Tag Archives: WebSphere

Geekery in 8-bits and more

In which I get misty-eyed and nostalgic, geek out over electronics, and think about mobile and the cloud.


On Saturday I went along to the Horizons 30th anniversary of the ZX Spectrum event, organised by Paul Squires and Leila Johnston and held at the BFI in London. The event ran on both days but I wasn’t able to stay on the Sunday, so I missed at least half of the fun!

Steven Goodwin reads Sinclair User

Although I’m full of nostalgia for the 8-bit era, I have to confess I never actually owned a Speccy or any Sinclair hardware. My friends did, but I was primarily an Acorn enthusiast and our first home computer was an Electron (although the first computer I used at primary school was a Commodore PET).

I fondly remember some of the hacks I did on/with/to the Electron, including soldering a pair of headphones into the motherboard to avoid annoying my parents with the music from various Superior Software titles 🙂

Regardless of “allegiance”, Horizons was a really great day. Highlights for me included a fantastic history of computing by PJ Evans from The National Museum of Computing at Bletchley Park (if you haven’t been there yet, you should visit!); Spectranet, an Ethernet adapter for the Spectrum which had me wanting one for no good excuse that I can come up with; and the mind-blowing live composition of a chip tune by Matt Westcott which I saw, but I struggled to comprehend. Matt’s ability to reverse engineer a tune in his head was remarkable.

Oh, and if you haven’t downloaded or bought MJ Hibbett‘s Hey Hey 16k yet, or at least streamed it, you really should.

aside: since Horizons was part of SciFi London, I tried to get Micro Men director Saul Metzstein to drop some hints about his upcoming S7 Dr Who episodes. All he would say was that the western episodes were filmed in Spain (knew that), and that the script for the Christmas episode hasn’t been written yet (didn’t know that).



After the event on Saturday evening, I found it a real struggle to avoid crazy, nostalgia-fuelled eBay purchases, but I did manage to resist! Instead, I resolved to finally get around to building the Fignition I’d picked up at the Hack to the Future event a couple of months ago.

For those who are not familiar with it, the Fignition is a credit card sized build-it-yourself 8-bit computer based around the ATMega chip (the same one used in the Arduino and Nanode Open Source hardware boards). It’s really a remarkable little device – I guess it took me about an hour to assemble and solder, although your mileage may vary. The build guide is excellent and very clear. After performing a couple of power on tests with and without the ICs inserted, it was time to connect up to the TV – and it worked first time. It boots into a simplified Forth environment, which was reminiscent of that BBC BASIC> prompt I am so familiar with from my childhood. The only real downside is that the keyboard – built from 8 clicker buttons – is a bit fiddly to get to grips with, but hey – I just assembled a complete 8-bit computer including video out and keyboard! It’s hard not to be excited.

The board I built was a RevD – the new RevE board has onboard audio in/out (get ready for some fun loading stuff from audio cassettes, again!), and is also slightly modified so that in principle, it is possible to add Arduino-footprint shields. That’s kind of cool, as it means that it might be possible to add a PS/2 keyboard or a network interface.

Ready to test!

What’s “the point” of something so simple, by today’s standards? Well, actually – the simplicity. I went from a bag of components, to a fully working computer in the palm of my hand – no surface-mount components – to a programmable device. It’s “primitive” by the standards of today’s machines, but it’s not that hard to understand how an 8-bit “brain” works, in comparison to the 32 or 64-bit mulitcore CPUs and GPUs in modern laptops and mobile phones. In my opinion, the Fignition, Arduino and Nanode fulfil an important role in helping youngsters to understand the basic principles of electronics and computing.


Last night I headed along to the fantastic Mozilla offices in London.

Mozilla Space, London

The main LJC event was Simon Maple from IBM showing off the new WebSphere 8.5 Liberty Profile running on a Raspberry Pi. I’d hooked Simon up with Sukkin Pang recently so that he could get one of the smart enclosures he provides for the Pi. It was pretty cool to see a full Java app server running on such a small computer – actually almost exactly the same size as the Fignition, only considerably more powerful of course.

The whole talk was live streamed on Mozilla Air – but if you missed it, there’s a video available (complete with semi-professional heckling from yours truly!)

Boot 2 Gecko

What stole the evening for me, though, was two other glimpses of what lies ahead. First, Tom Banks from IBM Hursley came on stage after Simon and showed off the Liberty profile running on a mobile phone. Let me clarify – he was running Android 2.3 on a Nexus One (an “old” phone), running Ubuntu Linux as a virtual image inside of that, and WebSphere inside of that. Kind of mind-blowing! A proof-of-concept and arguably not very useful… not sure when I would want to put a full JEE app server in a phone… but extremely cool. Finally, @cyberdees let Tom and I have a play with Boot to Gecko – Mozilla’s new mobile play. B2G was something I’d heard about, but not touched. I have to say that even in an early form, it’s looking very slick, boots extremely fast – much more quickly than any Android or iOS device I’ve seen – and the device integration (GPS, camera, access to hardware settings, etc) was impressive.

With the Open Web as the platform, ubiquitous mobile devices, and increasingly sophisticated cloud-based backends to interact with, the future is looking pretty cool.


When “end of an era” doesn’t cover it

This week, I tendered my resignation at IBM, after 10 years and 4 months, to a manager who has been my team leader and friend for the past 3 years. I can honestly say that it was a really hard moment; but also the right moment to make this particular transition.

As I’ve repeatedly written over the past few years – IBM has been a company I always aspired to work for, and once I had the chance, one that I’ve been immensely proud to represent. It’s a company that has endured over a century, and one that I was able to spend time with for a tenth of its existence – it was really the age of both WebSphere and the rise of IBM Software Group, and I’m grateful for the opportunity to have been there.

I have brilliant memories of the past decade. IBM is an amazing company and I will always value the chance to be a part of it, particularly in a wonderful location like the Hursley Lab. The people I’ve worked with, and with whom I’ve formed what I believe will be enduring friendships, have been simply outstanding. There were so many opportunities to do great things, not only in “the day job” but also as a BlueIQ Ambassador and social collaboration advocate, with IBM developer communities, in the universities programme representing IBM at careers fairs and as a guest lecturer in degree programmes, and the schools and community programme as a BlueFusion volunteer and mentor to kids at schools in deprived areas. I’ve also loved the chances to learn from others formally and informally, and to act as a mentor to others.

This will sound like a total paean, but it’s very true that there are amazing talents around IBM. In 7 years in IBM Software Services, and more than 3 years representing the development, strategy and product management teams in the lab back out to the field, I amassed a list of friends and colleagues from across continents, business units, and brands. It’s amazing to think of the broad reach of my network and I can’t help but be grateful for that.

My next steps are still forming; but I’m looking forward to spending more time with Open Source communities, with developers, with new technology, with connected systems and the Internet of Things, and as a speaker and writer. I’m also grateful to a range of friends for their support, particularly in taking over initiatives like eightbar, and in enabling me to remain involved in strands like Eclipse and MQTT.

Thanks for following me, reading my blog, sharing my thoughts, and joining the journey. I hope what comes next will be a continuation of the path I’ve been on; and an exciting next step in developing the direction I’ve been headed in.

WebSphere Message Broker version 8 is out!

Hot on the heels of the latest goodness in WebSphere MQ, it’s the turn of IBM’s Enterprise Service Bus – WebSphere Message Broker – to get a major new update.

WMB v8.0 was announced back in early October and has just arrived ready for  download in versions for distributed platforms, System z mainframes, and as a Hypervisor Edition for Linux and AIX (to be provisioned via the IBM Workload Deployer appliance).

As I did with WMQ last month, I wanted to take a moment to break out and highlight some of the key things in this release that you may have missed from the announcement letter. This won’t be a comprehensive list of everything, but I  want to point out some of the cooler features that you’ll want to be aware of. So, here we go…

(I’ve included a few screenshots to whet your appetite, click for larger versions!)

A simpler development experience

Version 8 brings a number of enhancements to the development experience, but one worth highlighting is what we call “Apps and Libs” – the idea that sets of message flows may be grouped into a unit called an Application which can be deployed, stopped and started as a whole. With Libraries, there are also truly re-usable assets like .esql files, or sub-flows, which can be deployed and updated separately, and invoked dynamically at runtime. This is a key change in the way that the Broker works – previously, sub-flows were compiled into the main flow and changing one required redeployment of all flows using it… they are now dynamically linked when needed, so they can be deployed and replaced more easily.

A new standards-based parser and message modeler

A new Data Format Description Language (DFDL, which you’ll sometimes hear called “daffodil”) enables any text or binary data to be understood within the message model. The Broker has had the “MRM” for many years, so of course could already do this, but DFDL is a new industry standard which can supersede the MRM (of course, you can continue to use your existing flows and message formats – you’re not forced to use DFDL). There’s a new mapper, too.

More importantly, coming along with DFDL and the mapper is a really, really nice set of utilities for testing message models inside the Toolkit – you’ll now be able to confirm that the model matches the test data without having to go through a full model->deploy-> test-at-runtime cycle. I saw this demo’ed at the WebSphere Technical Conference in Berlin during October and was blown away by it – it would have saved me a lot of time back in my consulting days!

Comprehensive .NET support

If you have .NET applications, assemblies, or services on the Windows platform, and you want to access those from your message flows – you can. If you want to write your message flow logic using C# or VB.NET or any .NET 4.0 CLR-supported language, using Visual Studio – you can.

If you don’t know how to get started with this stuff, the Toolkit has a new .NET Pattern to lead you by the hand and get you going quickly, and project wizards for Visual Studio.So, if you want a high-performance ESB platform that connects “anything to anything”, with minimal need to learn new skills, and run it on Windows with deep .NET integration – this release is going to cover your requirements.

Web administration

Delivered in version 8 is a first stage in making the Broker more easy to administer from a lightweight client – a web browser. Whilst power users and existing administrators can continue to use the Message Broker Explorer GUI, there is now an easy way to enable an optional web interface for basic administration tasks. Continuing the theme of simplicity the product has followed for a while, no additional moving parts (app or web servers) are required! Version 8.0 provides read-only views of running Applications and access to the log – more capabilities will be rolled into this interface in the future.

Record and Replay

Sometimes, when you are dealing with a set of end-to-end flows of data between applications, you may want the capability to record what is going on, and to replay specific scenarios and sets of events. This could be the case in audit, test, and many other scenarios. Another of the massive enhancements in version 8 is the Broker’s response to this requirement – again delivered using the same simple, lightweight interface offered by the web administration tool.

This also builds on technology around monitoring that has been progressively built into the Broker over the past couple of releases, so there are some really solid foundations and it is straightforward to set up.

Richer, yet easier to use

Just as I highlighted in my piece about WebSphere MQ 7.1, the Hursley teams have been strongly focused on “consumability” (translation for non-IBM-speakers = UX) for a number of years now. WMB continues to add capabilities that make it a richer, stronger integration platform, but also smooths out rough edges seen in earlier releases and is just… well… more productive to use. There’s even a drive to reduce the jargon and make the Broker logs more easy to understand, with new Activity Logging which aims to explain what a flow is doing in plain language (“GET message queue X”, “Update DB table Z”, and so on).

Taken together, the new wizards, web interfaces, integrated testing tools, message modelling tools, reduced dependencies, lightweight deployment with apps and libs… the combination just makes it a much more enjoyable experience for developer and administrators. And there’s a new installer, too.

The “papercuts” and node additions lists are huge: new JMSReceive node; new options for the File nodes; new Connect:Direct nodes; WS-ReliableMessaging support in the SOAP nodes; ability to install without root privileges; dynamic configuration of services without the need to restart execution groups… the list just goes on! Check out the product Information Center for more details on all of the features I just don’t have space to list.

… and finally…

Huge congratulations to some hard-working development teams in Hursley, Toronto and Bangalore in getting this release out there. As I’ve said before, I’ve been using the Broker for 10 years now and it just keeps getting better, and better. These guys are a very strong set of developers who turn out a fantastic, high quality product every time. Special thanks to MGK, @mqmatt, and @domstorey for some of the screenshots in this post 🙂

Footnote: version 8.0 is friendly to developers who use Ubuntu, too! 🙂 Anton (my go-to guy on all things Debian – listen to him!) has some good advice about running WMB or WMQ on Ubuntu and Debian.


Photo post: IBM IMPACT 2011

I didn’t have a “proper” camera with me at IMPACT this year, but I did snap a bunch of images using my iPhone – they were either tweeted via yfrog, or through Instagram. Don’t expect the most awesome photography, but … Continue reading

Rate this:

Connectivity and Integration podcasts

As well as being WebSphere Messaging Community Lead out of IBM Hursley right now, I’m also part of what we refer to as our “Connectivity and Integration” organisation (middleware… plumbing… the hidden inter-application messaging and adapter stuff, ensuring that systems can talk to one another reliably). Much of what we do in Hursley, and the software that we develop there, is part of the Connectivity space. It’s the software that joins up all the pieces of a Smarter Planet, and it’s an interesting space for a techie like me.

We thought it was about time to talk about some of the features that are in our WebSphere Messaging products – WebSphere MQ, Message Broker, and the family of software that fits around them. So, my colleague Leif Davidsen and I sat down and recorded a series of podcasts. Each episode zeroes in on a specific feature or capability, such as high availability, or telemetry, or security – you get the idea.

As we were talking, Leif and I were trying to keep the discussions bite-sized (about 10 minutes at a time); highlight things that users might not have heard about before; be interesting to administrators and developers as well as to architects; and we tried not to use too much “marketingese” – although I reckon you might spot that in some of the podcast episode titles! 🙂

You can start to subscribe to the Connectivity and Integration podcast series right now in iTunes or add the RSS feed to your favourite podcatcher. There should be some web content and show notes with links and references to follow soon – watch out for those, I’ll tweet about them and update this post when I know more.

NB did you check out my first and second columns for Sphere yet? More to come soon, and I’m hoping to join the GWC Lab Chat series for a future episode as well. Cool stuff.