James Powell’s Blog

After a very prolonged review by the EU Oracle has finally completed it’s acquisition of Sun and spent time on Wednesday outlining it’s plans for the combined company.

I must admit I was skeptical about Oracle’s plans when I first heard of the deal. Not that I knew what Sun should do but I doubted that Oracle would keep the hardware business, was worried about the future of Solaris and believed that the deal was all about customer acquisition and of course Java.

However having read about their published plans on Wed I think Oracle really understands what we (Thomson Reuters) need from our biggest technology partners. We’ve been talking about best of breed and heterogeneous systems for so long that we’ve missed just how expensive it is to do that integration.

I think being able to buy a fully integrated system (essentially “buying architecture”) and accepting the inevitable compromises associated with such a choice are key to they way we need to apply technology to solve our problems in the future. The recently announced partnership between HP & Microsoft is focussed on exactly the same issue.

There will always be special cases where we can and must build our own technology but being smarter about what we buy and driving leverage from those choices will be key to our success. We need to focus on being less special and making sure we’re well positioned to leverage the innovation and integration that our partners can provide (”skating to where the puck will be”).

I’m not yet convinced that Oracle will be successful with the Sun hardware (projects were cancelled and staff left while the deal was closing) but it’s encouraging that Oracle, HP & Cisco all seem to get this.

I was lucky enough to be invited to the launch of the iPad this week. I’ve always wanted to see one of these productions (as a friend put it - turning corporate America into a rock show) so even though with real time commentary freely available I made my way to San Francisco.

I won’t go into the usual review here of the new product - that’s been covered (and recovered) all over the web. Given the hype and interest it was almost inevitable that the gadget wouldn’t match up to people’s expectations. I loved the idea that the tablet was going to be more than just a better way to “read the web on the toilet” but as it turned out a bigger iPod touch is going to be perfect for doing just that.

The performance certainly impressed me (the decision to build their own proprietary silicon seems to have paid off) and I’m sure I’ll end up buying one at some point although CES had numerous similar devices from a hardware perspective.

However to me the interesting questions remaining (now we know what the hardware looks like) are all about the software. In some ways Apple is starting to make us rethink what a software platform looks like. As the device grows in size and performance beyond the iPhone, Apple is challenging many of our standard preconceptions about how we interact with a computer.

Some of these innovative ideas were in Apple’s previous attempt - the Newton (I still have a working one) where data was stored in a “soup” with associated metadata. It was a truly new way of thinking about data and as the iPhone/iPad platform evolves I expect to see a lot more challenging ideas to the old overlapping windows/hierarchical file systems way of thinking.

With the iPad there is no traditional file system - each app is sand boxed or there is also a shared folder. It’s also single tasking (for now) representing a real challenge for those of us with ADD. Apps have to be designed to adapt to changing device orientation and of course work in a “touch enabled” way. The new iWork suite demonstrates what’s possible and I believe ThomsonReuters will be able to build truly compelling products for our professional clients.

It’s also clearly a cloud based device - leveraging internet based capabilities just like Google’s Chrome OS. Many of the apps available on the iPhone today heavily leverage cloud based services. As we start to come to terms with this new platform - traditional platforms are going to evolve in a similar way.

Before long we’ll be out of the trough of disillusionment and start to truly understand what the device is capable off. It’s easy to get caught up with the hype and perhaps not understand that the device was never targeted at the usual high-tech early adopter crowd. We haven’t seen a significant change in how we interact with computers since the original Mac/Lisa/Alto - this could be the beginning of such a change.

Last week Google announced that they are building a new operating system that will leverage their Chrome browser and the Linux kernel.

There is plenty of discussion about why Google choose to announce this now - over a year from when it is expected to ship but I think this platform has some interest for TR. As an preview of how we’ll perhaps build applications in future and also as an indication of Google’s plans as they continue their quest to index the world’s information.

Choosing the platforms we build our applications on is sometimes straightforward (IE was dominant in the enterprise space for a long time), sometimes risky (Reuters’ decision to build a terminal on Windows 1.0 certainly caused me a lot of pain :-)), and sometimes complex (which mobile platform should we choose - Apple, RIM, Windows Mobile, Android, Symbian etc) without obvious answers.

Some of the critical issues are cost & how widely the platform is distributed. Chrome OS will be free but it’s going to be a long time before it or Chrome (currently standing at 2% market share) will be widely deployed within our clients.

Another key factor in our choice of application platform is how much leverage we get from it. How productive does it let our developers be? How much open source or 3rd party provided components & code is available that we can license and use? How easy is it to find developer talent? This has become a very fertile area for technology innovation (and recently VC investment) - whether it’s Ruby on Rails, Hadoop, or force.com, each approach has it’s believers (zealots) and of course much depends on what problem you’re trying to solve.

ChromeOS is interesting to me because it’s another example of “Web as OS” or Network as the platform. This has recently been adopted by Palm for their WebOS used in the Palm Pre phone.

Chrome is a modern browser - with a key focus on performance, security and support for Web Standards such as HTML, Javascript. By building an OS around it, Google lets developers leverage Web Standards to build applications whilst exposing key hardware features through non-portable web-based apis. There is obviously a lot of potential leverage - it may not be suitable for everything but with the capabilities of HTML 5 (such as offline storage and canvases) you can build impressive rich applications as was recently shown by the Google Wave demos.

By choosing this platform you get a standards based approach which should be deployable across a broad range of browsers and with a great price! I’m sure we’ll see an interesting range of devices appearing in 2010. Although I expect to see more & more applications built in this way I still expect “thick” clients to be around for a long time to come. I don’t see Chrome getting a big footprint in our client base soon and HTML 5 is facing competition from a pervasive competitor (Adobe Air based on Flash) and a rich environment from the incumbent (Microsoft’s Silverlight).

The other angle I find interesting is that by choosing to build an OS (actually their second after Android) Google are clearly eager to own (and critically instrument) more of the user experience. I believe that Google is willing to invest in these long term expensive projects because if successful they will give critical insight into how users are interacting with the web. This is of course motivated by their desire to deliver contextual (i.e. more valuable) advertising but I think it’s also very relevant to our mission to deliver intelligent information. There are lots of similarities between computational advertising and our need to deliver relevant contextual information. As such I think ChromeOS and Android are something we should keep a close eye on in the future. They’re not yet mainstream enough but could one day offer a compelling platform for intelligent information.

After a week of meetings back in NY I’ve had a chance to reflect on the D7 conference.

The core theme as announced by Walt Mosburg was “Web 3.0″. I’m not a big fan of this approach to naming waves of change but “Web 2.0″ has certainly weathered over time to be a generally understood term - user generated content, social networks, collaborative filtering etc so I guess they work at some level. If anyone had asked me before the event I would have assumed Walt was referring to the “Semantic Web” however he redefined it to be the trend around cloud based services coupled with lightweight mobile clients like smartphones or netbooks. As he put it - “it’s all about the applications”.

There is certainly a huge amount of innovation going on - particularly in the mobile space with Apple, Google, Blackberry, Nokia and now Palm all shipping operating systems, new devices and now application stores but I think the really interesting way to look at this is not about the applications at all.

One of the remarkable things about the internet and particularly the impact of move to digital information is the opportunity to reach remarkable economies of scale. A lot of the core Web 2.0 attributes need scale to be truly effective. If you’re trying to build an Web 2.0 style service - you not only have the challenge of building for scale but actually achieving scale. With the proliferation of devices & channels this is a real challenge.

I think a recent interesting innovation is the focus on platforms which can be widely leveraged across a range of devices and applications. Certainly standards such as web services have made it possible but it’s an interesting question as to what drove this. The mobile space is a difficult place to navigate for anyone right now. There is a vast variety of operating systems and devices not to mention the complexity of the network operators. Making the right choice with limited budgets and people is very hard.

Building platforms has always been a smart way to driving innovative use of your content or services - you can’t predict where the innovation will happen and as Bill Joy said - “No matter who you are, most of the smartest people work for someone else.” By building an open platform and enlisting others to build to it you can almost guarantee that people will come up with new ways to leverage what you’re doing. Of course it’s non trivial to build a compelling platform and then persuade people to build on top of it but if you can get that symbiotic process started remarkable things can happen. Witness the explosion of applications (almost 50,000 now) for the iPhone/iTouch.

But perhaps the focus on platform is driven by the need to get maximum coverage and not focus on a particular device or OS? Perhaps there are simply too many channels, too many devices, too many operating systems when you’re thinking about launching a service which requires scale to be successful. Sure you can use an abstraction layer such as Java or Flash but you end up delivering a lowest common denominator product - one which doesn’t leverage the features of the user’s precious device and just doesn’t look or feel quite right. If that’s the case then maybe building a open platform is the only way to drive the broad adoption without the need to make prescient bets on which devices or OS will win. If you build the platform using standards based APIs then you’ll allow your cloud based services to be wrapped in anyone’s favorite language and leveraged to deliver all kinds of interesting applications. If you’re smart you’ll take feedback and allow those same developers to drive the evolution of your api.

Take a look at the sheer proliferation of Twitter Apps on the Apple Application Store now. What’s remarkable to me is not just the number but the diversity of UI and creative ideas. The speed of innovation is fast with niche applications as the mainstream apps become over-featured. Of course that works both ways - one of the things about building a platform is if you are truly are open you can’t predict how that platform will get used.

The other big theme at D7 was the challenges facing traditional media. The consensus is that advertising won’t be enough to support the media as we make the transition from “analog dollars to digital dimes”. While Google is still making it work two of the most interesting platforms (twitter & facebook) continue to wrestle with how to translate a successful platform to revenues. There are plenty of ideas and it’ll be interesting to see how things work out.

While we were at D7 the latest in a series of platforms from Google was announced - Google Wave. It has all the right approaches to drive scale adoption - standards based, well defined protocols for interoperability and of course an eager developer community to innovate on top of it. No matter what drives this platform based approach I look forward to seeing the applications that will be available on it.

At the opening evening of the conference we got the chance to see the Twitter (everyone’s favorite asymmetric pub/sub publishing system) guys interviewed by Walt Mosberg and Kara Swisher.

Walt had made the point that D7 was focused on the Apps - in other words after the success of Apple’s App Store everyone sees this as the future. Android (Google), Blackberry & Nokia have all followed - with Microsoft to follow with their new Zune HD.

The conversation with the Twitter guys was interesting - mainly because they aren’t focused on the apps. They are instead building a platform - and have allowed hundreds of apps to bloom on top of of it. Sure they have a website but more than double the traffic comes from 3rd party applications.

I’ve always been a big believer of a platform strategy since my days at TIBCO & Teknekron where I saw the power of a set of APIS and was constantly amazed at how this platform could be leveraged.

We already bring platform thinking to our products - Common Platform and Novus to name a couple. As the number of channels for our content proliferates and the need to leverage context to provide better filters I expect a platform strategy will be a key part of our customer proposition.

This interesting article on the impact of more complex search queries suggests that the model of paying for words is starting to break down as users get more sophisticated in their use of search engines.

I’m not sure I agree about the queries becoming a conversation but it does show people are getting smarter about their use of this technology. It also seems to bode well for user interfaces which are driven by search (so called intent based computing). The more clues the user is willing to give us (beyond the inferred context which we build over time and experience) the better we should be able to do in delivering the information the users need at the right time.

Thomson Reuters is focussed on delivering “Intelligent Information”. One of the ways we can do this is by providing ways to see the patterns in the ever increasing amount of information we all have to filter.

This visualization caught my eye -

I’ve been interested in the debate about how social software could extend your ability to connect (increasing your Dunbar number) or if it is just a tool which allows us to be better at being connected. Either way this is a fascinating instrumentation of how people interacted with a live event.

One day I believe we’ll use social media such as Twitter, Facebook and IM to filter information in a truly contextual way.

I’m not sure if this is a good thing but it’s truly amazing.

I’m flying west on American Airlines to TED with wi-fi via gogoinflight.

Not bad connectivity either.

Even the Slingbox worked

I’ve always loved the movies and despite ipods and home theatre systems still love to go to a theatre when I can. Many of the movies I’ve seen over the last year have sadly been on airplanes (usually via an iPhone screen) and I can’t claim to be up to date on the current oscar nominations (I’ve only seen Slumdog which I loved).

I really liked this coverage of the top 10 shots from 2007 movies last year and was glad to see it repeated.

My personal favorite would have to be his #2 - maybe because of my love of architecture and/or London. Battersea Power Station was instantly recognizable in the movie.

(via Kottke).