The Software Slump

Riccardo Mori writing in an essay Raw power alone is not enough

Without innovation in software, all we’re doing with these new powerful machines is essentially the same we were doing 20 years ago on PowerPC G4 and G5 computers, but faster and more conveniently. Granted, it is progress, especially in those fields involving CPU- and GPU-intensive tasks and greatly benefitting by having lots and lots of calculations made in the shortest possible time.

I agree. Where is the modern software to take advantage of the extraordinary amount of power that computers, especially Macs have these days? Just look at this chart:

The MacBook Air multicore performance over time (source Geekbench)

Take a brand new MacBook Air or MacBook Pro out of the box in 2022 and despite it having more than double the raw compute power compared to 5 years ago, the included software is basically the same.

There are doubtless plenty of other examples, let's not even start on how stagnated Windows has become over the last decade.

* Yes, the title of this post is a play on one of my favourite albums by Grandaddy.


Forget the Metaverse, Facebook Wants to Build a Monoverse

Sami Fathi writing for MacRumours:

Yesterday, it was revealed that Meta, more commonly known as Facebook, plans to take a steep 47.5% commission for digital asset purchases made inside the so-called "metaverse."

This reminds me of how in the early 2000s Apple was able to make a deal with the record labels to sell all of their music while taking a healthy cut of their own. Since then they have been an unstoppable force in the digital music industry. The same deal never happened with TV and films. To this day, there is still no single streaming service that lets you watch any TV show and film you can think of, like there is for music. Not wanting to be beholden to Apple, the TV and movie industry learnt from the record companies' past mistakes. I think the lack of any significant new platforms since the smartphone could be a similar reaction from software developers. Why would anyone want to build software for a world run by Facebook, where they take a 50% cut? Short of groundbreaking hardware (unlikely), what does Facebook have to offer developers other than its now dated, millennial oriented, social graph?


Has Time Been Kind to the Apple Watch?

What do BBC News, Instagram, Twitter, Slack, Pokémon GO, eBay and Uber have in common? They all had watchOS apps that have since been discontinued. This week Uber became the latest big name to switch off their Apple Watch app. There are still many useful apps from big and small developers in the watchOS App Store: YouTube Music, Microsoft Outlook, Microsoft PowerPoint and Google Maps are still present (the latter making a comeback recently after being discontinued in 2017). There does however seem to be a trend towards big companies abandoning their ambitions for our wrists, the the remaining apps being far less ambitious.

In the early days of the Apple Watch, Apple would routinely demonstrate it being used to make phone calls, get directions, order food, open garage doors (with a live camera feed) and measure long hikes across the countryside without the need for a phone. The wrist was an existing new frontier for application developers to explore.

I remember in early 2015 - before I had the chance to own an Apple Watch - seeing someone pay for a coffee in Starbucks using their watch. This was before the release of Apple Pay in the UK, and so I concluded this individual must have been using the Starbucks app to make the payment. I thought this was extremely cool, and when I eventually managed to buy my own Apple Watch I installed as many apps as I could on it. Task list managers, calendars, train timetables, news apps, takeaway food services, Reddit clients and even a game of Pong. I even used the aforementioned Uber app during a trip to San Francisco, though remember it being somewhat limited when compared to the iPhone app. This excitement culminated with me upgrading to a cellular model of Apple Watch. I had hopes for this UNIX computer on my wrist to replace my phone in many situations. I could improve my mental health too, by carrying a phone less, and therefore spend less time looking at a screen.

It didn’t happen of course. Despite my cellular Apple Watch Series 4 being plenty fast enough to run 3rd party apps (something that could not be said for the original model) I found most of the applications built for it were either too limited, too buggy, or didn’t exist. The lack of an official WhatsApp client and inability to initiate a conversation from the Watch for example really hampers its usefulness. It was also too easy to run the battery down over the course of a few hours by visiting a location with poor 4G signal. A few times, when simultaneously using the watch’s features by having cellular data, GPS, workout tracking, music the screen flash as a light - all at the same time - the watch simply powered itself down under the stress of it all. I've never known an iPhone to do this.

These days, the Apple Watch fills a less ambitious role in my life. I still use the same Series 4 model from 2018. I no longer pay for a dedicated cellular plan - it just doesn’t seem worth the £5 - £7 a month the mobile operators charge. My usage boils down to just a few tasks now. Telling the time, tracking runs, listening to podcasts, using Siri to control HomeKit lights, and of course, notifications. The only 3rd party app I have installed now is an Authenticator app, which I’m not even sure I need as I seem to be able to confirm logins from the notification and never need to launch the app itself.

The Apple Watch in this respect is a bit of a disappointment. I can say the same for the iPad, which once promised to usher in an era of multitouch computing but ended up moving in the direction of a convertible laptop with much of the software looking and behaving like desktop software that came before. The Apple Watch once promised us a dream of wrist-based ambient computing that would fit so seamlessly into our lives that we would barely notice it, but it would be there when we need it. It has ended up being mostly a fitness tracker and notification device.

What can Apple do to turn this around? It’s difficult to say. Allowing the Apple Watch work without the need for an iPhone would probably be the biggest thing. This would mean the need for a better battery when on cellular. It might also encourage developers to build capable apps for the platform, as it would no longer be a given that the user also has a smartphone in their pocket. This might also mean Apple is more incentivised to support more capable APIs for the watch because suddenly the need is there. As it stands, the Apple Watch always has its older sibling, the iPhone standing over its shoulder as it attempts to make its mark on the world. It needs to be able to stand and compete on its own merits. Would today’s Apple do let this happen? While Apple in 2007 was happy to compete against its own iPod with release of the iPhone, today it seems less bold. The iPad is held back and after 11 years is still not as capable as a Mac despite having very similar hardware. While I would like to see an Apple Watch that truly ushers in a post-smartphone world, sadly I think if it happens it will likely come from another company rather than Apple.


20 Years of the iPod

My original iPod (left), alongside the iPod 5.5 and the Apple Watch, in many ways its successor.

It’s been 10 years since I wrote about it being 10 years since the iPod. That means of course that it’s now been 20 years since Steve Jobs first graced our 1024x768 screens announcing the iPod. I won’t rehash what I wrote 10 years ago, instead I thought it would be interesting delve into what the technology scene was like 20 years ago.

In 2001, I was 16 years old. I had left school in May and started college on September 10th 2001. I will always remember starting college the day before the 9/11 attacks. We had started on the Monday, and our tutor gave us the Tuesday off, telling us to come back on Wednesday. Nobody quite expected the world to have changed so much in the space of little over 24 hours. The gap between leaving school and the attacks was only a few short months. I mention this only to say that no matter the topic, it’s impossible to consider the world in 2001 without remembering that day.

My computer in 2001 was a G3 iMac of the Snow variety. I had bought this in the summer with some money my grandparents had given me. Back then, and still now, buying a new computer was not something I did often. The difference now is that I have many devices that could be considered a computer: a phone, tablet, laptop and even my Apple Watch. At the turn of the century a computer was something you kept on or under a desk.  Macs were not particularly cool, especially within tech circles. The arrival of Mac OS X, with its UNIX underpinnings gave it some street cred amongst geeks like me. I could justify buying one promising I'd be able to learn the Unix command line, while also running user friendly GUI software such as Microsoft Word, and connecting to printers and scanners - a task that was not easy on Linux at the time.

I bought the iMac from John Lewis (there was no Apple Store in the UK at the time) and still vividly remember the excitement of going to collect it, and how unexpectedly heavy the box was (it had a full CRT screen of course).  Initially Mac OS X was far too slow to actually use, and so my dreams of using the iMac to learn the Unix had to be put on hold. Around this time, my phone would have been a Motorola T191. A basic phone that had only one thing going for: it looked cool. To really satisfy my inner geek, I had managed to save up to buy a Palm m100. In 2001, if you wanted a PDA, you wanted a Palm.  For a cool one-hundred pounds, or £170 in today’s money, I had a device that could store phone numbers, take notes and play two-play pong over infrared. It supported email, but only when synchronised over from the Mac. There was no web browser to be seen, and certainly no music. The only audio the Palm m100 was capable of was a Casio watch style “beep”. Instead, my portable music needs were taken care of thanks to a £50 Panasonic portable CD player. It was capable of playing not only audio CDs, but also MP3 CDs. These were CDs stuffed with MP3 files. This meant you could fit more than the usual 10 or so songs that would fit onto a CD. It was portable, but with caveats. It’s bulk meant unless I was wearing a hoodie or fleece with large pockets, it had nowhere to go. It was also difficult to control the music while walking. Still, for 2001 it was great.

Then the iPod happened. I watched the Steve Jobs keynote on my iMac after reading about it in the news.  There was something about it being Mac exclusive ,and me being a Mac user that made it stand out to me. I was lucky enough to get one for Christmas that year. £300 Christmas presents were certainly not a regular thing for me - I don’t know what my parents were thinking!  Initially, my friends laughed at the white earbuds, it would take years for them to be normalised by Apple’s marketing campaigns. The original iPod was extremely easy to scratch, and so I sought a case to protect it. Unlike today where they are a myriad of case options available for every device under the sun, at the time of the iPod’s first release there was no such thing. I ended up buying a case from a specialist photography shop - it was designed to hold a camera flash.  From then on, my experience of music, along with millions others was that of buying the CD, feeding it into my computer, and then synchronise it with my iPod. The CD then went back on the shelf. A new generation of music lovers was born. To this day, when I think about playing a song I start with the artist, locate the album, and press play. One time, I was browsing CDs in HMV and frustrated at the thought of having to wait until I got home before I could listen to my new CD, I had the killer idea - what if the HMV (a bricks and mortar music retailer for anyone who doesn't know) had some kind of machine that would let you scan the barcode of the CD and plug in your iPod so you could listen to your purchase there and then? Of course this became somewhat of a reality 6 years later when the Apple introduced iPod touch and the iTunes music store. Now of course we have steaming services and music is easier to access than ever. We can now ask Siri (other assistants are available) to play virtually any song on demand. Where next for recorded music consumption? I have no idea, but I certainly look back fondly on the days of the iPod.


The Little Things: Teams vs. iMessage

Both Microsoft Teams and Apple’s iMessage offer the ability to ‘react’ or ‘tap back’ on a message - for those who don’t know, this allows you to mark a message with a quick emoji response instead of sending a full message back.

I’m a heavy user of both apps: Teams while I’m working, and iMessage for keeping in touch with friends and family.

Despite Teams being targeted at business - I find the iMessage implementation so much more well through-through and appropriate.

While they both have a thumbs up - check 👍 - the heart in Teams feels a bit too much like a ‘romantic heart’ and unless Microsoft are condoning workplace romances - probably best avoided lest you end up in trouble with HR - I can’t see what they were thinking here. On the other hand, iMessage manages to remain neutral - a small platonic heart that doesn’t throb is much more appropriate for work.

Teams also lacks a simple thumbs down, a way to convey the word ‘no’. Unbelievably, it instead opts for crying emoji, or even worse - a red-faced anger emoji. There aren’t many times I feel like crying in response to a co-worker’s message, or get so angry I turn red - but often I could use a simple thumbs down. The biggest omission from Team however is the humble question mark. Being able to simply question what someone said - that is ask them to expand on what they meant and clarify - is something that often needs to be communicated with colleagues.

All of this is to say that really spending the time to think through little details like this can make all the difference in how useful the feature is. Apple clearly thought more about it in my view. On this feature at least, Teams gets a big 😢 from me.

Footnote: None of these animated icons are technically emoji - but you know what I mean.


Developer Centric Design

The new version of Safari for iPad is a classic case of 'developer centric design'. That is, designing a product that you as a developer would love to have, but not considering how a less technical user base might approach the app. Take the new sidebar - technically, it's a marvel: I can access tab groups (a new feature), my bookmarks, the reading list and browsing history all from one convenient location. In addition, Private browsing is now just another tab group - how cool is that? I wonder if they inherit the same Objective-C interfaces under the hood? Of course they do - and that's the problem. The interface appears to have been designed around its technical implementation.

While new the new 'do everything' sidebar in Safari might sound great on paper, it just doesn't old up in real world usage. Since the release of iOS 15, I've found myself using the iPad far less because of this. Tasks that should be highly optimised are frustratingly slow to complete. Take opening a bookmark and then closing the sidebar in order to browse using the entire width of the screen. It takes up to 6 steps:

  1. Open the sidebar
  2. Optionally press back if the side bar was in another list (History or Reading List)
  3. Select Bookmarks
  4. Navigate to the bookmark and click it
  5. Navigate back up to the root of the sidebar
  6. Close the sidebar

Contrast this with the original version of Safari:

  1. Tap the bookmarks toolbar button
  2. Navigate to the bookmark and click it

That is 3, maybe 4 additional steps - for something which I imagine is a pretty common action.

While I'm not suggesting we all revert to a version of Safari that came out 10 years ago, I think the spirit of the original design: simplicity and usability are still worthwhile goals in 2021. Even with advanced features such as tab groups and browser extensions, there has to be a better way. A simple fix would be to add a bookmarks button back to the toolbar, and remove tab groups from the sidebar entirely (that functionality can also be accessed from the tab overview screen).

Importantly, Apple need to learn how to think like their users again, and not build products simply for themselves. While I can't claim to represent all Apple users, I'm certainly not the only one who finds the new design messy and confusing.


Glass: Half Empty

Like every other Apple-focused technology enthusiast with spare time on their hands, I’ve been trying out Glass, a new app for sharing photos. Its biggest selling point according to the creators, is that they are in fact selling it. Unlike rivals Instagram and Flickr, Glass is not for “selfies, stories, quick snaps, or influencers” and there is no free tier. Instead, Glass costs £4.49 per month or £25.99 for the year as an introductory offer, going up to around £48.00 per annum after the introductory period is over.

As someone who loves photography, but dislikes the way in which Instagram has turned into a late 90s glossy lifestyle magazine with its over-reliance on filters and airbrushing, Glass seemed like a breath of fresh air. In addition to selling this app instead of selling out their users, Glass also does away with traditional social media metrics: like counts and follower counts. While these seemed innocent and innovative in 2010, it has become apparent that such metrics though meaningless in themselves can be terrible for our mental health, and certainly do no correspond with any real sense of worth or value. The business model of selling the app instead of customer data means that in theory it doesn’t matter how long someone spends in Glass, because they are not optimising to maximise eyeballs on ads. This allows Glass to forego the algorithmic timeline that every other social media platform uses and instead offer a simple chronological timeline and feels more like Twitter circa 2009.

That said, upon signing opening Glass for the first time, I was surprised that there is no functionality available unless you sign up for a subscription. Not even ability to even view photos from other people without posting, which I would have expected. The first 2 weeks of the subscription are free however, and so I gave it a go.

First impressions were good: the app is mostly well designed. There are some odd quirks, the back button is in the bottom left of the screen, instead of the top left as it is in nearly every other iOS app, which even after 2 weeks still throws me. The portrait orientation of the iPhone doesn’t work well with photos that are taken in landscape. Glass attempts to remedy this by zooming in and allowing you to pan the image, but I would have preferred to be able to rotate my phone and browse in landscape instead.

Upon first launch I was invited to follow a selection of accounts. I’ve no idea whether this was a random list, or whether they have an algorithm that attempts to show you “interesting” people based on how many other people follow them or some other metric. After a day or two I’d built up list of about 15 people I was following. After that I uploaded some old photos that I was particularly proud of. It turned out this is exactly what everyone else did too. Nearly all of the photos the app displayed to me had dates on them from months and years before the app had launched. The chronological timeline seemed somewhat broken with photos from all seasons seemingly thrown in. This quirk will perhaps become less pronounced as the influx of new users cedes, but for now think of Glass as a place where people post their “best of” photos, rather than a window into what is happening in the world.

After a few days, boredom started to settle in. There were some nice photographs alright. Plenty in fact. But the web is not short of great photography. I found it difficult to discover anything interesting or above average. I particularly like landscape photography, but I couldn’t find a way to search for it, nor did Glass seem to know this and show me more of it. There are places I frequent and photograph a lot, and I wanted to see other people’s take on them. However there is no way to search by location either.  Instead, it’s up to me to follow the right people. It left me longing for an algorithm. Yes, an evil algorithm! The thing is, algorithms are not evil. It’s how they are used that determines whether they are good or bad. If an algorithm is used to optimise for “engagement” for the purposes of selling more advertisements, then this is bad. In the case of Glass, I’d quite like to see something pop up in my feed that is popular or of potential interest, but alas this did not happen.

Another thing I noticed, at least based on the list of people I see in the “follow people” tab, Glass seemed to be somewhat of a monoculture. This makes sense: everyone in the app is rich enough to own an iPhone, probably also owns an expensive dedicated camera, a laptop (probably a MacBook) capable of editing them, and has enough discretionary income to pay a monthly subscription to a social media platform. This is boring. While I’m no fan of Instagram, it does at least seem to attract people from all parts of society. Perhaps this will come with time, if Glass decide to expand beyond its initial user base of photography enthusiasts. For now, where Instagram is the economy section on a plane, Glass aspires to be the first class area with none of the undesirables. This may not be a conscious decision, but it’s a direct consequence of being iOS only and charging a monthly fee.

So is Glass for me? While I really wanted to like Glass, and I have genuinely found some great photographs on it, the staleness and sameness of the content leaves me wanting more for my £4.49/month. I'll be checking back in a few months, hopefully by then this glass will be brimming.


Google’s Native Advertising Problem

I've noticed that Google's search results are full of a lot more spam recently. When I scroll past the page of ads that Google itself places at the top of the results page and jump straight to the so called 'organic' search results, I find more than more that these links turn out to be nothing more than thinly veiled advertisements themselves.

In one example, after finding out my new router supports connecting my entire network to a VPN, I searched Google for the model name and the word 'VPN". The second result was an article that was quick point out that my router doesn't support Express VPN, but I should probably just use it anyway. There was of course a big fat button linking to Express VPN. Using the excellent wheregoes.com I was quickly able to verify that yes, this was a referral link. This article exists for no other reason than to make money from confused users.

Minutes later I searched for how to synchronise to Macs without using iCloud. The 2nd result was this page by MacPaw.

The article has runs through some rudimentary waffle about how cloud synchronisation works, and then about halfway down the pages suddenly breaks into an advert for their product "CleanMyMac X". An app that is completely unrelated to file synchronisation.

While I've nothing against so called "inbound marketing" articles, they can often have genuine value (Digital Ocean's vast array of tutorials for configuring Linux servers are some of the best on the web), it's clear that Google has a big problem on its hand in distinguishing the cruft from the genuinely useful content.


How to Remove Clickbait Headlines From Apple News

Apple News can be a nice way to catch up on the day’s events, however it can also be a source of clickbait headlines and celebrity gossip that I couldn’t care less about. I’m sure it’s the inverse for others too, perhaps you want to read celebrity gossip but keep getting exaggerated political controversies.

Thankfully, there is an option hidden away in the iOS Settings app called “Restrict Stories in Today”. Turning it on will cause News to limit articles to those from publishers you actively follow.

I definitely recommend following a wide variety of outlets to avoid a filter bubble, but I’ve found this option to be a great quality filter so far.


Fancier MacBooks Are on the Way, but You Probably Don’t Need One

Rumours are abundant, and so it must be true that in the coming months Apple will release updated versions its high-end computers. But for the majority of users, even us “professionals”, the current crop of M1 Macs will probably do just fine.

I've been using an M1 MacBook Pro as my own personal laptop now for 3 and a half months and I have to admit, the hype is real. It's a beast. Day to day it's exceedingly fast at nearly everything I throw at it. Anything you could possibly consider “day to day” use doesn’t even warrant a mention - it’s blindingly fast. It breezed through more advanced photo editing, compiling apps using Xcode, editing video in the Photos app, and pretty much every other task I've thrown at it. Contrary to many reports however is the fan. It does get extremely loud when performing intense tasks. In my case, when transcoding video that wasn't natively supported by the M1's instruction set.

In comparison to my other laptop, a work-provided 15" Dell XPS 9500 running Intel's 2020 i7 which has a bigger screen and costs £400 more but is far is hotter, slower and less reliable (frequently overheating while charging to a point it shuts itself off, and unable to stay in sleep mode without draining the battery completely over night) the latest MacBook Air and MacBook Pro are no-brainers for anyone who doesn't need Windows to do their job.

Unless you are playing (or making) AAA games, working with video at a professional level, or training large machine learning models, the current crop of M1 Macs will likely far exceed your needs. Even Marco Arment can get by with the latest MacBook Air for software development. As with any processor transition, it's not all plain sailing - I struggled to get an older version of TensorFlow working on the M1 Mac because Apple have forked the project for M1 at a later version - but these kind of teething issues will resolve themselves over time. As with anything like this, if you have specific software needs, check compatibility first.

So while the latest, shiniest, fastest laptop might seem temping, the mid-range is where the value is right now.