As YellowDog celebrates its 5th year in business, Simon, Chief Technology Officer, looks back at how YellowDog has navigated some of the key technology trends of the last five years such as cloud, Docker, and Cryptocurrency. He also looks ahead to 2020 and beyond.
Back in January 2015, Gareth started YellowDog. I joined him a few weeks later.
As a CTO for a new technology business, you would normally anticipate a typical six or nine month “ramp up time” to recruit and embed the right development team. But YellowDog was different. We raised money through crowdfunding and promised to have a service available to the market by the end of October that year.
Fortunately, Gareth had an existing relationship with Telesoftas, an enterprise software development house that he had worked with previously; he liked them, trusted them, and knew they could be mobilised quickly. This really accelerated our development schedule.
I can categorise YellowDog’s product development during 2015 into three stages, all of which happened within three days:
We missed our target product launch in October, but only by a few days. Not hitting the target date was a little disappointing, but we had achieved a lot, to come this far in a short space of time, and very soon we had paying customers.
When launching products or services it is common to offer a discount or a limited free trial, but it took almost no time for YellowDog to have repeat customers. We knew that people liked what we had built and they saw value in it. Everyone involved felt proud of the progress made, which really sums up our first year.
With a live service, our next task was to seek further investment.
At the time, it felt as though everyone, including venture capitalists (VCs), were talking about blockchain. Yet, neither Gareth nor I ever felt a need to move into blockchain or cryptocurrency. We had to defend our position against several potential investors across 2015 and 2016 who told us we were wrong. I distinctly remember being told to ‘stop being so old school’.
It took resilience. But we knew we didn’t want to go down the route of an Initial Coin Offering (ICO). There’s nothing wrong with blockchain. It’s a very interesting technology with lots of potential uses. But it wasn’t a fit for us.
My opinion is best summarised by an exchange I had with an author for a cryptocurrency book in 2016. He asked me to write the foreword. I declined with a response along the lines of:
“I don’t believe in using electricity to perform mathematical calculations that don’t advance science or solve a business problem.”
My comment ended up as the foreword of the book.
If early 2015 was all about Crypto, 2016 was all about Docker. They were the “unicorn”, with a billion-dollar valuation. At one point, it seemed everyone wanted a slice of the action. According to several technology analysts at the time, Docker was the future of applications.
Yet, while Docker enabled applications to run anywhere, compute power still needed to be provisioned and orchestrated to run the application. It made sense to some potential investors that we would be the perfect partners. Investors were obsessed. We found ourselves questioned with: “How can you work with Docker? How can we associate you with Docker?” And so on.
While Docker was at the forefront of people’s minds, in the July of 2015, Google launched Kubernetes and partnered with the Linux Foundation to form the Cloud Native Computing Foundation (CNCF).
Watching the landscape change over the past few years makes me glad we didn’t allow ourselves to get pushed into a Docker corner. While YellowDog works with Docker technology like Docker Compose today, Kubernetes has really won the applications battle and K8s (used for the management of containerized applications) will become mainstream for new business applications in the years to come.
Arguably, the analysts and VCs that thought YellowDog was a match for cryptocurrencies or an exclusive deal with Docker were wrong, but there are many technologies they were right about, such as the meteoric rise of cloud computing over the past five years and our fit within the ecosystem.
I believe there have been two key growth stages in the cloud thus far:
Growth Stage 1
Initially, functionality in the cloud was pretty limited, compared with what could be achieved on-premise, but early adopters were looking at the simplest elements of their IT infrastructure to migrate to the cloud.
Growth Stage 2
Now people are beginning to see that instead of just moving individual virtual machines (VMs) to the cloud, they can rebuild applications in a cloud environment. Application platforms and containers make this straightforward for developers. Once applications are cloud native they can begin to make use of high availability services, such as databases and serverless technologies. This can allow previously unrivalled uptime, fault tolerance and resource scaling.
Before thinking about the seismic trend of the last decade and perhaps the decade to come, I wanted to take a moment to reflect on, just a handful, of other wonderful analyst predictions. In no particular order…
“We’ll be 3D printing our own washing machine parts”
Back in 2015, if you’d believed the technology press, by now everyone would be 3D printing parts for domestic appliances when they went wrong. Clearly, that hasn’t happened. Why? People can’t be bothered to do the hard work for the upkeep of these household items. Plus, for the most part, manufacturers don’t want to give away their designs for the public to print.
3D print shops sprung up all over the UK. I remember seeing one in Bayswater where you could come in off the street, dress up, strike a pose and be photographed simultaneously from multiple angles. Hand over £20 and return an hour later to pick up a 3D printed figurine of yourself to go on the mantelpiece.
So that was 3D printing!
“We’re going to replace lithium batteries with new battery chemistry.”
Batteries have improved, but not by that much. Engineers have come up with better chip designs so they use less power to do the same things as before.
“Wireless charging will power our phones”
We were told that we’d stop plugging our phones in by now. Most us in 2020 still plug our phones in. Will it happen? Well, I always ask, “is plugging in your phone enough of a pain?” I don’t think people are too bothered about it. If they’re not really bothered, how much more will they pay, even if wireless charging is available?
“You can fold your laptop and your phone in any shape”
We’re still trying to get flexible screens right and for this tech to go mainstream. Again, is it a big enough pain for people? We’re being pulled in different directions by developers rather than consumers. Five or six years ago laptops were getting much lighter. Then touchscreens became mainstream and they got heavier again!
I think laptop users are most concerned about performance; they want something lightweight, in a convenient size, with a battery that lasts. They don’t really care about touchscreen if it means their machine will be heavier. In the same vein – they only want a folding screen if it helps with weight and battery life.
Our needs haven’t changed much when it comes to technology over the last 5 years. Or in fact, over the last 15. If I go back 15 years and think about my laptop use, I think about using Word, Excel, Outlook, and a few other applications. Today, I use Word, Excel, Outlook, and a few other applications.
What has transformed laptop use is a web browser. That’s it. One application. Simple, but transformative. Initially, a web browser was viewed as distraction – photos, blogs etc. Now, many of us run a business through this single application.
Bill Gates is famously quoted as saying:
“We always overestimate the change that will occur in the next 2 years and underestimate the change that will occur in the next 10. Don’t let yourself be lulled into inaction.”
The concept of High Performance Computing (HPC) is broadening and I think YellowDog is leading the conversation. HPC is traditionally associated with large “supercomputers”, all on-premise, with interconnected nodes. This is changing. If you mention doing High Performance Computing in the cloud, many HPC engineers just laugh. Cloud providers don’t typically have the network links or the machines in close enough proximity – or even in the same data centre.
Instead of defining HPC as a contained supercomputer in one location, we see a wider definition that focuses on an area of “High Performance Computing”. This will take effect in use cases when considerable compute resources are required, but they don’t necessarily need to be in the same data centre, with fibre optic links between nodes as they do with a supercomputer.
YellowDog represents this broadened definition of HPC. It can mean different things to different people and I’m excited to be there to guide others on that journey over the coming years.
And on that note, I’ll get back to Word, Excel, Email, and a few other things.
Please upgrade your browser
You are seeing this because you are using a browser that is not supported. The YellowDog website is built using modern technology and standards. We recommend upgrading your browser with one of the following to properly view our website:
Windows MacPlease note that this is not an exhaustive list of browsers. We also do not intend to recommend a particular manufacturer's browser over another's; only to suggest upgrading to a browser version that is compliant with current standards to give you the best and most secure browsing experience.