Digging out the Pi again.

This morning as I was running early for meetings, I dug out my old Raspberry Pi devices that I have sitting in a drawer. I’m not entirely sure if I have a 2 and a 3 or a 3 and a 4, but I think I’ll boot them later and check.

I was thinking about a Pi 5 because, well, it’s been a whole lifetime since I even booted them. So, with the money burning in my pocket I decided to check them out. Same hardware design (yes, I know it’s slightly different – heavy use of USB-A ports and only one USB-C used for Power.

What about performance. Oh. Dear God.

Yeah, that can wait.

It might be twice as fast as a Pi 4, but honestly I’ve got much more performant hardware lying around. So what would I be using the Pi for?
It’s for the one thing that is kinda missing on the iPad – terminal commands. I can get away with SSH to my server using one of the many ssh programs I’ve got on the iPad but it seems wasteful sending commands to a data centre thousands of kilometres away when I really want to mess with files that are sitting on a disk right beside me.

It can all wait. Maybe the Pi 7….

The Darwinistic Approach (and why everything worthwhile boils down to Natural Selection)

In my youth, I was lucky enough to study science, specifically Genetics. My thesis was about the change in populations due to natural drift – assuming a small rate of random mutation and some selection pressure. Selection pressure is what we would probably describe as “something going wrong” but it can also be a simple filter. Using a Darwinistic approach allows us to evaluate and iterate on a problem. What we are trying to achieve is “whatever can happen, will happen”. That’s the basis of Evolution and has led to a dizzying array of biodiversity in the natural world.

Mathematician Augustus De Morgan wrote on June 23, 1866: “The first experiment already illustrates a truth of the theory, well confirmed by practice, what-ever can happen will happen if we make trials enough.” In later publications “whatever can happen will happen” occasionally is termed “Murphy’s law”, which raises the possibility — if something went wrong—that “Murphy” is simply “De Morgan” misremembered.

The thing to remember about Evolution (and by extension Darwinistic Natural Selection) is that the possibilities generated must come before the selection pressure or nothing survives the selection filter. If the organism doesn’t survive to reproduce, then the line ends. The bank of possibilities must be there already.

This translates into “innovation” easily. An organisation must populate itself with a wide heterogeneity of minds in order to generate the ideas (the fundament of innovation) with sufficient diversity that can survive the selection pressure filter. The ideas should not initially be fettered by the selection pressure criteria (otherwise every problem that looks like a nail results in a solution that resembles a hammer).

After the ideation is complete (though, in truth, ideation and iteration should never stop – just like cell mutation), the selection filter can be applied. Ideas which don’t at first make the grade should be subjected to further iteration before they can be discarded. Only this way can you have a truly Darwinistic natural selection.

Natural selection in this way resulted in modern humans – but also resulted in pilot whales, baboons, golden eagles and bumblebees. Each of them adapted to the niche they occupy. If you apply your selection pressure with the single-minded aim of producing something that looks like a human, you’ll miss out on the entirely practical solutions that resemble bees, monkeys, birds and dolphins. In business terms, this means discarding every solution that doesn’t resemble “we have always done it this way”. What happens to organisms that discard new things? They die out.

We are in an unprecedented era with worldwide biodiversity loss. Organisms are simply unable to adapt to the new way of the world quickly enough. Evolution is simply too slow. Unfortunately for us, we are part of that. Humans are tremendously adaptable – mostly due to our brains and the technologies we develop – but it would be arrogant to assume that we are not under selection pressure right now. Humans continue to evolve but, like our counterparts in nature, we will not evolve quickly enough and, due to the way natural selection works, many of us won’t make it. The Selection Pressures of a changing climate (whether you think it is man-made or not is somewhat irrelevant) are presenting new challenges that we need new ideas to resolve.

Our technology may save us, but we are only fielding ideas that look like old ideas, under the same selection pressures. Great ideas (the baboons, bumblebees, eagles and whales) are discarded because the selection criteria are not fit for purpose. The Legacy pressures from “we have always done it this way” obstruct the effective solutions by discarding innovative ideas.

for example

Statement: We need food to survive
Selection Pressure: Food is not distributed equally
Legacy Pressure: We must grow food for profit not to resolve hunger
Result: A lot of people starve to death but we generate some value for shareholders

Statement: We need food to survive
Selection Pressure: Food is not distributed equally
Legacy Pressure: Only solutions resembling beef steaks will be considered
Result: A lot of people starve to death, but some people get steak

As you can see, Legacy pressures are artificial selection pressures. They limit innovation, they hinder success. They leave us without workable solutions and instead present us with short term distractions that bring us no closer to the result we need (avoiding mass extinction).

We have the opportunity to generate all of the ideas we need before the real selection pressure starts. But we have to rid ourselves of Legacy pressures.

A.I.

First they came for the writers, but I was not a writer.
Then they came for the artists, but I was not an artist.
Then they came for the customer success agents, but I did nothing as I was not a customer success agent.
Then they came for the developers and support, but as that was not my function, I did nothing.
Then they came for the sales team and I did nothing. Maximising efficiency sometimes means sacrifice.
Finally they came for the C-team, but as I had invested everything in the shareholding portfolio, I did nothing.
Finally I am a major shareholder in this company that no-one works at, where no-one can afford our products because we replaced all of the jobs with AI. Why won’t someone do something to help us poor shareholders?

So, Green Software.

Her we are, mid-July in Malta and the temperature is soaring. 34º which ‘feels like’ 39º when you consider the humidity. All over the world weather is breaking new records and none of it in a good way. Record breaking soil temperatures. Record breaking sea temperatures.

My point for today was to highlight the state of green software.

Green Software is software specifically designed (or maybe excellently coded) to reduce resources or to have a positive impact on the environment. You won’t find very many so-called Web3 technologies in the list due, in part, to their huge energy requirements and many of them on Github will be unfinished, non-functional or perhaps one-offs far below what would be considered an MVP.

I am reminded of the Code 4 Good project back in Northern Ireland. I originally kicked it off as Code 4 Pizza. The idea that we would have a hackathon, provide pizza and drinks and focus on an area that really needed fixing for the good of society. Back in 2009, it was a pretty new idea and we managed to hack the local transport monopoly timetables into something usable. Later, I pitched the idea to the Building Change Trust and they ran a series of successful mini-projects where they’d match a programmer with a charity who had a need. Small bits of code to produce web apps and phone apps to make lives easier for charities.

This needs revived. To take on the challenge of fighting environmental collapse during a period when large corporations (fearful of boycotts from climate deniers) are pulling back on their public commitments – and if not held to account – will likely pull back on their real world commitments.

Some of this will be invisible software. Like software that helps a 3D printed solar panel frame orient itself towards the sun. Some of it will be very explicit, like reducing the energy footprint of major apps (Honestly, I’m looking at Firefox and Chrome here). As some of you may remember, I’ve been a fan of ultra-low-power-computing for years. The idea that my Phone has enough computing power for all of my needs has been demonstrated multiple times, the idea that my iPad (with a 20W power supply) was able to outpace the desktop PCs in Film School (with their 400W PSUs) is not lost on me. The concept that the Raspberry Pi, running on a tiny 5V 1A power supply could conceivably replace all of an individuals desktop computing needs is again not lost on me. On the Pi, desktop usage is a little sluggish, maybe not so much that a novice user would notice, but I notice.

But these pieces of hardware use less energy when working hard than a desktop PC uses when idle and I don’t know anyone who’s happy with their current energy bills.

We have the hardware to run highly optimised software – we just need to optimise the software.

Damn, my iPad Pro (2020) is so slow now….

Words…that I’ve never said.

In fact, the number of times I’ve given this iPad Pro the stink-eye and thought “This is so slow” is precisely zero. Now, I do a decent amount of video editing (LumaFusion is a dream) and a little bit of graphics so processor time is something I do stress the machine with. Sometimes.

Over the last year that I’ve been using the iPad Pro 12.9 (2020) model, I’ve been consistently amazed by it. Coupled with a Magic Keyboard case, it’s simply a thing of grace. It makes me wonder why Apple don’t make Macs with the same level of precision and hardware fit as this thing. It’s just dreamy.

And last night, Apple revealed they’re bringing the M1 processor to the iPad Pro.

To illustrate what that means – moving from the A12Z to the M1 (graphics courtesy of Barefeats)

It’s going to be even faster. At everything.

It already is pretty amazing at exporting 4K footage. It’s already cutting through any photo manipulation stuff (using Pixelmator) with ease. But…speed….

A First For Apple

Much has been made of the new M1 processor-equipped Macs from Apple – the fabless MacBook Air, the diminutive Mac mini and the peerless MacBook Pro 13″. They’re currently wowing everyone with their power efficiency (and battery life in the laptops) and their heat characteristics (they’re silent or fanless) and lastly, and probably most relevantly, their speed. They’re toasting the vast majority of x86 (Intel and AMD)-based PCs out there and faster for most people than any previous Mac. And they’re available for new low prices which make it a no-brainier for businesses to invest in one for testing purposes.

The speed isn’t just about how quickly they can render video or process images, it’s all about the useful life of the device as well. My guess is that Apple just extended the useful life of a Mac by a couple of years – and this is a series of machines that routinely is a workhorse for 5-10 years.

I can’t give a better illustration than a render I did last night (I haven’t got my M1 Mac yet) from an 2016-vintage Intel MacBook Pro 13″. It took 2.5 times the length of the footage to render out (1080p). When I did a similar export from my iPad Pro 2020 model, it took a fifth of the length of the footage to render. These devices are a lot faster than the latest models and they’re even faster than the ones they’re intending to replace.

So, it’s great that Apple is currently vending the fastest laptops and micro-desktops on the market.

But that’s not the first I’m talking about.

Apple has always made great hay out of being a system designer. That the tight integration of their hardware and software makes for great computers. We were able to see this with iPhone and iPad – devices which are much lower specifications on paper, but which easily smoke the competition in benchmarks and real-world usage. And this is the thing – it’s only been a few years that Apple has been designing their own chips for the i-Devices. Before that they were like everyone else – using someone else’s chips.

They bought chips from Intel for the Intel Macs.

They bought chips from IBM for the PowerPC.

They bought chips from ARM and Intel for the Newtons.

And they bought the 68K processor series from Motorola for the original “classic” Macs.

This is the first time that Apple can honestly talk about an integrated system when it comes to the Mac. This is hardware designed for software. This is software designed for hardware. This is the dream that Apple has been flogging for nearly five decades. It came to light with the iPhone, it’s come to full truth with the M1 Macs.

Finally!

The M1 Processor is not designed for the high end. It’s fast, but look at that IO. Only one controller (for thunderbolt/USB) which means a limited pipe in. And it’s hardware wired for a maximum of 16 GB of RAM.
There’s room fo something else.

The next thing on the agenda is what happens with the higher-end Macs. Is there a need for an iMac Pro or is it just the “best” in Apple’s traditional Good/Better/Best product matrix? Will there now be room in the product roadmap for a thin and light return to the MacBook form factor? What’s a Mac Pro going to look like? Is there finally room for a Macintosh computer that’s got some expansion (like the Mac Pro) but sits on a desktop? Maybe for one Mac that can take an internal graphics card without having to buy a high end Workstation-class machine that consumes the power of a small mid-American town in winter.

Will this be like the i-Devices – with a M1X and an M1Z on the horizon? Or will Apple go all-in with a whole new no-holds-barred series (I’m calling them P1) which has multiple controllers, a special bridge to expandable RAM and up to 64GB of RAM on the SOC.

Will Apple be chasing the graphics market next? After decades of being poorly served by the graphics companies out there, isn’t it time they served themselves? They’ve been pushing Metal hard for exactly this reason. And Apple is always keen to provide a solution when the market has been providing something substandard and getting away with it.

Hydrogen

I don’t think that hydrogen is going to be a satisfactory alternative. We need to get out of the mindset of choosing finite resources (we only have a finite amount of hydrogen on earth). We end up just moving the problem down the road. Remembering that the water on our planet is finite and it’s a thin skin on the surface of the Earth should be sobering. Remembering that every hydrogen transaction is lossy. And every hydrogen escape is a permanent loss to the planet.

It’s the same with nuclear. Hydrogen is not renewable, nuclear is not clean.

Much of the energy we use is tied up in cars – they’re horrendously inefficient. Not just in the energy conversion (~70% loss) but in the transportation of 100 kgs of human, we have to also transport 900 kgs of car (and that’s for an average one). So, when you think about it, only 3% of the energy we put in the tank is used to transport a human. Electric cars won’t save us from this – they’re a band aid (and they’re even heavier)

(I question the symbolism of Wrightbus in the photo. I question the commitment of the minister from a party of climate-disruption denialism. I question the use of hydrogen for storage – after all, it’s not like hydrogen under pressure has been a problem in the past.)

Take the long view. Build civilisations to last.

Is the 30% AppStore tax fair?

The AppStore is a software aggregator. It’s not the first, it won’t be the last.

About 100 years ago, in the early 2000s, I was buying software for my phone and Palm Vx. I used various software aggregators (Handango being one) and I remember what a rigmarole it was. You really had no idea what you were buying, sometimes a “sticker” set was nothing more than clip art and there was zero recourse if it didn’t work or if you wanted a refund. What’s more, Handango was a revolution in itself as it only took 40% of the cover price. Other aggregators for other mobile operating systems took 40-90%. You’re reading that right. If you got significant success, you ended up paying 90% of cover price to the aggregator. It was like the sliding tax rate but you were really just lining pockets rather than paying for roads and schools.

Then Everything Changed

The AppStore came along and everything changed.

It was hailed as a revolution. A single point of access, installed on every device, taking only 30% and people could trust the vendor – it was the same vendor they’d spent the last few years buying music from. What could be better. Developers would get access to millions of customers overnight, for a small fee which included a code-signing certificate and there were all these dev tools. Sure, it wasn’t perfect by any means and it still isn’t but it’s a damn sight better than it was.

So what’s the fuss?

Some vendors who rely on tight margin subscriptions aren’t happy with the 30%. They’re just pissy with it. It’s hitting on an underlying business model of tight margins. (I’m talking mostly about Spotify and Amazon but it would equally apply to anyone). Their models don’t work if Apple takes 30%. But even then they adapted to while I can’t buy a Kindle book through the Amazon app or the Kindle app, I can just pop onto the web site and it works fine. But it is a little bit of friction.

So, what’s to be done for sharecroppers who are happy to give away an app for free and then use the platform for promotion and distribution without contributing to the baseline? Surely most developers who charge a reasonable amount for their apps would be thinking “No….Amazon shouldn’t be getting a free ride at all!”. But instead it’s really focused the camera on whether Apple’s 30% is reasonable in this day and age.

What percentage would mollify the masses?

There’s really no right answer for this. Some people would only be content with 0%. Thinking that they bring so much to the Apple ecosystem that Apple shouldn’t be making money off the ecosystem. Others might think that 20% or 10% is a good number considering that storage, network costs and compute costs have dropped considerably in the last 12 years. Some folks are happy enough with the 30% but would like the option to offer their own stores, with their own payment systems or the ability to side-load content. There will, like Brexit, be no one single solution that would make everyone happy (and by everyone I include Apple here).

But it does feel like the time is ripe for Apple to rethink the model. To re-evaluate what makes developing on their platform so compelling and to reward those who have made great apps. Maybe it’s dropping the percentage for those who support universal binaries? Or who have switched to Swift 100%. Or those who stick to some other set of rules? Maybe those who do well should get a sliding scale?

Gruber and Gassee used to agree with the Apple Store but have things changed?

Betteridges Law Applies. As Ever

I’m a fan of Betteridges Law. This rule of thumb decrees that the answer to a headline question is always “no” and so I’ll have to apply it here.

Times have changed. Costs have changed. The market has changed. So maybe it’s time to cut developers a bit more slack.

Announcing Digital Circle…

In 2007, industry, academia and government worked together to produce the Northern Ireland Digital Content Strategy – a timely document that paved the way to the Digital Circle collaborative network and several years of rapid growth in our local industry. Timely because it arrived just before a sea change in the market – the advent of the smartphone for everyone and the emergence of the App economy. With a steering group made up of industry professionals and solid support from the IT industry trade body, Digital Circle grew rapidly and was able to respond to changes in the market – from apps, to games, to augmented reality (it’s a big deal now but we had active projects in AR in Northern Ireland in 2011.)

The project officially ended in 2011 but continued to work until 2015 but after the demise of the IT industry trade body, and policy changes in government, practical support waned. Meanwhile, the skills shortages predicted in the strategy proved to be conservative. Salaries for developers have soared, leaving many to outsource their development to other countries – and with that, the opportunity cost in money flowing out of the region and skills and intellectual property being domiciled elsewhere.

With the ending of the Creative Industries Innovation Fund (and fewer interventions like HoneyComb, Creative Marketplace, Games on Film) there was a massive loss in terms of finance. This investment hadn’t always been a runaway success but there were some properties created which, in terms of value, exceeded the entire investment of the fund over its lifetime and all because of an initial £10,000 grant.

And one of the other pillars of Digital Circle (and the strategy) was internationalisation and access to markets. With the impending uncertainty over Brexit, investment has declined and we still don’t know what we have to prepare for. At the time of writing, we haven’t had a government in Northern Ireland in 1000 days. In the face of impending uncertainty and adversity, we feel there’s a need to invest.

Digital Circle is very pleased to announce an ongoing partnership with RAISE Ventures to give the network a home. RAISE brings a physical space, relationships, a programme of events, and a mechanism for private finance to the deal. We are working with RAISE for representation in government programmes (like the Digital Skills strategy, Future Screens NI) as well as championing startups and high potential startups within the region.

See you at the RAISE on Wednesday 16th October.