JavaScript - A new fever for open source - trends in the case of Bentley Systems
We do not really sell software, we sell the software result. People do not pay us for the software, they pay us for what they do
Bentley's growth has come largely through acquisitions. Two of this year were British. Synchro; the planning software, and Legion; the crowd and pedestrian mapping program, both widely known and respected in the United Kingdom. Its integration with Bentley's design and asset management systems will expand its use and bring added value to subscribers of infrastructure software. Bentley also produces some homemade product; 2019 will see the launch of iTwin Services that seeks to create the concept of "Digital Twin", which is the natural end product of Building Information Modeling (BIM), and the open source library iModel.js that will feed it. What was that? Open Source? Are we expected to believe that something we can not see and can not buy will generate money for its developers? Explain that.
Has there been a number of Bentley acquisitions this year, which has excited you the most?
I am easily moved about many things, but sitting down and looking back at what people currently do with our software is really sobering. There is an incredible potential for integrating these solutions with our product offerings. I find fascinating how Synchro has made a big difference for users. I have also been impressed by what people are saying about Legion. I think everyone should be using Legion!
In the United Kingdom, we now have the Geospatial Commission within the Government. What is it about the geospatial data that is causing governments to appreciate its value?
The concept of going digital is beginning to resonate. People are beginning to realize that if the information is there, it should be exploited and used as widely as possible. Only the existence of accurate and timely data is having more demand. That trend is sure to continue. People are going to demand more access to more information in time and with more form factors.
What is this thought that was behind the open source library iModel.js?
We learned that the information stored in files related to our design applications can be related to information from many other external sources; GIS, mapping, asset and road systems, for example. And we knew there was a call for better incident tracking and other kinds of live reporting. So it seemed natural to match the view of the road with the design of this road and with the most recent traffic on the road. People have daily experiences with using apps for this kind of information, and they can't understand why it should be difficult. We should be working on making those connections as easy as we can.
There is a lot of talk about "the dark data", what is that really?
In the world of engineering, each application is designed to solve a relatively specific problem, and many of them were conceived years ago. They store their data in a way that is easily accessed by the edited application. Most of the time - and I speak for our own applications - logic is like understanding that the information is in the application, not in the file. The file is just a series of bytes and when you try to understand it without the application, it is incoherent. The darkness is that other applications can not interpret it and visualize it perfectly.
We are as guilty of creating this situation as anyone. But the state of the world now is that we have an impressive number of applications that are needed to develop a cohesive whole of a stack of independent files. No one can achieve it. We have data and they are valuable, but we are wasting them.
Open source is a big step forward for Bentley, why now?
I have been advocating this for a long time, but you can not just open the code body that is in the encryption pond. If we had developed open source in our applications a few years ago, the construction process would have been very complicated. Just explaining how it works is below the ability of the casual observer - and the only successful open source applications are those that a casual observer can make sense of. Maybe that casual observer does not currently change anything, but they are the reason for open source - it's because people can use it for things it was not designed for.
When we started with our project in iModels, we thought it would not be valuable unless people could use it for things that it was not designed for. We needed a way in which people could use it without going to the "Bentley School". We chose JavaScript as the ideal language. JavaScript is everywhere. It's amazing how he has taken control of the IT world. We had then to convert a lot of code written previously, now in JavaScript. We had to invest a ton of time to look good, be well documented and well commented so that we could sell open source access as something of value. I can not tell you how many open source projects are advertised with fanfare and then ignored!
We do not expect that just because it exists, that people use it. We will have to work hard to prove that using iModel.js is worth the investment and time.
Did you encounter any resistance within Bentley over open source?
Quite! There was a strong current at Bentley Systems that said it was a terrible idea. We are a software company. We sell software People believed that I was giving away what they were trying to sell. And I kept trying to explain that we do not really sell software, we sell the software result. People do not pay us for the software, they pay us for what they do.
It has meant a change in the business model. It's similar to when Microsoft decided that Azure was a way to get money for helping people use Linux. With our new iTwin subscription, we can say; here is the whole source of the program that creates and visualizes the data, you don't need to pay for that, we will charge you for the iTwin subscription and with that you will have a vast sea of applications available. Some people will give it away. Some don't. But the ecosystem that we find everywhere in the JavaScript world is second to none. You couldn't create a closed source competitor for JavaScript. It would not work.
You said that a lot of open source software is ignored, what challenges do you face in gaining interest?
Make people find that priority is no.1. But that's just the beginning of the game. Then they will prove it. They will have questions. They are going to have problems. They will want to make changes. They will suggest alternative ideas. Being able to respond at all these levels is what makes an open source project work well.
Open source software must obtain a critical mass before people think they are part of a bigger issue. No one wants to be working on something if they think he is dying. Being open source does not mean that people will magically accompany us and become viral users of our products. We will have to make that come true.
I am always impressed with the amount of effort that Google and others put into their projects. They do something open source, and then they put a marketing team to sell it. If you ask something, someone answers you. Any problem you have, there is someone there to help you, not always from the original source in forums and online communities. They have a tremendous ecosystem of examples. It tends to feed itself.
Imagine you are writing a program. If you are not going to publish your source code, it can be opaque and complicated. If you work, work. But if you're going to say that users can put their layers of things on top of it, if you're going to suggest that it's a point of entry for other people's work, you have to prove it's worth their time. It is not an obvious step forward. Ten years ago I would have said; No way, it's very hard. But the combination with the iTwin subscription model and the fact that the ecosystem for the open source world has been established, means that we hope to capitalize on it.
In recent years we have seen more collaboration between the largest companies, Bentley works with Microsoft, Siemens and Topcon among others, why is that?
Until a few years ago we never really co-developed anything. For a while, we said that we were neutral and that we supported everyone equally. But Topcon and Siemens and the others came, and it seemed like a model that could work; we would both get profits. Sometimes we have debates about where the limits should be between what we do / do and how much they should pay us / how much they should be paid. But I think we're both better off than if we did not have those cooperation agreements.
In the case of Topcon, we work together when it aligns well with our priorities. We always try to keep them informed of where we are going, so as not to overlap. You can not do that with everyone. A special relationship is no longer special if you have that kind of relationship with everyone. That idea of a cooperation agreement, where we currently unite developments, has become a model that is working very well. I could not have predicted it. Frankly, I was not a believer in the concept, but I'm happy that they could prove that I was wrong.
As the founder of Bentley, what are you most proud of?
We have made 105 acquisitions, some of them are more prolific or have lasted longer than others. But what we acquire many times is really good people. A large percentage of our colleagues came through these acquisitions. If you are a small business and assimilate a larger company, then there are two routes you can follow: follow your way out and return to a small company, or see the opportunity. We have managed to convince some very smart people to stay.
We are an amalgamation of 105 companies that have come together over the years. I may have started it, but I can't take much credit for what we have become. When I sit in the back of the audience and watch a Synchro demo, which is now called "Bentley Synchro," I think to myself, man, those guys are so smart. I am living in his reflected glory. I felt the same way about acquiring Acute3D a few years ago. Those guys are brilliant. They created this wonderful tool. We acquire it. I look at her, and I say to myself, man, my name is there. That's very good.
How do you feel about the size of Bentley now?
When we started, I just tried to stay in business long enough to pay the bills. At one point of time, I knew every person who worked for Bentley Systems. I knew what they were doing. He knew his children. That is different now. We have expanded into spaces of problems that are not the ones we faced in the beginning. We have expanded to markets that would not have been our normal market. Our reach is much wider than it would have been if we had only grown organically. What was the premise to start Bentley? I was working for DuPont, who was an Integraph user. My brother Barry had started his own software company, and I left DuPont to work for him. Meanwhile, DuPont asked me to improve some software that I had written while working there. I told them I would improve it if they gave me the right to sell it. And that was the beginning. I started Bentley Systems and started selling CAD software.
We interviewed Greg Bentley back at 2016 and asked him what it was like to work with his brothers, how did it seem to you?
I advise you not to do it! But it has turned out relatively well. We never really had a complete plan. When we started the company, there were five of us working there at that time and my mom was petrified. She could not believe that software was real. You could not make the idea that people would pay for something they did not look at. She was genuinely concerned that all her five children would be unemployed and return home.
What do you expect most from Bentley in 2019?
The concept of digital twin. Someone is going to make it. Whoever develops it really well is going to have a bigger market opportunity than what exists now. This opportunity, this break point in the industry where there is a great transition between the current disconnected world and the digital twin world is a market that we have to embrace as quickly as possible. 2019 could be Year One for us.
I was there in the beginning of the computer days. The computer was brand new, and everyone was speculating what things might be possible. I think we are at the beginning door again with digital twins. It is not a new concept, construction and infrastructure are the laggards in this. If I look at the way business develops in 2018, it doesn't look that different from when we started in 1984. Yes, we have digital paper. Yes, we have the 3D models. But contracts say the same thing, and people generally build in the same sequential way as before. Things like Synchro are revolutionary, but they are not widely used. In this next phase, many things are going to be different.
Any result that comes out of the opportunities created in the digital twin world, is going to be an open source world. I'm sure of it. I would be petrified to compete with him anyway, so we want to take the lead. It's easy, after almost 35 years now, to say, I'm done. But I feel we are on the starting line of a race that is going to turn into the next gold rush.
Keith Bentley, Founder and CTO, Bentley Systems, talking with Darrell Smart and Abigail Tomkins.
CES December 2018 / January 2019