build ARM-based Macs that’s part of a bigger plan to change the whole character of the Mac. For years, MacOS has looked decidedly frumpy and unloved, and its few significant improvements (such as TouchID) have been iOS’s hand-me-downs. Maybe that’s because Apple has been sitting on some huge and wonderful ideas that’ll boost the Mac into a higher orbit, and they’ve put off rebuilding MacOS until they had a good reason to tear it all down first.

Or…maybe Apple’s longterm goal isn’t to transition MacOS into the next decade (or, hell, even just our present one). Maybe its goal is to transition Mac users to iOS. Apple’s obsessive love for the iPad has been made clear to me by both my observations of the product line and my conversations with people inside the company (present and former). It doesn’t seem ridiculous that Apple might push the Mac much closer to the character of the iPad, with the iPad Pro picking up enough of the Mac’s character and functions that the whole consumer Mac line would become redundant.

As you may have guessed, it’s this last paragraph what most concerns me. It’s something I’ve been pondering as well since reading that Bloomberg article, but Andy — amazing tech writer as he is — was able to put it all down in such a way as to actually fill this veteran Mac user with dread.

I’ve experienced all the transitions the Mac has gone through. The first one in 1994 — from the Motorola 68K architecture to PowerPC — was perhaps the most painless for me as a user. The Mac-oriented businesses I was collaborating with at the time had upgraded most of their machines, so I was always given rather up-to-date Macs to work on — and in 1994–95 they were all PowerPC Macs (I remember a Power Macintosh 7100 and an 8500). My personal Macs were still 68K machines, instead: a Macintosh Classic and a PowerBook 150. But they were still capable machines for my needs, and I had plenty of 68K software to choose from that was still working well.

I remember that in those years (at least in my corner of the world) the 68K-to-PowerPC transition felt like something that was mandatory for professionals who needed the cutting-edge Macs, the latest and greatest, and most importantly who could really afford the upgrade ($4,000 for a Power Macintosh 8500 in 1995 was already a lot of money, something like $6,500 today, and in Europe Macs were sold at even higher prices). While home users weren’t really pushed to upgrade as it would happen nowadays. Even the ‘affordable’ Macs of the mid-1990s weren’t exactly affordable; peripherals and software applications used to cost much more; a Mac ecosystem was a much more pondered investment in general, and for consumer/prosumer users in particular. I remember purchasing that PowerBook 150 second-hand, and I paid a lot of money for it nonetheless, not to mention the investment for a StyleWriter, a scanner, an external SCSI hard drive and CD-ROM drive. I wasn’t about to drop everything and get a PowerPC Mac. I finally upgraded in late 1998, when PowerPC Macs had been around for four years already, and somehow I didn’t feel I was late to the party. The pace was slower at the time, kids.

The second transition — in 2001–2002 — was an operating system transition, from Mac OS 9 to Mac OS X. This I remember to be more of a bumpy ride for some friends, mostly because certain software they depended on wasn’t yet being ported to Mac OS X and they were forced to either configure their Macs to dual-boot in Mac OS 9 and X, or to open those applications in the Classic environment. But again, this transition took long enough not to be abrupt, despite how some people felt at the time. Mac OS X 10.1 (Puma) was released in September 2001, the last update for Mac OS 9 came out in December 2001, and in 2003 there were still Macs capable of booting into Mac OS 9. On the software side, Mac OS X 10.4 Tiger, introduced in April 2005, was the last release to include the Classic environment; and since Mac OS X 10.5 Leopard was introduced in October 2007, those few users who still relied on very few, very specific tools not updated to Mac OS X, could still run them in Classic as late as 2007, six years after the first usable iteration of the new operating system.

The third transition — in 2006 — was again a CPU architecture transition, from PowerPC to Intel x86; and again, it wasn’t smooth for everyone. I was concerned at first: my then main machine was a somewhat recent PowerBook G4, and I worried that maybe it was already time to save my hard-earned money for yet another Mac with an Intel CPU. Thankfully, by focussing on my needs and being patient, I was able to make that PowerBook last for much more than anticipated. It also helped that in 2007 I was able to acquire a Power Mac G4 Cube for a really low sum: having a second Mac that was powerful enough to take care of secondary tasks, I managed to hold on to my PowerPC setup until 2009, three years after the transition.

From a software standpoint, I still contend that the PowerPC-to-Intel transition was undoubtedly smooth. In the 2006–2009 period, Universal Binary apps were plenty, and when Mac OS X 10.6 Snow Leopard came out in August 2009, it was indeed an Intel-only software, but thanks to Rosetta, Intel Macs running Snow Leopard could still run PowerPC applications. That was particularly useful because it helped extend the life of a few software suites I had regularly purchased, which weren’t exactly cheap: Adobe CS2, Microsoft Office 2004, and Quark XPress 6. PowerPC software was definitely left behind with the release of Mac OS X 10.7 Lion, which didn’t contain the Rosetta interpreter, and Lion was released in July 2011, giving to PowerPC software a five-year grace period. (Even more, since a lot of people preferred to stay on Snow Leopard rather than update to Lion right away).

All these major transitions have common characteristics:

They were all rather user-friendly and customer-friendly.
They weren’t particularly rushed: there was both preparation and confidence on Apple’s part, and they unfolded over a long period of time and at an acceptable pace. Users had to update eventually, but they were given plenty of time to do so.
All these transitions were for the better. PowerPC Macs (except maybe the first generation) displayed a noticeable performance leap compared with 68K machines. Mac OS X had a more robust foundation than Mac OS 9, and especially from Mac OS X 10.3 Panther to 10.6 Snow Leopard, its stability was remarkable. After the initial growing pains, it was clear that Mac OS X would turn out to be much better than Mac OS 9. And Intel Macs were simply much better performers than PowerPC machines almost from the start. Even the 1.5 GHz Intel Core Solo CPU in the Early 2006 Mac mini was twice as fast as the 1.5 GHz PowerPC G4 in the Late 2005 Mac mini. Added benefits of having the same Intel x86 architecture as PCs allowed installing Windows on Macs and configuring dual-boot Mac OS X/Windows Macs via Boot Camp; people who needed to work with both systems were quite happy about that. Oh, and the Mac became a better gaming platform, since more PC titles were more easily ported to work on Macs. (These are just the first things off the top of my head).

This rumoured next transition — from Intel-based Macs to ARM-based Macs — is once again for the better, at least on paper. In his piece, Andy Ihnatko rightly points out that having custom, Apple-designed chips inside Macs could bring the same beneficial effects we’ve all seen in iOS devices: stellar performance, great optimisation of the internal components, even tighter integration between hardware and software. But things have changed in the meantime. For one, today Mac OS evidently isn’t the primary focus of the company. Those past transitions were all done to benefit the Mac; the idea was The Mac shall advance. We’re changing and improving things under the bonnet, but the Mac is still the Mac and its identity won’t change. Instead, this theoretical Intel-to-ARM transition doesn’t feel as such. It feels as there are impending changes to the Mac operating system and platform that are clearly influenced by iOS. This makes me uneasy.

Let me tell you a couple of things straight away: One, there is nothing wrong with the Mac platform, except what Apple has been doing to it in recent years. Two, since Steve Jobs’s passing, my impression is that Apple has been progressively unable to properly handle their two major platforms, Mac OS and iOS. It’s like they can’t keep a balance of resources, development, and attention between Mac OS and iOS. Instead of envisaging a plan where the two platforms progress in parallel, and flourish by making the most of their respective strengths, what I’ve seen is a clear preference for iOS, and a clear progressive neglect of Mac OS. As a Mac user, this frustrates me. Apple’s secrecy, mixed with vague statements from the company’s top executives, is making me more and more anxious about the Mac’s future.

The signals couldn’t be more mixed. Apple seems interested in pleasing again their pro users (the iMac Pro, the new Mac Pro hopefully next year, the Pro Workflow Team, etc.). Tim Cook recently said in an interview that We don’t believe in sort of watering down one for the other. Both [The Mac and iPad] are incredible. […] And if you begin to merge the two… you begin to make trade offs and compromises. So maybe the company would be more efficient at the end of the day. But that’s not what it’s about. […] So this merger thing that some folks are fixated on, I don’t think that’s what users want.

And yet this doesn’t fully reassure me. It’s like watching someone nod vigorously while also saying “No.” Cook’s words tell me that the Mac is not going away. Good. But in what state the Mac is going to stick around — that’s my concern. For both the hardware and the operating system. The ideal scenario for this possible Intel-to-ARM transition could be a new generation of powerful Macs that can run both Mac OS and iOS applications.

Now, this can run both Mac OS and iOS apps is something that can take many forms and directions. If the year were 2011, for example, such new direction would probably be framed as an advantage for the Mac, essentially. Imagine being able to develop and test an iOS app running natively on a Mac instead of inside a simulator. Or imagine playing that iOS game on the Mac directly, taking advantage of the more powerful Mac hardware. Or the benefits of the Mac/iOS integration, where the Mac can run iOS widgets and little utilities as extensions that, for example, can directly interface with another iOS mobile device. If this were 2011, the narrative would probably be, Here’s what iOS can do for the Mac.

But it’s 2018, and seeing where things are going, I worry that the narrative is going to turn out as this: Here’s what iOS will do to the Mac. Because I have the distinct feeling that if iOS is allowed to run on Macs, it will be the end for Mac OS. Maybe not entirely, maybe not overnight, but who’s going to develop Mac OS apps when one can develop a iOS universal app that can work on iPhones, iPads, and Macs? Sure, the Mac version will have to adapt to the non-touch, keyboard + mouse (or trackpad) user interface and interaction, but perhaps it will be less of a headache than having a separate SuperGroovy Text Editor for iOS and SuperGroovy Text Editor for Mac OS.

The only cases where I see Mac OS development survive is a) thanks to the tireless work of long-time, die-hard Mac developers; b) to make very complex Mac professional applications that take advantage of the sheer power of the Mac family of pro machines, and whose tasks and UI design require the non-touch desktop paradigm of a Mac instead of the iPad’s multi-touch interface and relatively small screen real estate. And so maybe Ihnatko is not wrong when he predicts:

I can see a near future in which the only Apple hardware that behaves anything like what we’d recognize as a Mac today are the Mac Pros. These are the pricey workstation-class machines that benefit the most from Intel’s top horsepower; I don’t want to even take a guess as to how long it would take for Apple to build CPUs strong enough to do what an iMac Pro does. So if you use Macs for making movies, apps, visual effects, and machine learning models, rest easy: I think your Macs are safe from the red banner of revolution.

The rest of us are in an itchy spot. The functions of a basic laptop such as the MacBook Air could easily be met by an ARM chip, I reckon. And the latest generation of consumer MacBooks already look like iPads (and involve painful sacrifices, at least if you aren’t a fan of their almost-zero-travel keyboards). Why wouldn’t Apple make them act like iPads, too? In the absence of any visible sign of Apple’s eagerness to make great consumer Macs, it’s a tough question.

Going back to the afore-mentioned three main common characteristics of the previous major transitions — they were customer-friendly, they weren’t particularly rushed, they were all for the better ultimately — while I think the possible Intel-to-ARM transition will be for the better, I sincerely hope it will also be customer-friendly, and that Apple won’t rush things, or approach this transition with the attitude I’ve seen too often in recent times, the “let’s try this and see if it sticks”, make-it-up-as-we-go-along kind of plan. I can stomach another transition, provided it’s done properly.

As a closing consideration, I have to reiterate just how silly and disheartening all the recent treatment of the Mac has become. That it’s inadequate, and has to be phased out, is just empty talk by all-too-eager iOS-only pundits. Obviously, everyone is free to use what’s best for them and speak about their preferences, but things like The Mac is too cumbersome and difficult to use, or that it’s inadequate for modern tasks, or that iOS is a superior platform are very subjective opinions, and not statements of facts. It’s also a bit hypocritical to invite Mac users to be more open-minded towards iOS as a professional tool, while iOS-only proponents aren’t similarly inclined to maybe get to know the Mac better before dismissing it as inadequate and awkward. As I’ve previously, repeatedly said, this iOS vs. Mac OS debate is toxic; Mac OS doesn’t need to be put aside to make iOS shine. It’s not a zero-sum game.^[1]

This insistence that, between iOS and Mac OS, ‘only one shall prevail’ is so misplaced. Both platforms have a specific kind of versatility and a specific set of strengths. If you ask me, the smart position is Better both worlds than the best of both worlds — but both worlds need to be taken care equally. Currently, that doesn’t seem to be happening, with the Mac losing ground, and Apple executives not giving very strong signals that they love the Mac as much as they say they do. This rumoured next transition will be crucial and revealing in this regard. As Ihnatko concludes, either Apple has a big, revolutionary plan in store for the Mac, or it’s preparing for the last season of Mac OS.

Never before has a rumour made me so anxious, though.

1. After reading and appreciating Jason Snell’s feedback on Twitter, I have decided to revise this paragraph, eliminating a couple of particularly snarky comments. I’ve been having a lot of private conversations with too many folks who don’t seem to know Mac OS all that well, yet they’re happy to criticise it to promote their message iOS is the future. Adapt or perish. I wrote this piece after another recent conversation of this kind that got me even more frustrated and irritated, and so the first version of this paragraph contained my embittered reaction to all this nonsense. But the tone was gratuitous and out of line. I had no intention of insulting anyone’s intelligence. ↩︎

Microblogging and fragmentation

April 17, 2018

by Riccardo Mori

Tech Life

I get routinely annoyed at Twitter for something they do to, or don’t do for, their users. During my annoyance periods, I look around in search of ideas or alternatives. This is how I got into App.net in late 2012. This is how I ended up backing Manton Reece’s Micro.blog project on Kickstarter a few months ago.

The cool idea behind Micro.blog is that, while it looks like a Twitter clone on the surface, it actually goes deeper than that, as each account is an independent microblog. Each user is actually publishing very brief blog posts, from websites that are hosted either on the Micro.blog platform (you pay a small monthly fee to have your hosted [username].micro.blog space), or on spaces users already own (you can integrate your microblog with your current blog or website).

This approach has some advantages. For example, you own your space in a way or another, and you’re not dependent on a centralised infrastructure and on the whims of companies like Twitter or Facebook. You can publish on your blog or microblog first, and then broadcast elsewhere. People can follow you on Micro.blog itself by joining the platform, but if they don’t, they still can subscribe to your RSS feed.

This is an intriguing path I urge you to explore, but I must say that so far I’m struggling to make the most of it and optimise my experience. The main reason is fragmentation. Now, it’s not entirely the platform itself at fault here. My particular workflows and habits as someone who publish posts and follows other people who do so, make it difficult to achieve a seamless experience. But I also think that a new type of RSS reader or application could greatly help smoothen things out.

Fragmentation, I was saying. Let’s have a look at my habits first. I’m currently most active on Twitter, but I’m also a member of two other social networks mainly populated by ex App.net folks: pnut.io and 10Centuries. These two other places have slower-moving timelines than Twitter, but still, after adding yet another timeline (Micro.blog’s), I tried monitoring everything by keeping different windows and client apps open on my Mac, only to find it was all a bit overwhelming. I had to do what I do on iOS: check things on each different network by keeping only one app open at a time, maybe two if I’m taking a break from work.

This is not a problem of Micro.blog in itself, of course; I’m just pointing out the fact that if you don’t want to give up other social spaces, adding yet another one can become cumbersome. Sure, you can crosspost to Twitter and Facebook from Micro.blog but, if I understood correctly, you can’t handle Twitter or Facebook interactions on Micro.blog; you still have to go to Twitter or Facebook to do that. In its current state, Micro.blog is an excellent, robust solution for those who are dissatisfied with Twitter, are leaving it or putting it on the back burner, and at the same time are considering the idea of a single space from where to blog, micro-blog, and be social. (Suggested reading: A Guide to Micro.blog For People Who Have A Love/Hate Relationship With Twitter by Jean MacDonald).

Fragmentation, for me, is an issue also when it comes to writing. At first I had thought about integrating Micro.blog into my main website (this one), but I quickly decided against it. Perhaps I’m in the minority here, but I actually like to keep things separated. I like to have a place (this one) where I publish long-form pieces, and another where I broadcast status updates, share photos, links, etc.; for now this other place is mainly Twitter, but even if I got tired, gave up everything and used Micro.blog exclusively, I would still maintain two separate spaces for brief ‘social material’ and for long articles.

Separation and fragmentation tend to form anyway, even if you just stick to blogging and microblogging. The Micro.blog client (or Web interface) is good, but I think it works at its best if you limit yourself to status updates and to the social aspect of Micro.blog. Whereas if you frequently write longer posts and long-form pieces on your main blog, you either:

Keep blog and microblog separated using e.g. MarsEdit for your blog and the Micro.blog client for your tweet-sized updates and for reading other people’s updates.
Keep blog and microblog integrated, and at this point it might be easier to just use MarsEdit to write everything; but to read other people’s updates and interact socially, you’d still need to switch to a different app (Micro.blog app or other client).

Earlier I mentioned that a great feature of microblogging is that you can follow someone’s activity by simply subscribing to their RSS feeds. The problem is that current RSS readers have user interfaces tailored to handle longer posts with titles, mostly. This is because such interfaces all follow the email client paradigm. You have a list of sources on the left, then a list of post/articles you can browse once you’ve selected a source, and then a bigger pane on the right where you usually read the article directly in the RSS reader.

If you follow somebody who mixes short micro-post and longer, traditional blog posts, you’ll easily end up having dozens of unread items, but of course 98% of these will be Twitter-like updates (typically untitled), while the meatier stuff will be the remaining 2%:

And this is for just one source. Imagine following even just 25–30 people. It would become impractical very soon, especially with people who micro-blog as frequently as they tweet. A traditional RSS reader would have an inadequate interface. On the other hand, a microblogging app would need a good overhaul to feature a full-blown set of tools to write long-form posts and not just micro updates.

I still haven’t refined this idea, but for me a more practical Mac RSS reader to follow microbloggers would be an application capable of detecting micro-posts (status updates) and present them in a Twitter-timeline-like manner; while long-form posts and articles would be separated and available to browse and read in the traditional RSS reader view (similar to email clients). Or — even better — an application that looked like a Twitter client at first, but when you click on a post announcing a longer article, a pane would open on the side, where you could read the full article with ease. If you’re only interested in longer articles, a toggle could hide the ‘chaotic’ Twitter-like timeline and present everyone you follow as an RSS source, highlighting only the articles and omitting the status updates.

This is a roughly-sketched mockup of such an application. It’s a sort of Tweetdeck/NetNewsWire hybrid. And yes, the OS X interface looks older because I created this image on my 12-inch PowerBook G4 running Mac OS X 10.5 Leopard.

I believe microblogging has a future, and Manton Reece is doing a great job with Micro.blog as a platform. There are, however, still different levels of friction involved that might make transitions less smooth; not only because the ideal would be for people to just embrace microblogging and leave proprietary, centralised social networks behind — and that’s not going to happen very soon — but also because microblogging as a platform needs specialised applications (readers, clients) that can handle timelines, short updates, and long-form pieces, ideally in one single place, with a homogeneous UI. And by handling I mean both reading and writing. I’m aware that the era of monolithic applications capable of managing different tasks is at its sunset, but the overall (micro)blogging experience would certainly feel less fragmented.

Apple needs polycarbonate again

April 6, 2018

by Riccardo Mori

Tech Life

The last polycarbonate Mac was the 13-inch unibody MacBook released in May 2010 and discontinued in July 2011. It cost $999 but was powerful enough to meet even prosumer needs. Its weaker points were perhaps the graphics card — an NVIDIA GeForce 320 M sharing 256 MB with the main memory — and the lack of a FireWire port. For the rest, it was a capable machine, with a 2.4 GHz Core 2 Duo CPU, and with the RAM that turned out to be expandable to 16 GB. The most notable weakness, though, was ironically the ‘durable polycarbonate unibody enclosure’, which was prone to ugly cracks in the most-stressed spots (such as the hinge area on the back). But that’s not the point.

Let me get to the point by first going back in time. In 1999, when Apple was back on track under Steve Jobs’s direction, all Macs were made of some kind of plastic/polycarbonate material. The separation between consumer and pro machines was indicated by the name scheme, essentially. If the Mac’s name started with an ‘i’, then it was a consumer/prosumer model (iMac, iBook). If it started with ‘Power’, then it was a professional model (Power Mac, PowerBook). This was, of course, reflected in the machine’s specs and prices. The entry-level Mac was the base iMac, with a G3/350 MHz CPU and a starting price of $999. The iBook was more expensive than you might remember: $1,599 for the base 300 MHz model. It was, however, very affordable compared with its professional counterpart: the base 333 MHz PowerBook G3 (“Lombard”) cost $2,499, and the 400 MHz model a jaw-dropping $3,499. Similar prices were found in the pro desktop line — the more high-end Power Mac G3 configurations exceeded $2,000, and the Power Mac G4 had a range of configurations that cost $2,499 to $3,499.

As time passed, and Apple researched new building materials for its computers, the separation between consumer and pro Macs was also conveyed by the materials employed and the enclosure designs. Around 2003–2004, the distinction between polycarbonate (indicating a consumer/prosumer machine) and aluminium (indicating a pro machine) had fully evolved. In this period, the entry-level Macs are the base 12-inch 800 MHz iBook G3 model (April-October 2003), priced at $999; then the base 12-inch 800 MHz iBook G4 model (October 2003 — April 2004), priced at $1,099; and the eMac (only $799 for the base G4/800 MHz model). This is probably the period where you can find the most affordable Macs in recent history that are also not ridiculously crippled when it comes to specifications.

In 2006, after the transition to Intel architecture, the situation isn’t much different: the white polycarbonate MacBook takes the slot that belonged to the iBook, and again the base model is sold at $1,099. The base 17-inch iMac isn’t much more expensive, by the way, and can be had for $1,199. But of course the most affordable Mac keeps being the Mac mini, introduced a year earlier, whose base model is only $599. Its chassis, however, is a mix of polycarbonate and aluminium, so it represents an exception to the “polycarbonate = consumer, aluminium = pro” rule.

But it doesn’t matter, because when I say that Apple needs polycarbonate again, I mean that in a sort of symbolic way — like in the era of the iBook and pre-retina MacBook, Apple should provide a true entry-level Mac made of a different material than aluminium. A Mac that isn’t too terribly crippled specs-wise, but affordable, a little more rugged and — why not? — a little more cheerful. For that, Apple won’t lose its status as a ‘premium’ tech company. If anything, this new ‘low-end’ Mac will signal that the company can smile and not take itself too seriously, just as it happened sometimes under Steve Jobs; and it will also help sell even more Macs.

Yes, I still think that using $300–400 iPads as affordable consumer solutions isn’t always (and everywhere) the right strategy. Where I live, I keep seeing a lot of university students preferring cheap laptops over premium tablets as their portable solution. I know, Apple will never make a $400 MacBook, but a $799–899 model, in my opinion, could really be a hit in the low-end consumer and education slots. Apple’s recent trend in portable Macs has seen prices go up again, and I’ve been hearing old stereotypes come up again (e.g. Apple makes trendy products for rich people).

How to position it, though? The current laptop line would become even messier with a cheap MacBook, an expensive retina MacBook, the comparatively affordable MacBook Air, and the overpriced MacBook Pros. So here’s my slightly radical proposition.

Discontinue the MacBook Air.
Turn the current 12-inch retina MacBook into a 12-inch retina MacBook Pro.
Added bonus: make that 12-inch retina MacBook Pro a bit thicker, a bit more powerful, and give it an additional USB‑C port.
Make the MacBook the most affordable line again, even visually, with this hypothetical new ‘polycarbonate MacBook’; make it in different colours (but bright and vivid ones, like the iMac G3, the first iBooks, the iPod nanos, not the usual and boring space grey, silver and gold); make it reasonably thin (but thicker than the MacBook Pros) and give it at least a USB‑A port; cut costs by using a non-retina display, but perhaps something a bit better than the display in the current MacBook Air; and finally, give it a great battery life, something similar — or even better — than the MacBook Air.

This hypothetical new MacBook could be a sort of spiritual successor of both the old Mid-2010 white unibody MacBook and the MacBook Air line. A convergence of materials, designs, colours, for a product that would hark back to the 1999 era of Apple resurgence in the consumer market; a product that could turn out to be equally successful.

P.S. — Also, a product with a decent keyboard, please.

A few stray observations on voice assistants

March 8, 2018

by Riccardo Mori

Tech Life

To keep the title of this piece short, I have used the somewhat generic term ‘voice assistants’ instead of something like speech recognition and intelligent personal assistant applications. Now that I’ve hopefully made that clear, here are a few thoughts I’ve been having on the subject, in no particular order of importance.

Every time I interact with people who are excited by voice assistants and the underlying technology, they often like to include references to past science fiction series and films, ‘pioneering’ the voice-based human-computer interaction. I’m more and more of the opinion that those series and films have been a bad influence on tech people, that they gave them the wrong idea about what the future of computing should be about. On several occasions, I’ve been baffled by how few of my tech enthusiast interlocutors failed to recognise that in series like Star Trek or Space: 1999, the voice-based interaction with computers is essentially a dramatic device, a way to deliver information to the viewers that is quick and effective. Instead of having boring close-ups of displays where you see queries typed by a Starfleet officer and the computer’s responses, it’s easier to conceive the computer as another character — a very erudite one — who can be queried on the spot and whose response is equally fast. Sometimes this is taken even further by the introduction of an android, a computer with a human shape.

In other cases, presenting a voice-based human-computer interaction is a sci-fi trope to convey a general idea of technological advancement, especially when combined with the absence (or very reduced footprint) of physical tech gadgets/devices. Voice assistants might be ‘the future’, but their current state is little past the mimicking stage of this fictional interaction. That is, we’re aping what we saw in those sci-fi shows, but we’re still at a stage where the form is nice, yet the substance is lacking. There’s little depth beyond the surface. Speech recognition is passable, but reliability is still poor, and the scope of actionable tasks is still limited. It’s like playing one of the early text adventure games, where the parser can’t interpret commands that are more sophisticated than GO EAST, TAKE LAMP, OPEN DOOR, etc. Sure, “we’ll get there” someday, but I can’t shake the feeling that it’s not worth the amount of energies Silicon Valley is pouring into this, and the amount of data we’re feeding to machines to improve Artificial Erudition (I’m still not ready to call it Artificial Intelligence, sorry).
I have this theory about the current limited usefulness of voice assistants, and their relatively slow rate at getting better. I tentatively call this theory ‘the Google Glass fallacy’. It has been pointed out how Google Glass has turned out to be a failed attempt as a general-purpose device aimed at the general public, but a more successful one in limited, specialised applications and environments. I believe voice assistants have started with the wrong foot — as I wrote on Twitter yesterday, I think that if voice assistants had been originally designed having people with disabilities as first and sole target audience (instead of lazy tech dudes), and then gradually extended to everyone else, today they’d be a bit better.

Joe Cieplinski’s 3‑tweet response to that really resonated with me, because it touches on one aspect that inspired my observation in the first place. Here’s his response (emphasis mine):

I think you may be on to something there. Another problem that enthusiasts who think “voice will one day replace your screen” never consider is that those with hearing difficulties would be locked out entirely. I’ve always felt that voice will find its place, but never be the “only” way to interface with computers. Even the folks who wrote Star Trek knew that. It’s also worth noting that there are two different things at play here. Voice recognition, and then artificial intelligence. There’s no reason the two have to be permanently linked. We could just as easily type to Siri or Alexa. Or show it images to interpret.

[Link to tweet 1 | Link to tweet 2 | Link to tweet 3]

I think that in the creation and initial development of these voice assistants, there hasn’t been given enough thought to the ‘assistive’ part, because the design mainly referenced able-bodied people. Simplifying, there’s a big difference when your goal is to develop a tool that makes your life-as-an-able-bodied-person easier (read: spoiled) instead of a tool that makes the life of a disabled person more tolerable. Your able-bodied person’s ‘friction’ is bullshit compared to the real friction of a person with any disability. A useful virtual assistant is one that, first and foremost, addresses a few crucial types of impairments. Design with that in mind, give precedence to solving problems related to the interaction between a person with impairments, develop against those, test against those, then worry about perfectly healthy twenty-somethings who are too inconvenienced to manually select the music they want to play.
When my dad was still around, every now and then I used to tease him into discussing tech-related topics. He was an extraordinarily intellectually curious person, always willing to learn new things, and often approaching questions with great common sense. We only had the chance to talk about voice assistants once, briefly. I was explaining to him the technology and the current capabilities of Siri, Alexa, Cortana, Google Assistant and the like. So, what do you think? – I remember asking.

He fell silent for a bit, then he said: These things can be really useful to people who, for one reason or another, need assistance in their lives. I mean, real assistance: because they’re blind, or can’t move, or are simply too busy to use their hands. For someone like me they’re mostly useless. You know me, I’m quicker if I just use my hands.

– I use Siri to set a timer when I cook. Sometimes for a reminder. Nothing else.

– Simple things, he nodded. – I probably wouldn’t even use it while cooking. I’d just clean my hands, take the phone, set the timer myself.

– Yeah, there are times when I still have to do that anyway. Siri doesn’t always understand what I say.

He slowly shook his head, then said: – Reliability must be put first with these assistants. They ought to understand you at once, and if they don’t, they ought to allow you to correct them as quickly as possible. Otherwise they’re just like that subordinate at the office who is supposed to help you do the work, but he doesn’t understand or misunderstands what you want him to do, and you end up doing more work to fix the misunderstandings.

– Yes, that’s Siri right there.

We laughed, then he observed: Now, if this Siri misunderstands you, you are absolutely able to take matters into your hands, and you just do the thing. You just open the app for the weather forecast, or you set the timer yourself, or you type your internet search. You do that in no time. Now imagine those who truly need this kind of technology in their lives, they are already frustrated enough by their condition. When these assistants fail, it’s even worse. They need them to work. If tech companies want to help these people, they have to work hard at this stuff, or just drop it. Half good doesn’t work here.

— Or, you know, at least recognise your limits and rethink the project. Make something that’s really good at one thing…

– Yes, like something that’s really good at reading things for blind people. You develop one piece, then maybe another company develops something that’s really good at taking your dictation — but really good, something that gets you even if you stutter, or have a lisp… Then maybe one day these two companies collaborate, put the pieces together, and make something very good at more than one thing… and everybody wins.

– Instead, everyone wants to compete, each company trying to offer a finished product that does many things out of the box, and they’re all more or less mediocre.

I miss these chats with my dad.
I wrote this in Siri’s fuzziness and friction — October 2015. Nothing has improved on this front, and this kind of criticism can be extended to other assistants:

And indeed, Siri is the kind of interface where, when everything works, there’s a complete lack of friction. But when it does not work, the amount of friction involved rapidly increases: you have to repeat or rephrase the whole request (sometimes more than once), or take the device and correct the written transcription. Both actions are tedious — and defeat the purpose. It’s like having a flesh-and-bone assistant with hearing problems. Furthermore, whatever you do to correct Siri, you’re never quite sure whether your correcting action will have an impact on similar interactions in the future (it doesn’t seem to have one, from my experience). Then, there’s always what I usually consider the crux of the matter when interacting with Siri: the moment my voice request is misunderstood, it’s typically faster for me to carry out the action myself via the device’s Multi-touch interface, rather than repeat or rephrase the request and hope for the best.

[…]

Siri’s scope is still rather limited. What is the reward for my continued use of this technology despite its immaturity? That sometime in the future it’ll be able to properly write a text message or a reminder? Time is too precious a resource for me to keep trying to have Siri understand simple requests. Not only does the friction in interacting with this particular fuzzy interface have to disappear, but the scope, applications and usefulness of Siri must expand as well — it has to offer enough flexibility and reliability to engage the user. It has to offer more, to provide an advantage over performing the same tasks manually. Otherwise, I think it’s difficult to expect users to invest time and energy in something that still feels non-essential.
While this other article — Siri, wake up — is five years old. Five years. It has aged rather well, which is not a compliment to Siri. In this MacRumors article from September 2017, among other things, there’s this:

Siri is powered by deep learning and AI, technology that has much improved her speech recognition capabilities. According to Wired, Siri’s raw voice recognition capabilities are now able to correctly identify 95 percent of users’ speech, on par with rivals like Alexa and Cortana.

Not my experience at all. I’ve tried Siri, Google Assistant, and Cortana, interacting in English (which is not my first language, yet I believe my pronunciation to be fairly decent) and performing the same requests. Google Assistant and Cortana both performed better and more consistently on this front. Cortana (on Windows Phone 8.1, Windows 10 Mobile, and iOS) even understood me while whispering to it at dead of night. There’s more: even Dragon Dictation under iOS 4.2.1 on my old iPhone 3G was able to correctly understand 99.8% of the text for a short email I was preparing to send.

[…] Joswiak says Apple’s aim from the beginning has been to make Siri a “get‑s**t‑done” machine. “We didn’t engineer this thing to be Trivial Pursuit!” he told Wired. Apple wants Siri to serve as an automated friend that can help people do more.

Maybe it should have been engineered to be (also) Trivial Pursuit. At least it would be good at the Artificial Erudition part of this whole machine learning thing. At this stage, Siri is indeed an automated friend, but a quirky, unhelpful one. Apple is in a difficult position here, because they have decided to integrate this half-baked assistant in too many points of their ecosystem to just pull out now. And the pace of improvement of this automated friend is frustratingly slow.
Speaking of quirky, unhelpful automated friends: apparently, Alexa laughs at you, unprompted. Nick Heer comments:

But why is this possible at all? Is there some sort of hidden maniacal laughter mode? Is that something people would ever want to trigger intentionally, let alone have the device invoke accidentally? Is this a prank? And could you trust Amazon’s virtual assistant to not do anything like this again?

Today it’s an unprompted laugh, tomorrow may be something else, equally unexpected, but perhaps not as innocuous. If I have to renounce a big slice of privacy for these ever-listening devices (I’m not putting any of them in my home, by the way), is it too much to ask for some useful assistance in return?
This has just popped into my head as I was about to publish the article: How nice and useful it would be to have the ability to define ‘shortcuts’ with these assistants, so that common, repetitive tasks can be carried out with shortened queries. Stupid example: you’re often making pizza, or heating a pre-packaged meal in the oven, and you always set a timer to the same amount of time. You could ask the assistant to Define task. The assistant would respond: Name of the task?, and you’d say Pizza time; then the assistant would ask: What do you want me to do for ‘Pizza time’, and you’d reply: Set a timer for 25 minutes. Once confirmed, every time you say Hey [Assistant], let’s do ‘Pizza time’, the assistant carries out the pre-recorded task.

Yes, yes, the idea needs refinement (I just thought of it, bear with me), and it would also need a step further towards sophistication: a more context-aware assistant, the ability to perform a longer exchange than a simple challenge/response, the ability to store tasks in a database on the device and/or the cloud, and a better parser, so that when a particular phrase is invoked, it triggers the assistant to expect a task shortcut. But we’ll have to make them stop laughing at us, first.