The InfoQ Podcast Hosts Take a Look Back at 2019, Discussing Teal, Edge, Quantum Computing, and more

This podcast is part of our 2019/2020 trends overview. The insights come from our editorial team, all of whom are software engineers, who push the barrier of innovation in their professional lives. Read and reflect on their insights to inspire your tech visions and roadmap for 2020.

Show Notes

• 01:40 Wes: Shane, last year, you and Evan Leybourn published a book "No Projects" – what would you write about this year?
• 01:50 Shane: I would want to call it something like: "Teal gets Read: business agility starting to becoming a real thing" - organisations are starting to make a real transition.
• 02:05 Shane: A think that has annoyed me over the last few years is the number of organisational "transformations" that have been nothing to do with transforming the ways of working, just putting a veneer over the old way of doing things, with no change in outcomes.
• 02:25 Shane: We're starting to see (even in the big traditional organisations) a desire to make real change that does have an impact on customer satisfaction, employee engagement and business outcomes.
• 02:45 Daniel: As a shameless plug, a year ago I wrote a book with Abraham Marín-Pérez called Continuous Delivery in Java, and we talked a lot of tech in there, but also principles and practices as well.
• 03:00 Daniel: To Shane's point, there was a lot of questions regarding "Should we use Jenkins?" or "Should we use Java 8 or 11?" - I've noticed at conferences, people are moving to higher questions: business value is the more important thing.
• 03:20 Daniel: As awesome as Jenkins and Kubernetes are, it's more about delivering value to the company.
• 03:30 Daniel: People are now asking how to deliver value quickly with the tools, minimise downtime, test complexity - which made me happier.
• 03:40 Daniel: By focussing on the practices and goals, rather than the specific tools, made me smile.
• 03:50 Shane: Outcomes over outputs.
• 03:55 Wes: Adrian Cockroft tweeted "People can copy Netflix, but they only copy what people see - the results, not the process."
• 04:10 Charles: A more general observation, but perhaps to Daniel's point; the people and process tracks at QCon, which used to be in the smallest rooms, are now in the largest rooms we have in the venues.
• 04:30 Charles: This is indicative of the shift that there is an acknowledgement that often in software people are the hard bit, and managing people and projects are difficult.
• 04:45 Charles: Maybe it's a sign of a growing maturity in the industry that I'm delighted to see.
• 05:00 Shane: I'm going to drop a 'yes, and' here - we're seeing a lot of improvement, but there's still a lot of really bad behaviour.
• 05:15 Shane: Some of the best attended tracks at QCon have been in the ethics track this year, and it was great to see people exploring ethics this year.
• 05:20 Shane: We've known about the AI ethics challenges, but it goes a lot deeper and bigger than that - people have (deservedly) gone to prison for that.
• 05:30 Shane: We've been talking about codes of conduct for software engineering - what does it mean to be an ethical software engineer?
• 05:40 Shane: Is it a profession - are we craftspeople?
• 05:45 Charles: We had the ethics track at QCon London 2017, which to my knowledge was the first cross-discipline conference to do an ethics track.
• 06:05 Charles: After that, Anne Currie, who championed that, spun up a co-ed ethics one-day event on ethics.
• 06:10 Charles: We've since followed that up at other QCons, and published articles and emag on ethics and GDPR.
• 06:25 Charles: That's part of growing up as an industry, because software is so pervasive and affects so many people's lives, we have to take the ethical implications seriously.
• 06:40 Wes: I'm really proud of what we're doing with the ethics tracks and socially conscious software, we're pushing a few things QCon.
• 06:45 Daniel: We're spoilt for choice when it comes to QCons around the world - I go to QCon London, QCon New York, QCon San Francisco: what's everyone's favourite presentation?
• 07:00 Wes: I think the keynotes in 2019 were absolutely incredible: London's Sarah Wells did a keynote on mature microservices and how to operate them.
• 07:20 Wes: It was a run book on how the Financial Times runs and operates their microservices.
• 07:30 Wes: I couldn't keep up writing down and tweeting tips and techniques as she kept laying them down.
• 07:35 Wes: At QCon San Francisco, Pamela Gay did an amazing closing keynote, which was an impassioned plea for citizen scientists to help AI move forward.
• 07:50 Wes: It was an amazing keynote.
• 07:55 Wes: The one that really stands out from all the talks and keynotes had to be Nick Caldwell's QCon New York keynote on "Ignite the Fire: How managers can spark new leaders".
• 08:20 Wes: You can watch it on InfoQ.
• 08:30 Wes: You can learn things and immediately apply them when you get back to your desk.
• 08:40 Daniel: It kept me chuckling at some points, but made me think - a perfect talk.
• 08:50 Charles: It was one of my highlights as well; lots of practical advice, things you could immediately apply, properly deep - it was a superb job.
• 09:05 Shane: For me, it was some of the ethics and socially conscious software - Holistic Ed Tech by Antoine Patton
• 09:10 Shane: As a convicted criminal, and coming out from that, was an amazing deep powerful conversations.
• 09:30 Wes: Alex Qin gave an incredible talk on "How do we heal"
• 09:35 Charles: We've just published that - it's absolutely outstanding.
• 09:45 Charles: It was really raw, emotional and powerful; but it was also really deep - that would be my talk of the year.
• 10:05 Wes: Multiple people came up and said afterwards that there was a standing ovation at the end of the talk.
• 10:35 Shane: There was a talk by Elizabeth Schneider on "Optimizing yourself: neurodiversity in tech"
• 10:45 Shane: I think C4Media/QCon/InfoQ can pat ourselves on the back; we have been championing these things for longer than everyone else has.
• 11:00 Daniel: Bringing your whole self to work is the key thing, I think - being yourself in your workplace is so valuable.
• 11:15 Wes: Having the right and the confidence to be yourself at work - whether you choose to or not; it's up to you - having that comfort level is important at work and conferences.
• 11:30 Wes: Speaking of conferences, Daniel, you're running the QCon London conference this year: how's that looking?
• 11:40 Daniel: I've got an increased appreciation on what plate spinning you were doing in the past.
• 11:45 Daniel: There's so many folks behind the scenes that keep QCon working behind the scenes.
• 12:00 Daniel: We've got a bunch of folks on the program chair and track chairs.
• 12:20 Daniel: We've got a great Kubernetes track, organised by Crystal Hirschorn who is at Conde Nast.
• 12:30 Daniel: Charles is looking after the Chaos track along with Nora Jones.
• 12:50 Daniel: The final one I'll mention (there are many others on the website) is Sid Anand on Leaving the Ivory Tower.
• 13:05 Daniel: Sid ran an excellent track last year at QCon New York.
• 13:35 Wes: It's been a pleasure to see you and David doing an excellent job at QCon London.
• 13:45 Wes: So, Charles, looking back at InfoQ over 2019 - what are some of the big stories that stood out for you?
• 13:50 Charles: We started running trend reports over a year ago, and they found their feet this year.
• 14:00 Charles: We have discussions with the editors at InfoQ, looking at surveys, private polling and traffic patterns, trying to get a sense of where our readers are going.
• 14:20 Charles: It took a bit of time to get the format right - we don't do a lot of opinion content on InfoQ, so it was a bit novel.
• 14:30 Charles: We did one this year about programming languages, excluding Java and JavaScript, and we got a lot of really positive coverage.
• 14:50 Charles: That was the stand-out one; we're looking at some of the things we're seeing in infrastructure: Ballerina, Pollumi, DartLang.
• 15:05 Charles: We're seeing Rust move from innovator to early adopter.
• 15:10 Charles: We're seeing Python evolve as a teaching language, and also the growth of Swift.
• 15:15 Charles: We talked a bit about .Net and .NetCore, which has become popular on InfoQ recently.
• 15:25 Charles: It wasn't that long ago that .Net wasn't hugely trend setting, but .NetCore has.
• 15:40 Wes: I want to jump back to Python - I recently read that the Tiobe index with Python overtaking Java.
• 15:50 Wes: We've been hearing about the demise of Java for years - and I say this as a Java developer - are we finally reaching the demise of Java?
• 15:55 Charles: I've worked in computing as a programmer for more than 20 years, and for the entire of that type, Java has apparently been dying.
• 16:05 Charles: I don't think Java is dying; it might not be getting bigger than it is, but depending on which your benchmark is, Java is either the biggest or second biggest language used today.
• 16:15 Charles: Yes, Python has grown in popularity; it's worth also noting that JavaScript is possibly more widely used than Java is, but not the dominant language that people write systems in.
• 16:35 Charles: Java is the most widely used language in the InfoQ readership, and it's fair to say that Tiobe index is a signal, but not the only one.
• 16:50 Charles: Python and Java are about the same age [Python was created in 1990, Java in 1994].
• 17:00 Charles: Python's sudden rise in popularity has been driven by data science recently.
• 17:05 Charles: If you're a Java developer, you don't need to and look for a new language just yet.
• 17:10 Wes: Python's rise is more a rise of the data science and machine learning explosion that we've seen over the last few years.
• 17:25 Daniel: "Java is dead, long live Java" has been the refrain for the last decade, but it's pervasive.
• 17:30 Daniel: There's clearly going to be maintenance work for ages to come, and the language is evolving with lambdas.
• 17:40 Daniel: There's always incremental work going on; we had a post from Brian Goetz recently on InfoQ what's been going on with types.
• 17:50 Daniel; A lot of hipsters in the Java space that I talk to are looking at Kotlin, not necessarily in production, but they like it.
• 18:00 Daniel: Kotlin is quite easy to write, easy to read, quite expressive in some ways that Java isn't, but I think outside of Android, Kotlin's adoption is still a year off.
• 18:10 Daniel: I'm with Charles; I write Java as my language, but it's good for a few more years.
• 18:30 Wes: There's innovation on the JVM with GraalVM, which we've had at QCon for a few years, and the ahead of time compilation that Graal provides.
• 18:45 Daniel: The podcast that you recorded with the GraalVM folks was excellent, I thought.
• 18:50 Shane: I can't comment a lot in there - last time I wrote code in anger was in C++.
• 18:55 Shane: I discovered a while ago to my horror that I can still read assembler code in a hex dump.
• 19:05 Wes: That code isn't too far off, with web assembly being able to put code in a browser - you could put an application in there like AutoCad or Photoshop.
• 19:30 Wes: Shane, let's switch gears; what is business agility keynote about?
• 19:45 Shane: The keynote is "The edge of agility" - I start with an assertion that 'agile is dead' and pull out InfoQ articles that address that.
• 20:00 Shane: So we look and see that maybe 95% of organisations are using agile in one form or another.
• 20:10 Shane: In our InfoQ trend report, we've pushed vanilla agile to the late majority stage.
• 20:20 Shane: What we're seeing in the early adopter and innovator spaces are organisations getting real about making the genuine shift, building on the platform of strong technical agility to deliver business benefits.
• 20:40 Shane: The state of business agility report indicated what business are getting from this; increased revenue, faster turnaround time, better relationships, greater transparency, customer and employee satisfaction being two of the biggest measurable benefits of adopting these new ways of working and thinking; sustainable success, market success, profitability
• 21:05 Shane: There's also a whole lot of misconceptions and self-deception in the report that are self-reported - they have five categories of 'how well are you are doing'.
• 21:30 Shane: In the C level they reckon they are about a 5.4 out of 10, but the business leaders are saying they're a 4.1, and middle managers are saying they're a 5.1.
• 21:50 Shane: The individual contributors, the people doing the work on the ground, think they're at 3.9.
• 22:00 Shane: The external partners reckon they are at 4.3; so there's a lot of self-deception going on at the organisational manager levels.
• 22:15 Wes: That sounds like an important trend; what are some of the other trends that we're seeing for the upcoming year?
• 22:25 Daniel: I do a lot of work with Kubernetes, and we're seeing a realisation of a lot of the complexity.
• 22:35 Daniel: The complexity often correlates with bad business outcomes; developers are overwhelmed with building good architectures and SRE fighting fires.
• 22:50 Daniel: Kubernetes is an amazing tool, but it is quite complicated.
• 22:55 Daniel: I'm seeing a move towards building tools that supports reducing that complexity; people are picking up custom resource definitions as a standard way to define their infrastructure.
• 23:10 Daniel: Giving that Kubernetes is a standard now with all the public clouds supporting it, they are now homogonizing that as their compute foundation.
• 23:20 Daniel: We're building tools to not only make it easy to build services going into the platforms, but also to deploy routing (both ingress and service-to-service).
• 23:35 Daniel: Using the same workflow; the way you deploy to kubernetes is separate from the way you release; they should be separate, but allow you to use the same workflow.
• 23:45 Daniel: I'm hoping that trend continues where we reduce the workflows and standardise on them.
• 24:10 Charles: I've been tracking quantum computing; IBM published a paper in Nature in May on how to deal with decoherence.
• 24:30 Charles: The issue with decoherence is the fast decay of a wave function has an undesirable side effects generating noise.
• 24:35 Charles: In September, Google announced in a paper in Nature that they had achieved quantum supremacy, which is an overblown term that means that it can outperform what a classical computer can do.
• 24:50 Charles: It's worth saying that the paper has been disputed by IBM and others as well, but it points to the fact that this technology is moving quickly.
• 25:00 Charles: Amazon have announced quantum bracket, which is a way of dealing with different providers of quantum computers from different manufacturers.
• 25:15 Charles: Azure has a quantum provider, and IBM have had theirs on the cloud for some time, but it still blows my mind we can be sitting at home but running actual quantum code on real quantum computers.
• 25:30 Charles: It speaks volumes to this shift to the cloud, and what it means in terms of running costs; they're esoteric, difficult to run and manage, but you can have IBM/Amazon/Google do that for you.
• 25:50 Charles: That combination of ubiquitous cloud computing and connectivity of these quantum machines is worth keeping an eye on.
• 26:05 Wes: The trends that I'm tracking - I agree with Daniel that Kubernetes is the fabric that we're deploying to now.
• 26:20 Wes: With edge computing, the first question is: what edge are we talking about?
• 26:30 Wes: We're talking about edge coming into the cloud provider.
• 26:40 Wes: Between the last mile of the telco provider and the cloud edge, there's an amazing richness of compute functionality there.
• 26:45 Wes: What if we could run a kubernetes cluster at the cell towers instead?
• 26:50 Wes: What if we could take the machine learning models trained up in Python, and have them aggregate the decisions at the cell tower?
• 27:05 Wes: I didn't cover the device edge, the billions of the devices that are consuming services.
• 27:25 Wes: Separately, machine learning is becoming more approachable to build models without needing a PhD.
• 27:40 Shane: Two things I want to add; first thing is that your company is not Spotify.
• 27:45 Shane: There's a whole lot of pushing out there to tribes/squads/guilds to be the latest cargo cult.
• 28:00 Shane: Be inspired by, but do not copy, the Spotify model.
• 28:05 Shane: In 2012 when Hendrik spoke about the Spotify model, he said it was temporary and in flux and changing even then.
• 28:15 Shane: I'm going to blame the big consultancy firms that over-popularised it as a transformational model: it's lipstick on a pig again.
• 28:25 Shane: On a personal level, there's a piece that we reported on InfoQ that I'd like to draw people's attention to: https://www.infoq.com/news/2019/10/traits-agile-nation/
• 28:35 Shane: I'm a New Zealander, this was in response to the terrorist attack in Feb 2019, and the way that our country responded.
• 28:50 Shane: Dr Rashinda Hoda is a muslim woman living in Auckland, New Zealand, and gave a TEDx talk in August, where she looked at the way New Zealand society responded and linked it to the four values of the agile society.
• 29:15 Shane: I'm going to read them out; people and interactions over protocols and rules; community collaboration over closed decision making; policies and actions over speeches and promises, and responding to change over following the status quo.
• 29:40 Wes: What are your new year's resolutions?
• 30:00 Daniel: My personal one is not to travel as much - from a software perspective, encouraging folks to focus on the business outcomes.
• 30:30 Charles: I was going to mention travelling less; might be a theme - the other big thing is focussing on back on the core of InfoQ and improve it; it's a privilege and I want to focus on that.
• 31:15 Shane: What can we do to make more joyous workplaces - Richard Sheredon published a book on the customer values being derived from happy employees.
• 31:35 Wes: I'm sensing a trend about working at a higher level, less on the tools, focussing on the people, processes and journey to delivering software.
• 31:50 Wes: As a software developer, I want to investigate machine learning projects with Python; I'd love to get involved more with WASM and WASI.
• 32:30 Wes: I'm humbled and privileged to work with each one of you: Shane, you inspire me; Charles, you keep me on track; Daniel, you keep pushing deeper and I wish you a happy New Year.