We earlier wrote a fact sheet summarizing what we believe to be the significant topics in 2021: psychological safety, scaling, platforms and design systems, SaaS, edge cloud, 5G, and data challenges. In this podcast, Kalle Mäkelä and Marc Dillon are doing a so-called "mid-year review" to evaluate whether our crystal ball was properly calibrated.
I have been taught 10 years that Okay, CICD is a push oriented way to deliver value to the operational capabilities, and value foods. But it's actually now a pull based, and I see this is also a trend that everything is going to be, transforming to this pool base, which is this kind of lean way of thinking already has shown us so well that push based is never a good idea. Whatever you want to do, whatever you want to change or manage. It's always better to do pull based systems.
Hello there and welcome to the DevOps Sauna podcast. DevOps started as an initiative to break down the silos between development and operations. But these days, DevOps encompasses everything that makes modern software engineering faster, safer and better. We at Eficode earlier wrote a fact sheet summarizing what we believe to be the significant topics in 2021. In this podcast, Kalle Makela, and Marc Dillon are doing a so called media review to evaluate whether our crystal ball was properly calibrated. Kalle and Marc floor is yours.
I was actually reading to document three documents simultaneously earlier today on this topic, there was the DevOps, the puppet state of DevOps report that just came out recently, there was a report coming from GitHub, which is called Upskilling Enterprise DevOps Skills. And then naturally, I was also reading the one that you folks wrote, So Kalle, and Johan wrote back in time, or Kalle and Johan has been marked as authors for the DevOps trends reports for 2021. Tell us a bit about this process that you experienced before the holiday break, late last year, out of which the DevOps trends report came, how was the process of writing that report?
Sure. The motivation basically, was to lift up the pains and gains from the practice level, what we do in our daily lives in working in DevOps and DevOps kinds of activities. So I feel that the difference between these reports that are by nature more scientific, and they are gathered data points and trying to find out pains from the target groups, but we wanted to bring our own experiences, to as a Christmas gift to our customers, and our new customers. And basically, me, Johan, and Marko Klemetti CTO, we had 20 different items. And then we fought about the items that we want to actually include because we couldn't include all of them.
And then we started to work on them, so we had to think about items that were very close to our hearts, basically, and then continue from those. And it was this kind of tournament, a document that all of the greatest and most important ideas from our free point of view. So it's a smaller group that wants to elaborate new ideas that people should think about in their companies and in their organizations, but then we started to do the content, and then because nobody from the three of us couldn't draw anything. So we wanted, of course, to have a little bit more professional way to visualize the things. So I don't know if I forgot something. But a lot.
But the beautiful thing is, we don't know if you forgot, because neither of us, but you said that not one of three of you could draw. And Marc remind, do I remember correctly, Marc that when you are in the meetings do you draw in your notepad?
And I do it still sometimes in online meetings, and of course, the there's a lot of online tools that have made the collaboration of drawing online a lot easier. And just as we have a conversation, writing down a few keywords that any have say as we go along, it gives us all something that we can point to, and come back to, and work around, and figure out "is this the thing that we're talking about?" What are the terms that kind of come?
Actually now remembered what I forgot. So all of these items were we had this one fact that Okay, we have a fact that was some somebody has proven that this is a number that everybody agrees on. And then we want to elaborate on a critical way on that fact, and that item. So and we wanted to make it very visual, so Lauri you said, they can be seldom messy, where you have all the data points and everything. And then you have some kind of abstract in the end or summary, and obstruct in the beginning, of course, so we wanted to make it visual that everybody remembers it.
And the DevOps trends report is a mash up of a white paper underneath infographics. So I like that the way how you have done it, you cite the sources, you have the graphics, and then you have the statement that makes you think. I'd be curious to know what those 13 and where you did not include because I think in my calculation, there were seven: psychological safety, scaling for community building, internal platforms, SAS consumption, edge cloud, 5g, and focus and data choice, yes, seven of them, well, we hope you'll be able to drop some of those 13 into you left out during this podcast,
I have to take the document.
No, we will trust your memory. I have organized these one by one. So I was hoping that we will do a little bit of a mid year review, so we would look at each one of them and have a little bit of this discussion about those and hear your thoughts one way, or the other. But I wanted to before going there more generally ask both of you what has been most surprising about this year in the DevOps space, disregard whatever seven, or 20 topics they are, but for sure, none of us would have could have predicted how a year has unfolded. So I wanted to start with a soft question. Marc, you go first.
Thank you. One of the things I think that was most surprising, and really interesting to me is how many companies are really getting towards state of the art practices and tools in terms of DevOps and software development, the bar is really raising from some companies where they are taking the way that software is developed much more seriously than in the past. And I was joking with Kalle like 10 or 15 years ago, it seemed DevOps was kind of the bass player of the software organization. And now it's become something that it's more like the music theory of the software organization, it's something that every developer now is taking into account the fact that you must have pace, rhythm and timing, to use musical terms.
But this is one of the most delightful things and then this is also something that companies that are just coming on to this fact that their competition is now starting to move quite quickly. They may even have an advantage kind of drafting with the bicycle where they've been doing okay with their previous practices or legacy practices, and now starting to do things in a modern fashion using DevOps for example, can give them a great boost of speed and advantage.
Actually, I want to collaborate on that point. So what I see is that the capabilities that you can do is so large and they are so evolved already that you can actually front run your older competition based on that they have the momentum on going to a different kind of runway or they cannot change their way of working so well, or so fast. And the capabilities, you can just you don't need to do either the whole change anymore. You can just step one, you can jump over 100 steps, and just sprint after that. So from the past year exactly that those kind of even everybody knows these kinds of feed the DevOps practices that you need to do them, like Marc said that it is like, de facto, set of state of mind. But then, if you haven't been able to date big for before you can do it now, it'd be very much in a very nice speed.
Yeah, it must be supported, at least by the fact that it's sort of a self serving prophecy that because everything is becoming so easy. As a consequence, it has become so easy to build niche tools, and niche software as a service, which then fuels a virtuous circle, that whatever niche you're looking at, you always have a choice. And because you have a choice of so many niches, that means you can focus on your actual business problem so much more quickly, instead of having to tinker with the fundamentals that has nothing to do with your business case, it just back in time you had to, if you were to be a big engineering organization, you had to build your operating system distribution packages yourself. And now there's an abundance of choices for different Linux distribution, it's a matter of picking the right one. And that has now permeated through all areas of software, nobody in the right mind would start over on any niche before looking at those 10 options available out there.
Yeah, like me 10 years ago, we're watching this telco, industry specific CICD pipeline, and for that this is the greatest and this is what we can do and nothing is almost better. But now actually, what we are going to talk about more in the edge cloud area, for example, I feel that again, this is the greatest of everything that we can achieve. And let's see one year forward what we can achieve more.
Should we start with that edge cloud thing because I have been trying to wrap my head around that one. And truth be told edge cloud enable SAS for OnPrem Solutions that has been the hardest for me to capture of all of these trends, so why don't-
Why, it's so easy.
Exactly. So it give us a 30 second rundown on, let's start with what it is, and where are we going with that?
Yeah, let's go with very concrete and very, this kind of simple example. Everybody hates or loves electric vehicles, so let's think about the way Tesla actually is delivering and using its cloud which is an actually car itself. In the future millions of cars are managed with automation, they are getting the software deliveries, they are generating data, they are sending the data back, they are actually calculating different models already in the car, you need to update the software in the car to do that, you need to have the connectivity and all around the cars and then the software side where you actually use all of your assets in the world. All of them are when they are driving, they are your it's cloud. So how do you manage the software there? How do you actually think about in the development part already all of these capabilities that okay, you have some kind of platform there, you manage the version etc. So I hope that actually is enough already for you, but we can continue, but edge cloud is everything that is near the user. It is not centralized, usually, capacity, it is very decentralized capacity in the world. It can be in the space, it can be underwater, it can be anywhere.
Yeah, one of the interesting things it, you can think of it as cloud computing is way out there somewhere that it was actually thought of as fog computing this edge for quite some time. But but the marketing isn't so good to see you don't want to come out of the clouds and into the fog. Do you?
It's very interesting, and if I'm little bit continent because the 5g, and the edge cloud, were the two what I was holding into this, or I want to push these two items into this document because I have seen in couple of years now the huge tsunami of different industries adopting software in different kind of machinery in the forest, in the air, on the sea, and undersea. And they have the same problem base that have software as an asset that they want to manage, wherever they have connectivity. So you need to have the connectivity, and then the game is open, but all of these DevOps problems are then on in your lap, you need to somehow code your asset, you need to somehow develop a delivery them, you need to release them.
And delivery and releasing is totally different thing, which again is a problem in fleet management scenarios, where you don't want to push the same binaries all the time on over the mobile network because it's going to be terabytes of data, it's going to cost you millions to update your cars, your planes, whatever. So the whole system is your new problem field, your all of the assets in your system needs to be managed in an automated way, with a scalable way that is not even possible to do by hand. So this is just this is so large tsunami that needs to be built in the world. And there are front runners, and then there are laggers that are still sending USB sticks with people into the rain forest to update the software in some hardware. So that's not the way to deliver software.
Which used to be the problem for the IT department. And I think this is where we're sort of my mental model is, has expired such a long time ago, that I'm thinking that my conventional way of thinking is that you have the development team, and they deliver something for IT. And then IT has the problem of endpoint management, and then suddenly come DevOps. And then you say, Well, imagine there is no IT, imagine there is this development team, and then there's this ops team, and it's their shared problem. There is not the development team that comes to IT and say, here's the new build, can you deliver this to all of these endpoints, and then the IT is trying to figure the endpoint management out. So I'm deliberately a little bit of a devil's advocate here. But it sounds to me we have gone a full circle, went from client server model into a cloud computing, and then we're coming back into something we call this cloud.
I kind of look at it, would you agree with this Kalle, I like what you inferred Lapa but to me, it's more client client, or server server. It's like the tools are such that it is no longer just a singular kind of connection. Now, it's a parallel kind of connection.
Yeah we had in the DevOps 2021 conference, we had this Groke Technologies on stage, they were where we were actually, or they were presenting their way that how they are delivering software on their vessels through the vessels are in this poor connectivity kind of way. So it is also that the entities, the assets that you have in the world needs to actually be able to be self maintaining, in the same way that you have connectivity problems, they are going to be ping the master, or ping that type of stories, what they actually configured, and then they are going to update themselves.
So I have been taught 10 years that, okay, CICD is a push, it's a push oriented way to deliver value to the operational capabilities and value streams. But it's actually now pull based. And I see this is also a trend that everything is going to be transforming into this pool base, which is this kind of lean way of thinking already has shown us so well that push based is never a good idea. Whatever you want to do, whatever you want to change, or manage. It's always better to do pull-based systems. And this edge cloud is perfect way that the fleet management functionality for example, you have clear time windows, when you to actually do the download so that the intelligence is in built into the, asset in the field.
Well safe to say that trend has happened.
Yeah, basically we have all the time growing this kind of case amount in the area of this. So and it is industry independent. so it doesn't care if you are banking sector, or a manufacturing sector, the same thinking, and same solutions actually applies.
Cool. Well, we have one down. And that was your hobbyhorse Kalle. You had two hobbyhorses edge cloud and 5g.
Yeah, but let's go 5g later. Yes. Okay.
Right. Marc, should we pick up your hobbyhorse in that case?
I'm guessing you want to talk about psychological safety? I think that is one of the most exciting ones for me. And I've witnessed this firsthand, during this year, where I see companies giving a lot more attention towards the safety of their employees in terms of the ability to be themselves, the ability to express themselves. And we threw around these words of empowerment in for 20 years and in various different kinds of management circles. But now it's really working in a lot of organizations, we've had things like the pandemic happen, and you can see to this is a wonderful way to look through this because right now you see two kinds of companies, you see those that are just working remotely, and supporting their employees, and creating everything they can to to enable this kind of work.
And then you see some companies that are either sitting on their hands, or like, well, now we will allow you to work remotely two days a week once the pandemic is over things like this. So there's been a great deal of allowing people to work in the methods that they choose themselves, actually being empowered to choose not only the methods, but even in many cases, the content of the work. And the consequences of this, I'll just kind of reiterate is that it's becoming nearly ubiquitous, so companies that try to buck this trend are in fact losing people. So as many companies are getting better at supporting the employees, and allowing them to be themselves, and thrive. Some are looking back at older ways of working command and control hierarchies, things like this, and realizing that it's not that competitive in this landscape anymore.
Last spring, we conducted a what we call remote work survey, both for our own employees as well as our customers. And we had a question, I think there were two questions that that were interesting in this perspective, one was, what are the reasons in the order of importance to increase remote work possibilities, number one reason was to accommodate employees preferences. The number two reason was to for employees to better combine work and other life. And then number three, and four, respectively, was improved productivity, and their ability to hire talent from broader geographical area. And number five, which is the last one on the list is cost savings.
So if you look at this from an employee employers perspective, this small sample tells that thinking employees first is a way forward. And the cost savings which in a traditional corporate management would probably be a prime directive now comes as a consequence and I'm I it personally feels common sense to me that organizations who embrace have a culture like this, what our customers have said, then they will think it, or not improve psychological safety in their organization that people can more feel at ease with the working circumstances.
For that point, remote center, or remotely working one year has decreased our capability to talk as a group in my mind very much. So it's a problem for me that I have seen personally that when you don't have that kind of one to many collaboration all the time, seeing different people even and outside your team. It decreases your psychological safety so that you are thinking more that you should, you are the ideas are coming to your head without any reason that okay is my work, appreciate it, why I'm not connected, or do I need to conduct some sorry, talk to somebody. And when I have seen it because I'm very physical, this has been a very, not very difficult time for me. But because I'm an extrovert person, I want to see people and not be alone in one room, so there's a different kind of fears, or ideas coming to my head.
But I see there can be kind of an opposite effect where I see, well, first off, one thing that has relaxed a lot is the formality of the first five minutes of the meetings, where I see a lot more of this, instead of the meeting starting at one minute sharp, or everybody just waiting for five minutes. And well, we are waiting for some others, there's banter. And there is this kind of getting to know each other more in this first five minutes of online meetings after the pandemic has been around for a while.
And I think that's really interesting. And there's this kind of another side to that too, which is we're seeing inside of each other's homes all the time, and we're getting a glimpse of someone within their home environment. And to me, it makes me feel better about myself because I'm kind of also sharing and having that that shared with me. So it kind of feels we're still serious about our work. But this level of putting up a front of formality of I am a professional, and I will only talk about the work that's kind of dropped some and I think we're more human because of it.
Yeah, it was basically the news anchor in a TV, so couple of years ago, where they had this kind of small kid coming to them, that broke the ice, then you could have in CNBC, CNN, Sky News, all are collaborating from the living room, or something. It was interesting, and I feel the world needed that one. So I agree with you.
I remember one of the original underwriters of the Agile Manifesto. I learned this from somebody that he had a habit that he joined online meetings ahead of time. And he was playing guitar on his own in the online meeting with nobody on the line, and then when people joined they, in the course of time, they learned that this person is playing guitar in anticipation of the upcoming meeting. So they began joining early because they wanted to hear that person playing guitar, which was a nudge of motivating people to join on time, or maybe even early, it's a very human way of looking at that.
I still want to highlight that psychological safety is not the same thing as as remote work. You even though that's a very strong embodiment of psychological safety. But if we look at the definition of what that is, there's one one definition, which is that you can express your point of view, without being afraid of the consequences. One of the best expressions, or definitions I have heard is that the psychological safety is the ability, and the environment where you can take risks with interpersonal relations. I do personally think that, or everything that we have discussed have made it easier to take risks with interpersonal relations.
Yeah, and I want to emphasize that point wonder on a written format on slack, or teams, people actually that they don't want to take risk. They want to write the message in such a way that an email in a time that it's not a it's not a collaboration chat. It's kind of that it's locked, every loose end of the text is locked so that nobody can misunderstand it. Nobody can not feel bad about it, and be very clear on it, and it has, it's very difficult I have seen that in work because in a remote work you have not you don't have the face to face communication, you read, and write, and then you have misunderstanding possibilities in that kind of communication also.
But this ability to not know everything, this ability to ask a wrong question, this understanding that just within the DevOps landscape, the number of tools and technology is so vast that no one can be an expert in all of this stuff. So it's okay, if you don't know how to do this thing, and to ask about that. And the one of the classical tricks here as well, as, sometimes you ask for help, and you, or you asked a question, and you don't get an answer. But if you make a wrong statement, then everyone will come and say, correct you, which is another way to feel safe, and learn to feel safe doing this, and still get the kind of help that you need.
Great. So it sounds despite the circumstances, we have been able to increase psychological safety, or I hope other organizations have found that as well. It's slowly again, in addition to the DevOps trends report, our teams at Eficode have written a hugely popular guide DevOps for executives, I am adding a link to the show notes, and I encourage you to read it at your leisure. Also, you might find a few more links that interest you, take a look. Now let's get back to the discussion with Kalle and Marc. When I joined Eficode there was a person that Eficode who taught me a great lesson, which was that the psychological safety is not a characteristic of an individual, is the characteristic of an organization, and the organization develop psychological safety, and not individual. And from time to time, I'm coming back and thinking about that.
Yeah, and it's good point that this kind of team building activities have been very hard to keep. And in our company, I'm so glad that these kind of company wide parties, they have been quite fun and actually quite open and in my mind successful, and that it needs much more effort to keep the motivation of people, and actually build this kind of psychological safety in the company when you are not present.
So far, we have been on safe ground. It's almost like everything we say, we agree and things have happened. So should we try and pick something on the list, which has been more on unbalanced, or wobbly throughout the year? What are the trends haven't really come through as we expected?
Let's say basically this community building is, it's a difficult subject, it's so important that we actually wanted it to be here. It's like self evident that it's like a better format of your organization, and how you can cultivate the value out of the employees is through the community, and the culture, of course, but it is so hard to change. And it has been very hard even more in this Corona time, so but we can talk about that.
I have a few words, to start with on the community building. One of the things that I've seen is like practice communities, or and some other some places, they're called guilds, or whatever, which are internal communities where people come that have like interests. So it could be Kubernetes, or edge clouds, or lots of different kinds of things. But then it can work like it's a volunteer force without ownership that then doesn't really have a mandate, or power and kind of dies on the vine, or could even rot, and the rot could spread a little bit. So communities - this is something I've seen in the last year where it's we can't just expect them to grow on their own. You can't always push, or pull or guide them too strongly. You can give them if necessary mandates, or ownership of a community.
If you don't have a benevolent dictator, or a charismatic leader that is pushing something in a direction, and the community is not moving, what do you do? So this has been one of the challenges that I've seen this year, as a lot of companies, they want to have internal communities, or they've tried it, and it didn't work, or they have they've created volunteer, cross team, DevOps communities, or whatever. And then they don't really get it off the ground. And there's lots of different reasons that can contribute to that, the ownership, or the mandate, or the time allocation is one that we face as well. We're consultants, we work with customers, so where do we allocate the time in order to do internal things, or community related work.
I somehow in my mind, I associate community building with innovation management through communities. And if you do innovation management, through communities, or if you do innovation management throughout the organization, almost inevitably, you will talk about allocating certain resources on the basis of democratic decision making, basically, everyone who joins the company are given a certain virtual tokens that they can use to fund certain initiatives, or certain innovations. And so it's virtual money, but everybody has it, and that you can freely divide your equity to whatever internal initiatives there are. And then, by way of allocating virtual equity, then those initiatives will get the highest amount of virtual equity, they actually get actual resources. And, when you say that, you need a benevolent dictator, or what if a community dries out, then what then somehow, in my mind, I think something like that has to exist for it to live longer, that that virtual equity is somehow converted into real resources. Otherwise, it's a risk of just becoming a chit chat forum, and nothing ever happened.
Yeah, I but I feel that it is mandatory to have that kind of open space to talk about the certain topics, and then people can pull, or collaborate on the ideas. And it's the capability to execute on that I don't want to say execute innovations because it should be this kind of sorry, organic theme, where you actually notice that, hey, what we just did, we just did something valuable for us and our customers. And it is very hard to achieve that kind of situation and it can be broken so easily. It's a fragile theme, I have experienced from my previous company, we had this spark innovation, this kind of competition, we had 400 ideas, and the top 20 were given this possibility to actually have this kind of start up whole training, we had beats competitions, we had prize money in the end, we had everything and some of those actually ideas are now generating revenue as we speak.
So it needs to be, there has to be incentive, and time and then this kind of framework where you can breed, breed and do that as a sorry, planned way, so that everybody understands that this is the reason why we want to do these kind of communities also. And that actually, in my mind, the greatest cultural building blocks or is the fact that when you succeed to build something, you have a such a great team, that nobody's going to stop that team in that company, or in some other way it's going to be when they succeed in those kind of things. That's very nice things, and very nice to see that kind of team building.
I would still probably my role to be a little bit of a devil's advocate, but I would imagine that if this trend has really broken through, we would have seen examples of it from our customer side.
I usually our customers are not, are they have some kind of problems usually. That's why you use consultancy is the reality that you need to some kind of change the internal way of working, and I know that they are companies that are using practices. But what's the there is a thin line in my mind and center of excellence, and in the minds of many people, they don't understand the difference between center of excellence, and practice community. So they are totally to do different things, or they should be at least in my mind. And but in reality my kind of projects are not the "we are fine, please call Kalle to mix things up."
Other than the mandates and ownership and other things that we talked about, one of the things I've noticed about community building, in my experience has been, the people are attracted when they can interact with something, with the community, with the company, as a customer, you become involved in the community, when the company gives you a way to interact with them, or you can do peer to peer interaction, of course, but if you are a member of a community, are you then able to interact with gurus, or interact with maybe a management layer, or something that makes you feel you can get ahead, or have professional growth, or something like that. So the interactivity of community, I think, is something that does in addition to these other things that help them enable, value for the humans, which then translates to value for the company.
So we must believe that when the working mode becomes more hybrid, then that face to face part becomes more of the vehicle to community building. And which is, by the way, well supported by their remote work survey, which I referred earlier, where we the other question we asked was, where would you prefer Eficode people to perform the following activities. We asked that from our customers. And by far, the single individual activity that our customers preferred Eficode people to perform in the customers office was a team workshop. If you contrast that option with the different options of independent work, or a day full of meetings, or day for your meetings, or frequent the meetings, even frequent the meeting was preferred to be taken from home office.
But team workshop, which I imagine that when you are building a community, you end up doing a little bit of you go face to face, do something unplanned, do something which is very highly dynamic in terms of interacting, which has it benefits from being face to face. And the customers prefer that to happen face to face. And one last thing about the survey is, by far, the biggest option, for all practically all of these apart from this one was I don't have an opinion where the work should be happening. Where do you guys want to go from here?
I could quickly take the internal platforms, and design systems. This is something I've had a lot of customer experience with lately, and I find it fascinating that I've seen a couple of different things on the trends here. One is that more and more companies enabling developers to build, and utilize their own tool chains. So we don't care how you do the work, we just want you to deliver the roadmap, and then allowing the developers to use completely their own tools. And then I've also seen the opposite thing. So this is a fragmentation enables efficiency, and then I've seen the opposite where there are a lot of companies that are providing end to end solutions. So if you're going to build a physical product, would the first thing that you do be buy a factory? Well, no, that would probably not be the most efficient way to start making physical products. But when you are developing software, one of the most efficient things to do is to set up real proper professional pipelines day one, and then all the work that you do will be more efficient.
Now how does one do that? Well, one can allow the developers to go and pick and choose whatever tools they like. Or they can go to a single solution provider that provides a platform, and then they can build their pipelines on top of that. So I've seen a lot of work in these kinds of areas where it can go in both directions, it can be having an internal platform that is essentially a set of rules that you can use whatever you like, as long as you fit within these, these minimal things, or an internal platform, which is here is a standard tool chain, you can update it or, maybe you must use it as it is, but it is provided to you as a service. And now there are companies that can even provide the end of that entire service to you from change management all the way through to deployment.
And on the design systems, one of the things that I found in Eficode for example, if you look at our look and feel, we have quite well crafted design systems, that when I am trying to express an idea for a customer, I have a library of graphics, and I have a library of slides, and I have a library of layouts. And I have a library of icons, and I have a set of typography, and I can go from the paper to a professional looking set of slides in minutes because we have a design system that allows me to do so. And that's something that when it's done properly, it is a real luxury and a huge efficiency boost for being able to express an idea to another person.
What do you think about Low-Code? Gartner is saying that they almost every software line is not written in manual way that it's generated from Low-Code. So it's going to be interesting also, the design systems is kind of our first step on that direction, I believe that there is going to be some kind of Low-Code change coming. But let's see there is also always excuses, why don't you want to use low-Code. But I see in systems where you have millions of lines in database, or users in database, but you have couple of users actually for example, a government. Those kinds of systems, nobody's paying for that kind of manual software design development anymore on those kind of systems.
We will go from low-Code to the point where the AI actually looks at your data and writes that code, and gives you a bunch of used cases that it can find. And that's the next one.
That is the next one. But testing is never going to go away, and requirement management is never going to go away, and the customer feed back and that kind of thing. The manual labor actually could be minimizing but I want to point out this one idea that actually I met, or I was told last year, was it Johan or Henrik, I'd actually don't remember now, but socio technological system. So I have heard and read about systems thinking before this, but the socio technological system actually opened my eyes so much. And this is the very oldest 60 years old concept. But when you go to the wiki page and open the space history, you can see that the past two years there is exponential growth of changes to this page.
And if I would read one from DevOps point of view, if I would want to learn about how you want your organization to use technology in your behavior, or how do you want to rotate jobs, or how do you want to spread the knowledge, this is the system thinking from technological, and sociological point of view. And it is a key, in my mind, also, in these kind of systems, where you have platforms, you need to also use them, you need to teach people to use, and how do you accommodate that process to use those kinds of tools. So this is what DevOps is all about, I believe a little bit continue on this one.
I figured that the technology part is always a little bit in front of the sociological part because if you want to enforce a process to be to some people that don't have the capability to execute on that process, it's going to go so soft, so fast that you actually think about your, all of your actual transformation projects. Where the reason was that you are not able to do that your circle of your work that you can do is smaller than what is expected from you. So in a way that the technology makes your circle much bigger, and then you can execute on those things that are expected of you. So and this is a little bit idea from both from the bottom to top, or top the bottom which way you want to transform your organization. And the answer is both ways.
So my takeaway from you Kalle is that I had a question in closing, what trends would you like to see come up? The socio technical system might be one trend that you would like to see come up based on-
Yes. And the optimization of this one, so that you listen what the architects are saying, and also that architects and technical guys needs to think about the whole system. How do we manage this one in production coming back to the age cloud, for example, if you only build a technological asset that is not capable of being managed by the process of the organization, then the whole thing is a proof of concept, and not maintainable in operations.
Cool. Guys, we have five minutes and three topics. So shall we stop here? Or do you want to take one more?
I would vote for the 5g.
Go ahead on the 5g, then.
I started my career in actually in creating software for doctors, where I was confronted of the fact that if you don't create the software for the doctors very well, and it's not usable, they don't use it. So two to three seconds is a time that it is they evaluate the software that is usable or not. But the point is that I actually started my real career in the telco area, where I was building these huge, very costly systems for big vendor, a Finish vendor, you can guess who it is. Now 15 years later, there is a transformation on going so that for example, Ericsson and Nokia, they are not selling their hardware anymore to operators, or to connection service providers. Basically, what they are now selling, they're selling software to companies, and talk about vertical inter integration, and owning your operational value stream where you need a connectivity to your assets. So this is if you combine the edge and 5g, do you have a copy of the actually the PDF there, what was the market cap that is untapped? Can you please check it quickly? It was huge number.
I do have it, just give me one second. The Edge cap market will be worth 43 billion by 2027.
The point is that this is a totally untapped market. And it's when cloud was coming, you can imagine what kind of numbers were and it's still growing and cloud services that you can use, and indicate to your operational value stream, that when your customer comes to your service, it is actually running in cloud, that kind of transformation of all of the companies, and the revenues, and where the actually money comes from. And now you can imagine what kind of value stream changes on operational side you can have, when you have, for example, I'm going to take an example like a port now.
The maritime port handling logistics, for example, they don't need to anymore have a partnership on the country level, on the operational side, they can actually own their own network, and manage everything, and managed scaling of their business is not any more depending on the native localization problems of the connectivity, and spectrum usage. So it's hard to imagine what kind of event changes this can, and business models this can give to the companies so for example, also the factory floors. So there's three mega trends in 5g, there is massive IoT, then there is this more broadband large, very large broadband and the latency. So I have this dream to play Angry Birds on my mobile phone, so that I'm augmenting my reality so that I can pull my bird, my rate to the nothing now at this time better, more momentum and tip it over and the latency is so good that you are not even seeing the difference.
And that kind of change, or from user experience to build the augmented reality that we can have glasses finally, that way I can see, of course, I'm married, but if I was looking for a spouse, I would see immediately that Okay, there is a possible spouse candidate, these kind of they have been a dream, but they know they are coming on this kind of Blade Runner. Blade Runner, cyber punk reality is going to be able to build on this 5g and explode capability. Sorry for being a little bit rambling here, but I'm a this kind of, I think, far to the future.
But it's here and now you said there are examples of this happening, and 5g is basically, again, a virtualization of resources to another higher abstraction level compared to before, which is very natural development of technology.
We haven't really seen the killer apps yet, unless you count the COVID vaccine. For the killer apps for 5g. That was a joke if nobody's laughing. Thank you, Kalle and Lapa. So anyway the interesting thing that I'm kind of playing with here in my mind, is that, okay, so this could go two different directions, and I think it will tip one way or the other, whereby the device that you are using the endpoint, the terminal, the laptop, or the phone, or the smart glasses, or the electric vehicle, or whatnot, has due to the latency, and the bandwidth has almost no processing power, and no storage and everything is streamed. And we're seeing this in the gaming industry a lot as well it could go tip in this direction to whereby you don't own anything, everything is a service that is provided to you, and you have a very thin layer client, and then everything else is on the, there is no cloud, it's just somebody else's computer.
It's somebody else's traffic sign.
People that want free internet, they don't understand that everybody has to buy it from somebody else. But then on the other side, of course, is the world's greatest supercomputer, the conjunction of the billions of connected devices on 5g all over the world that now can solve all of life's problems. And I did a presentation on this some years ago, where the Seymour Cray, the inventor of the supercomputer, compared a single human cell with a computer and put it into those terms. And by my math, all the smartphones in the world at that time, if given good connectivity, what add up about to the brain of a rabbit. So as time goes on, if we do go towards more power on the edge, we may not even need the cloud anymore, and everything is kind of distributed in the fog computing.
And you Lapa actually asked about the ideas, what were the trends, so one trend was actually this kind of decentralization and of course, the block lectures, and cryptocurrencies are going to be a part of that also as a decentralized solution. So that was one thing that we dropped, actually, so but my ideas were quite a little bit technology and enablement point of view and Johan was this, Johan bring this kind of soft elements that this these publications.
This the social technological, where two of you represented one side, one respective side of the social technological?
Wonderful. This is great. Thank you folks. And this leaves two more for people to read in the DevOps trends, but you said it earlier, SAS is pretty much preaching the choir, and the last one is increased focus on data challenges, which I think we have to talk about data at some other time because it goes into building information architecture and technology architecture side by side. And it goes into talking about building data lake and things that as part of your continuous deployment, and how do you do things like feature adoption as part of your continuous deployment and things like that. So that definitely deserves an episode of its own right. Any last words? Kella and Marc you'd like to say before we finish off?
We have touched already this little bit, but I was emphasized that when Coronavirus spread, the economy was taking a hit, and then a couple of months later, everybody was panicking, that how we are going to work, how we are going to deliver software. In this kind of scenarios where we can actually have a flight we are stuck here in the Nordics. And this has increased, or amplified the pace in the development part, in the collaboration, the psychological safety, how do you build the teams? And then these operational practices how do you update software? How do you monitor actually the software? How do you get the feedback when you cannot even go to the field for example, the updating software in the factory side? So this has even more giving me this kind of, I don't know, it's more self evident that these DevOps practices is needed. And all of this paper was written in a perfect timing in that sense.
Thank you Kalle, I'll just say a couple of words that I think one of the most interesting things that is moving with momentum, and the trend is increasing as that software complexity is rising, but the ease of use of creating many different kinds of even sophisticated software applications, the ease of use of creation is also increasing. And I think we're finally starting to break through one of The Mythical Man-Month Brooks laws about complexity. So the complexity of a system is the square of its interfaces. But we're DevOps, and pipelines, and the current kinds of tools, and platforms are actually simplifying the interface between the developer, the code, and the delivery, rather than scaling exponentially with the complexity of the system so that developers operating, and I think that's a really fascinating trend right now.
Good point. And I feel we are moving from this strongly centralized way of running things to the more decentralized anyway, in the technology, and in the sociological aspects.
Thank you for listening, you can find links to the social media profiles of Kalle and Marc in the show notes. Alongside related materials, that could be of interest you. If you haven't already, please subscribe to our podcast, and give us a rating on the platform. It means the world to us. Also, check out our other episodes for interesting and exciting talks. Before I let you go, I want to offer a floor to Kalle and Marc to introduce them properly. All I say now is take care of yourself, and remember that the future is in the edge cloud.
Hi, my name is Kalle and I'm lead consultant in Eficode. I am here to make organizations to be better in delivering the value to their end customers, and everything that is in the way I want to remove.
Hi, my name is Marc Dillon. I am a lead consultant at Eficode. I have been building software for many years now. And I'm really excited about the way that DevOps and software trends are allowing the creation of greater things with higher quality, and less cost.