Developer

Taking developers into the interface

In the second half of our interview with Matt Thompson, director of Sun Developer Network, we discuss JavaFX phones, Sun's view of Google and Adobe, Swing's appearance and just how much of a bubble the industry is in.

In the second half of our interview with Matt Thompson, director of Sun Developer Network, we discuss JavaFX phones, Sun's view of Google and Adobe, Swing's appearance and just how much of a bubble the industry is in.

When will there be a JavaFX enabled mobile phone in Australia?

A number of us have prototypes of phones that run a portion of the JavaFX stack. Unless our plans change we will be able to show you something when we are back here in March.

Some of this has to do with our strategy and some of this has to do with the fact that we have been working with phone manufacturers and carriers for a number of years. Without commenting on other companies strategies, we know that the successful strategy in the mobile space relies on incredibly strong partnerships with the carriers and manufacturers.

Having said that, you can assume that we followed a strategy ... and that those are the conversation that we are deeply involved with right now.

The one thing I can tell you is for Mobile FX is based on one platform for all comers, that this makes both the carrier's lives and the manufacturer's lives much easier.

So the conversations that we are having are not of the "here is a new technology, can you please use it?" It's more along the lines of from the carrier's side, from the manufacturer's side "how fast can we have it?"

How big of an issue is the dilution of the Java brand?

Java is actually more prevalent now than ever before. We've done a couple of brand studies that show that people recognise the Java logo and the Java brand as singularly synonymous with technology and advanced capabilities.

Whether that be because the Java logo is booting up on your phone — I have a Blackberry, when I go to "about my Blackberry" it first shows information about RIM, on the next keystroke you hit brings up the Java logo because all the applications are written in Java.

The open sourcing of the Java platform, and there are a lot people that said that it would be detrimental to the Java brand etc.

By open sourcing the technology we've allowed the technology into a lot more people's hands; you can do a lot more interesting things and the brand is still very secure.

The concept is that as you extend the Java platform you have to maintain compatibility — compatibility is what controls the brand. The way the Java brand would be diluted is if that compatibility was not maintained.

It's pretty much a guarantee that we will continue being compatible going forward.

You mentioned previously that you do not see Microsoft as a competitor. Is Sun's big competitor Adobe?

I would say Adobe is extremely well positioned to lead developers or designers, who are now starting to add behaviour to their "applications" and I use "applications" loosely because designers work with an interface, an interface is usually connected to behaviour, that behaviour is what defines the business logic of an application.

In the past, a designer usually would not be involved with behaviour, it would be all "when this happens, we want to display x".

With JavaFX, and the platform and JavaFXScript, I would say as Adobe starts to lead designers into the ability to create behaviour behind their design, we believe it is a better model to take developers into being able to create easy, lightweight front-end, high quality UI designs for applications that they already have.

Do I consider them a competitor? No. I think that they are in a very extreme space right now, and I think between the companies we will be offering developers more choice than they have today.

Do you see Flex encroaching upon Java?

I would say that they are certainly moving into a similar space, a competitor is really someone that you look at and go, a competitor doesn't instil fear, but you have to see a path by which they can lead developers down to where everything works together.

I think that one of the fundamental challenges that Adobe has shown over and over in this space is that they own the space for designers — there is no doubt about it, brilliant. The thing is that what designers want is easily manageable behaviour and the moment you step into any real world application and have 50 million users touching it, behaviour just gets a button and you are going to find users who do things that you don't expect, and that means that your backend needs to be more robust.

It's whether you design an application from the front to the back or the back to the front. We believe in many cases that the business logic is what drives the value of an application or Web service. There are companies like Yahoo and Google, and even in some cases Amazon, that are partners and competitors for us in that space.

Adobe is in an interesting place because they have the ability to pull these things together, but I don't think they're there yet.

Why do you see AJAX and Google's use of AJAX as a transitionary step?

Today what they are putting forward to developers is an AJAX-centric model for client-side development. It'll be interesting when they start producing code for other platforms — we've all heard rumours about whether there will be a Google stack for a phone — does AJAX play a similar role there or will it change to something else?

The Google Web Toolkit is a fantastic technology, we use it, we show developers how to use it but there are other technologies that are even more compelling.

You've mentioned that Sun's business model was based on increase infrastructure demand, how long is this sustainable?

If Jonathan Schwartz was sitting here today he would say that "we are at the infancy of this build-up", and I actually truly believe that.

It stuns me, it really stuns me, to look in the datacentre of some of these start-ups. Twitter started on a very small scale server, if Twitter continues to grow we are talking about literally dozens of machines to handling the load.

I hate to use Google as an example, but it's an easy one for people to figure out. Even though Google is highly secretive about the infrastructure that they run and how they build it, all you have to do is look at the real estate development that Google has gone through recently in building out datacentres. They're not empty, they don't build space for one machine and go "done!" The reason why there is such demand for things like cooling and power is because as the load grows, their infrastructure demands grow.

The question that we then get asked usually is "Those are great, are there one or two or three or five or 10 of those?" The answer is that we believe that there are thousands, if not tens of thousands of companies that are going along the same lines. Now some of those people will not build their own datacentre, and they shouldn't have to. So we have network.com, which for us is going to be built out as an answer for some of these companies who do not want to run their own datacentre.

You sit on a board with venture capitalists in San Francisco, in your view, are we in a bubble?

Right before I got on the plane, I was reading in the local newspaper that the question was whether China was in a bubble. There's so much venture capital going into China right now, and right now it has all turned into gold, almost every major investment made by anyone of value has paid off through acquisition or just revenue. So if you look at China as an example and then you throw in India, South East Asia, Eastern Europe, South America — it's highly speculative. But that's really what venture capitalism is based on, if one out of your 10 bets payoff, you are doing phenomenally well.

And the fact is, that there is more money being put into venture capitalism today than ever before, but it's actually less money per risk, because they are actually placing more bets. The payoffs are so potentially huge, the risk of each individual investment is actually quite low relative to the entire portfolio.

As a venture capitalist, if you make 100 bad bets then yes it is going to hurt. But if one out of one hundred works out of the investment levels you are making, then that's great.

Are we in a bubble? I would argue that the venture capitalists have become a lot smarter in how they invest. They are investing for short cycles upfront, there are many people investing in an eight week lead. Step one, go from idea to reality in eight weeks. Eight weeks later, over the next three months build you user base to something significant. After that, what's the strategy for turning profitable?

Those conversations in the height of the bubble where taking place at the beginning of the bubble and might be two years out. At the end of the bubble people were having them almost everyday, like "what is our exit strategy?"

Today, the VC is having those conversations within the first eight weeks.

There was this joke roaming around the VC circles a year ago that the plan was to be bought by Google. I bought this up with friend of mine who's a VC, and he winced when I said it. He said "We are getting a very bad reputation, that's not the plan, it should be the plan." The fact that Google has picked up some of the high-fliers that have been funded by VCs recently is really the exception more than the rule.

I don't think we are in a bubble, I think the VCs are a lot smarter this time and entrepreneurs are begin asked to deliver a lot more, a lot quicker.

The Darwinistic effect of this is that if you are not fully into it, you will be gone within three months rather than burning money for two years. So that limits the downside.

Why is the Swing default look and feel so weak?

Have you had this conversation with the Java SE guys? What was their answer?

They told me that there was a new one in the works, and then I asked someone else and they said "really?"

[laughs] So, both those people were telling the truth.

In any company Sun's size, in the investments we make around the Java platform, we have to place bets. And I think when the question was answered the first time, they were referring to a project that we had to really reinvigorate the desktop.

With the evolution of JavaFX, I think you've seen a lot of that work go on the backburner. We believe that JavaFXScript is actually a great way to extend the Java front-end. Does that mean we will never go back and fix some of the issues people have brought up? The answer is no.

I think what we are seeing is an evolution where JavaFXScript and Swing co-inhabit in front-end apps, and that's going to bring a lot of capability. Some of the things that Swing was really poor at, have been replaced by JavaFX.

Look and feel is, well, I'll leave that one alone.

About Chris Duckett

Some would say that it is a long way from software engineering to journalism, others would correctly argue that it is a mere 10 metres according to the floor plan.During his first five years with CBS Interactive, Chris started his journalistic advent...

Editor's Picks

Free Newsletters, In your Inbox