The Spring Framework creator speaks to JAXenter about his startup portfolio, VMware frustrations and rediscovering his love of piano.
Following the news that Spring Framework creator Rod Johnson would be joining Hazelcast’s board of directors – the fourth such startup he has joined in the space of a year – we caught up with the ex-SpringSource CEO about commercializing open source projects, the internal politics of big companies like VMware, and, of course, his PhD in classical music.
JAXenter: Why join Hazelcast? What drew you to the company?
Johnson: I guess one of the things that really interested me in Hazelcast is the size and strength of the open source community. I mean, obviously I have had quite a bit of experience in open source businesses, most notably of course SpringSource. But I think one of the things that really stands out about open source businesses is [that] it’s really hard to create a business and then grow a community. Whereas the right way to do it is [to] create something that developers love, build a great community around that, and then think about, y’know, what may be the business opportunities that arise from that.
So, I don’t think there’s a huge number of things out there that developers feel strongly about, and I think Hazelcast is one of those things where the community is quite large and growing rapidly, and also where a lot of developers just feel really, really good about what it does, and the existence of that capability in open source.
How will you be contributing to the company (besides a portion of the $2.5m investment)?
Well, as a board member, I think the role varies depending on the company and depending on the stage. So, one of the reasons I actually enjoy being a board member is that you can’t really plan anything over the next, say, six months, and say, “hey, this what I’m going to be doing”.
Typically, what I’d expect to do is work with Talip [Ozturk, Hazelcast CEO] and the rest of the team in terms of making the transition to it not only being an open source project, but to also being a company [Note: sentence clarified 22/9]… work with the executives and the other board members in terms of getting the marketing and sales engine working… and also working with the Hazelcast team to try and make sure that we are doing the right thing on the product for existing and future customers.
Why hasn’t the open source movement disrupted the in-memory data grid market already?
There have been a couple of open source projects. But I think one of the issues has been a focus in terms of what problems developers wanted to solve. I think, broadly speaking, there’s been less open source activity than you’d expect in this space, and I think some of the other players have been kind of more focused on going into a direction they found exciting than a direction that the community found exciting.
I think also – Hazelcast is Apache licensed, which I think’s a positive thing in terms of community adoption. And I think one of the things that’s just incredibly important in growing a community, is Hazelcast is really, really easy to use and get started with. And I think that not only attracts people to use it, [but] it also makes people feel really good about “wow, these guys have created something that’s actually making my job easier”.
I think that kind of ease of use, and helping developers become more productive, actually creates a community that really wants to give back. So, a community that thinks “wow, these guys are really doing good stuff, maybe I should tell my colleagues about it, maybe I should eventually contribute”.
Do you think it’s important to turn community projects into profitable businesses?
I think it is. I mean, obviously this is a matter of some ideological debate. But the reality is that, if you take any piece of open source that is both complex and important – so, starting with Linux – the majority of the people who work on that software are employed by companies that derive some benefit from it, direct or indirect.
I do not believe that it’s really viable to rely purely on volunteerism in terms of open source. And especially when you look at something like Hazelcast, where you’re dealing with very complex problems like replicating data in-memory across potentially a very large cluster. This is not a very easy thing to do, and I think it’s something important that the product can have a very predictable amount of investment. And it’s important as well that there’s the existence of y’know, commercial support for the companies that would feel more comfortable if they had that.
I’ve always had a pretty strong view that open source is definitely a good, and that given a choice, it’s much better for something to be out there in the open and enable people to work with and contribute to, but I certainly think that – just as in any complex piece of software – I think you’re safer if you’ve got at least one company that’s incentivized to develop the software.
Hazelcast is one of many investments you’ve made over the past year…
I wouldn’t say one of many. Neo Technology [in 2011], Meteor, Elasticsearch and Typesafe. And I haven’t invested in any infrastructure software beyond those things.
How do you split your time between those five?
It varies quite a bit. You want to be helpful, but you don’t want to interfere. So, I think for example, if I was putting too much of my time into any one thing, I think management might think that I’ve crossed that line. In general, how much time you put in depends on what the company’s doing.
There can be a number of things that can involve putting in a lot of time. For example, recruitment is one of those classic things where a board member will really try to help a lot with trying to get people on board. Also, say, trying to brainstorm around a product roadmap, those kind of vision issues often can take a bit of time. And then there’s situations where everything’s going great, and it’s more a weekly call. So, yeah, it really varies quite a bit.
You’ve made an investment in each company, and you’re sitting on the boards of each company, right?
Essentially, yes. But the investments in some cases are really pretty small. So, I mean essentially in all of those companies, I am really an independent board member rather than an investor. I am actually not on the board because I invested, I’m on the board because I have some knowledge of the space and can hopefully help. The exception is Meteor, where I actually represent Andreesen Horowitz. So I am actually the board member representing Andreesen rather than myself personally.
So to you, it’s less about the money, and more about wanting to be involved in these companies?
Yeah, I mean, to be honest, I would be very happy if I make money out of these different companies. But the biggest motivation is, for me, frankly, trying to help other entrepreneurs. I did quite well out of SpringSource [Note: VMware paid US$362m in cash plus US$58m in stock and options]. So I would like to make more money but it’s not the most important goal I have in my life. So I actually find the process of working with talented technologists who are setting up businesses – I find that a very rewarding thing.
How do you choose which companies to become involved in?
Well, three things. The first is the technology: do I believe in it, do I think it’s a good technology? I mean, I definitely couldn’t imagine myself being interested and being involved in something that I thought was inferior. Secondly, do I like the people? Do I think they’re smart people capable of running a company, and do I actually want to spend time working with them? And thirdly, whether the company, I think, has the potential to succeed. I wouldn’t see myself investing in something that I thought was not very likely to succeed.
Plus, of course, I ideally look at things where I honestly think that I can contribute. There are probably plenty of companies where I’m not sure that I could really help that much. Whereas I think with every one that I’m involved in, they’re using the space and the kind of business I think I can help with.
It was just over a year ago that you left VMware. Was this to pursue these entrepreneurial ventures?
I just don’t think that I’m a fit in a big company. I actually – there were definitely things I liked about VMware, but I felt to the point that I’d looked after the people that I’d brought into VMware, I felt like I’d proven that the business that we brought in would grow very nicely, and just to be honest, it was just not that exciting to me, to be an executive in a big company.
So, when I looked at what I wanted to do after that, I just find the whole startup experience way more satisfying. When you look at any big company, you see a lot of politics. You see a lot of people who, frankly, probably aren’t doing that much constructive. And I’m certainly not singling out VMware in particular! I think this is true in general.
Whereas the marvelous thing about a little company or a startup is, with few exceptions, everybody there is working very hard to do the best for the technology and for the company. You don’t tend to get a lot of individual agendas. If you get an individual agenda from someone in a startup, that’s the exception.
Whereas in a big company, yeah, you get a pretty good chance of sitting in a meeting with people of whom two have an agenda that is very much around themselves and their group, rather than the company at a whole.
I mean, the other thing is – in big companies, it’s frankly more difficult to get things done. You have this strange contradiction between having a lot of resources and finding it incredibly hard to get things done due to politics. [laughs]
The Spring Framework passed its tenth birthday this year — what do you think its legacy has been?
Well, I think certainly – it’s a little early to talk about legacy. It’s pretty much alive. I mean, if you drop by the SpringOne conference in Santa Clara, I think they had something like 1200 people. So I think Spring is still quite alive.
I think the biggest legacy is that it really played a very important role in simplifying what people did in enterprise Java. I think people now forget just how bad it was before Spring. In, say, 2000 through to 2005, when there was EJB2 and the heavyweight application servers were dominant, I think Spring greatly simplified things, and essentially killed off the traditional Java EE application server. And I think both of those things will be very beneficial to developers.
In July, you said Scala needed “to see more evolution in the community than in the language itself”. Do you think your comments have been taken on board, despite the initial backlash?
Yes, I think so, and I think it’s important to note that I wanted to start a debate rather than finish the debate. So with some of the things I was saying, I wasn’t so much saying “this is the answer”. I was saying “this is something we as a community really, really need to discuss”. And I was actually very encouraged by the panel discussion [“Scala in 2018” at the San Francisco meetup group] last Thursday night.
Because essentially we had a very open and very civil discussion on things like: Do we want Scala to be mainstream? Or do we want it to be something that’s used by a smaller group of people? How do we bring on new developers in a way that they will progress towards what we might regard as more idiomatic Scala? So yeah, I think overall it has actually contributed to community discussion, which is what I’d hoped for.
In an interview with Markus Eisele, you described yourself as a musician. Do you still make music?
Yes I do, I play the piano, and in fact my highest education qualification is a PhD on French piano music between 1830 and 1848. So, I still play. Not as often as I should, but I play predominantly classical music, especially romantic music.
Why didn’t you pursue a career in music?
You know, at some point, I realised – I mean, my first full-time job was in music. So, my first full-time job was teaching music history at the Sydney Conservatorium of Music, and I think at some point I realised that I needed my personal happiness to continue with both music and computer science.
Because at that time – that was the mid-90s – I was writing Windows shareware. And I kind of realised at one point that I had it the wrong way around, that I’d be better to have a career in the one that had more solid job prospects. And then of course, I got so busy that I didn’t play the piano for ten years.
Since you’ve stepped back from Spring, have you found time to play piano some more?
I definitely have had some more time. Actually, I started – I bought myself a piano as a Christmas present in 2008. So for the last year of SpringSource as an independent company I’d actually started playing again. And I found it was marvelously therapeutic. So, I can remember actually being at home and having ten minutes between conference calls, and calming myself down by playing Brahms.
So, yeah, I definitely have more time now. I’ve also started to do a lot more exercise, and I’m also trying to pay back some of the debt to my kids in terms of spending the first years with their father being a CEO who travelled 150 days of the year.