Tuesday, December 27, 2005

At home in Liverpool

I spent Christmas at my parents' home in Liverpool. It was good to see my parents, sister, brother and cousin again and I had a pleasant and relaxing time.

My cousin is visiting Europe and working in Ireland before returning to Australia sometime next year. He works in the mining sector and it was fascinating to hear about the effect that high commodity prices are having on areas such as Western Australia - abandoned mines that I visited when I was there ten years ago are now economic again.

I am always amazed at how much Liverpool changes from one visit to the next. Antony Gormley's "Another Place" sculptures litter the beachfront by my parents' house (indeed, you can see a few from their bedroom window) and the effect of 100 identical life-size statues strewn across a large beach is very impressive.

The centre of Liverpool is also in full-scale redevelopment mode... the tell-tale outline of cranes was visible in the drive from John Lennon Airport to home. I discovered by accident that VLM have an offer on at present: my ticket will allow me to use the airport's lounge free of charge on my return trip - I'm not expecting anything like Virgin's Heathrow Clubhouse but it will make a nice change from the litter-strewn departure area that I encoutered last time I used LPL.

I met up with an old school friend and his girlfriend last night. They are working abroad at present and both trying to find roles in London so they can move back to the UK. I asked what it would take to persuade them to move back to Liverpool and work here instead. We concluded that we couldn't think of any scenarios where we would do that.

This isn't the fault of Liverpool - or any other city in the country (indeed, both he and I have spent time living and working on the South Coast). Rather, it's a realisation that the sheer size of London's economy means that it would be madness to live or work anywhere else. There is more competition for skilled labour there, the salaries are higher and there is more to do.

Those who know me may find it strange, therefore, that I still work for the company I worked for when I lived in Southampton. That's not really contradictory - the mere fact that I live in London is enough to keep my options open. And, contrary to what I may sometimes suggest, I actually quite like my work.

Wednesday, December 14, 2005

We don't all work for Global Services

Microsoft's Rich Turner is being playful again...

He is responding to an email from James Governor about one of his previous articles.

It's a long article and he makes some bold assertions.... I'll deal with a few of the more extravagant claims :-) (from my own perspective, as always.... none of this is an official response, etc, etc)

Firstly, he believes that the forthcoming WebSphere Enterprise Service Bus is essentially a triumph of marketing over technical innovation. I agree with him that our marketing team have done a stellar job of late. For what seems like the first time in a long time, we have managed to consistently describe our offerings, relate them to each other and to wider industry themes - and do it in a way that addresses our clients' real issues. However, it's a little unfair to imply that Kareem's team are also developing the product in their spare time. They're good... but they're not that good!

Rather, there is a lot of real innovation in there that will make this a compelling proposition to a large number of clients. Take one example: this will be the first ESB product in the market to deliver Service Component Architecture (SCA) functionality. *

Secondly, we haven't renamed Message Broker to "Advanced ESB". Of course, we did tinker with the product's name (it's now called "WebSphere Message Broker"). That seems to be something of an annual ritual for this product... I think it's had five different names in the five years I've been at IBM. However, we did do something important in this area in relation to the concept of an ESB. What Rich has noticed is that we now explicitly talk about broker in terms of ESB. To my mind, this is just pragmatic. Like it or not (and I happen to like it), the concept of an ESB is very important right now and we would be doing a dis-service to our existing customers if we didn't make it clear that they're on the right track (for they are) and we would confuse potential customers if we didn't explain to them that they can build an ESB on this technology. In other words, if everybody has settled on a name for a concept, it would be somewhat obstinate of me if I refused to explain how my stuff related. So the bottom line here is that message broker's name hasn't changed in any material sense but we do now make a point of emphasizing how well it allows an advanced enterprise service bus to be constructed.

Finally, Rich makes the mistake of assuming every consultant in IBM works for Global Services. IBM Global Services is a fine organisation but it isn't the case that we all work for them. I, for example, work for our Software Group - in IBM Software Services for WebSphere. I am measured on client success.... successful projects.... "enablement" (a fancy word for how well I help clients become self-sufficient), etc, etc. A very similar measurement model to MCS by the sounds of it. Indeed, if I spend more than a few weeks at a single client, it would be unusual. We, like Microsoft, simply want our clients to be successful with our products. Of course we partner with Global Services where appropriate - that's good for us and good for our clients. But to claim that the world's second biggest software company (which is what we are) exists solely to cross-sell services hours is absurd.



* It's not the first product in any class - our WebSphere Process Server wins that prize

Tuesday, December 13, 2005

Emotion and politics

Love the final line in Ed's piece here....

Monday, December 12, 2005

IBM Systems Journal on SOA

A colleague (thanks Isabelle!) sent me a link to the current issue of the IBM Systems Journal. This issue is completely dedicated to SOA. I haven't read any of the articles yet but some of the author names suggest it should be a very interesting read and provide useful pointers on where we're going. Peter Niblett, Beth Hutchison, Peter Lambros, Rob Phippen are just some of the senior technical leaders at Hursley who have contributed.

[EDIT: minor typo]

One down, more to go....

No.... it's not another "Heathrow should be demolished" diatribe. Rather, it's a celebration and time to rejoice.

IBM UK's very own Andy Piper is joining the conversation and you can read his musings here. Andy is particularly well connected within IBM's global community of internal bloggers and I'm sure his positive style and inclusive attitude will encourage many more of them to join this side too (it's not an either/or, after all)

A deep WebSphere Message Broker expert (among other things), he has been one of our key thinkers in the ESB space and I look forward to his contributions in this area. He has also promised to treat us to selected highlights of his past output. And if that wasn't enough to get you over to his blog (----->http://thelostoutpost.blogspot.com/<-----), he also links to some other IBMers who I was previously unaware of. So, what are you waiting for?!

Sunday, December 11, 2005

Famous people in our midst

It has been a source of disappointment to me that, in almost two years of London living, I have encountered almost no famous people at all. Indeed, with the exception of Jeffrey Archer, I have encountered none.

That changed this evening. A few friends took me along to "Rebel Rebel" - where I encountered not only T4's Simon Amstell but also Neil Tennant and Kele from Bloc Party (well.... "encounter" is a strong word. "Observed from a distance" would probably be more accurate). I should go to alternative college Indie nights more often, it would seem.

Of course, going out on an inadvertent celebrity hunt meant that I didn't go to Scoble's Geek Dinner..... a shame since there were several people on the sign-up list that I would have liked to meet. There's always next time I guess....

Saturday, December 10, 2005

Celebrity Columnists

I'm not sure if Tim Harford counts as a celebrity columnist or not.... but the sudden recollection that his two excellent columns would be in the FT magazine was sufficient for me to put down my copy of the Guardian and pick up the Financial Times in Tesco just now.

Sorry Alan.... I hope I didn't miss anything interesting in today's issue...

Of course, I was probably being completely irrational since most of Tim's content is online anyway but, still.... I've never experienced this before. It's rare that I'll buy a different newspaper because of the promise of one column.

Incidentally, my friend Polly is also a fan. This gives me confidence that my decision was correct as her taste is far superior to mine.

[Update 101205 1529] D'oh! Looks like I was right.... one of his two pieces is online.

Thursday, December 08, 2005

Right. This time I mean it.

Andy, Aidy, Brian, Darren and the rest of you Hursley BlogCentral lot..... get your asses outside the firewall!

Brian: Don't you just itch to counter Jonathan Schwarz?

Andy and Aidy: Are you going to take this lying down?! We can't leave Governor to stand up for us.... he's an analyst. They're fickle :-p I've written something here.... but I've read your internal post, Andy, and it blows mine out of the water...

[EDIT: oops... forgot about eightbar.... sorry Darren!]

It's not just an abstraction layer!

James Governor writes about the Service Component Architecture here. He also linked to Neil Ward-Dutton's take.

He is, of course, right to be sceptical... a major announcement like this should be critically assessed and the the technology evaluated on how well it solves real problems... It's our job to both argue for its merits and, more importantly, demonstrate that our clients are achieving real value from using it. As an IBM Software Services consultant, my job falls primarily into the second bucket. I wish I could blog about some of the exciting projects I'm seeing right now.

From reading the press coverage, however, I see two misconceptions appear again and again.

Misconception 1) "It's just a spec". Not quite... WebSphere Process Server is built on SCA and is shipping today.

Misconception 2) "It's a new abstraction layer". Not quite... it is more than that. I tried to articulate what I mean here but I'm not convinced I've done a good job. This point is very important so I'll try again soon.

Up in the air

I see that my esteemed colleague Ed Brill discovered the joy of Canary Wharf yesterday. Sadly, the description of his cabbie's behaviour is not as unusual as it should be :-(

However, London City Airport - and VLM airlines - are indeed two of the pleasures of living in London's Docklands. Being able to check in 20 minutes before takeoff - and never having to wait for any checked luggage make it my favourite place to fly to and from. The extension of the DLR to LCY is the icing on the cake.

As for that photo.... I think there's only one place it could have been taken from - but I guess it depends on your Point of View... :-p

Tuesday, December 06, 2005

ESBs...

Rich Turner takes a pop at WebSphere Message Broker here.

I think he's being a little playful in his implication that you can't build an ESB with broker because of its support for WebSphere MQ - but he does cover some important points.

Firstly, of course he's right that an ESB has to support open standards - how else are any new open-standards-based applications expected to take advantage of the services on offer via the bus? And in many cases, focussing primarily on open standards, web services, etc., is just what you want when implementing your ESB. Our forthcoming WebSphere ESB product provides just this functionality.

However, there are also large numbers of cases where you need support that goes beyond this. We offer WebSphere Message Broker as our recommended tool for delivering an ESB that integrates both standards and non-standards based applications. We can argue about where MQ sits in the standards spectrum but broker's deep support for MQ is obviously a huge plus.

As for whether adding SOAP support was a "stretch" for Broker's Hursley developers... I'm sure they'll take the suggestion good-naturedly as testament to their technical prowess and of the superior architecture of the product :-p


Incidentally, my first job in IBM was testing Message Broker (before it was called that...). I'm still officially based in Hursley but now live in London. I wonder if that means I count as a Hursley Blogger. James?

[EDIT 071205 1502 to fix link]

Amen!

I first discovered Microsoft's Rich Turner when I listened to a PodCast about an Indigo roadshow he was running. I'd like to say he stuck out because of his clear articulation of the technology or his clear passion for his work (although both were the case) but he first dragged me out of my early-morning-commute-stupor because of his familiar-sounding British accent. At least I *think* it was him.... if it wasn't, I have no idea at all how I came to find his blog. Anyhow....

I subscribed to his blog shortly after hearing the podcast and have given it passing attention since. Of late, however, he has been catching my attention increasingly often.

Take this. I suspect he and I may disagree over quite how far web services should be deployed and over how suitable they are for various applications but his core argument is spot on: if performance is the only thing holding you back from using a SOAP/HTTP web service (meaning that you're satisfied with the qualities of service offered and all other aspects of the decision) then do some tests - realistic ones. You may be surprised to find your assumptions challenged.


Monday, December 05, 2005

IBM Software Group Analyst Conference

The news articles are beginning to trickle out about IBM's annual analyst conference. Here's Bloor's Phil Howard's initial take.

The annual analyst event is not the kind of event to which mere mortals like me are invited. So, I relive it vicariously through the writeups...

From what I've read so far, it's heartening that there's nothing surprising here.... everything I'm reading (e.g. in Phil's initial writeup) is consistent both with the message I hear internally - and with what I am telling my clients. Phew!

One thing did surprise me slightly, however: Phil thought that our lack of discussion of web services in the context of SOA was noteworthy. He is correct, of course, to justify this in the way he did - SOA and Web Services are not synonymous and you can have one without the other - but it suggests he believes his target audience need to be told this.

Incidentally, Joe McKendrick linked to a good stab at defining SOA last week. I need to check how close it is to our definition but it looks promising to me. And, of course, it doesn't mention web services.

Loosely Coupled has an interesting story today about a meeting organised by J.P. Rangaswami, the CIO of Dresdner Kleinwork Wasserstein, in London last week.

Rangaswami's over-riding theme is the need to settle on a vendor-neutral, generic architecture for their companies and their applications.... I couldn't agree more.... I'm looking forward to hearing more about this event. We vendors can make a lot of noise in the market place - an initiative such as this... where the consumers of our technologies unite into one voice... is a rare and valuable thing.

Business State Machines

One of the new features in WebSphere Process Server is support for the concept of a "Business State Machine". The concept of a state machine is hardly new but this is the first time I've seen an engine implement them in the context of a business process.

The idea is that many business processes are inherently event-driven and are not linear. For example, the processing of a mortgage application follows a pretty rigid set of steps. If you want to model your process with a view to automating it, it makes sense to draw it as a logical sequence of steps.

However, some business processes are not like this at all. I saw a colleague use the great example of a DVD-rental store today. If we treat the lifecycle of a DVD rental as a process, then we quickly see that the process is entirely driven by events - the hiring of a DVD, the return of a DVD, the expiry of a rental period, the making of a reservation, etc, etc.

Until now, modelling such a process with standard business modeling tools was difficult.

WebSphere Process Server allows you, instead, to model the process in a far more natural manner and the tooling and runtime work together to take care of the execution for you.

It's an area I'm still investigating but I can think of several client engagements from the past that would have been a lot easier with this technology.

Friday, December 02, 2005

It is not enough for information to be available...

... it's only valuable to you if you consume it.

When I started thinking about making equity investments a few months ago, I struggled with the asymmetry I faced in terms of information. Professional investors had access to all the information I did (i.e. the internet) plus their proprietary information feeds plus the informal information they could pick up due to their position.

I therefore reasoned that I stood the best chance of doing well in areas where I could discover useful information (legally) for myself.

So, I chose to concentrate a large portion of my pot in retail. Why retail? Well, I am already heavily exposed to IT (all my human capital is invested there for a start...) so that wasn't an option if I wanted some diversification. I chose retail as the second best as it's the industry I probably have most exposure to.... I visit shops every day... I have friends who have worked for retailers at various levels and it's a simple business to understand. I figured that my own experience as a shopper would give me a good handle on which retailers were doing well and which were on the slide long before their profit figures showed a problem.

However, my best investment turns out to have been one of my smaller ones. It was a bet I placed on a Japanese Market index. Why did that perform so well? A lot of it was, of course, luck.

But I have belatedly realised that there is a second reason. My initial theory... that I would do best when I had access to information that others don't have... wasn't the whole story. It wasn't that I needed proprietary information... it was that I just needed it before most other people.

It's obvious in hindsight but I had been assuming that information travels - and is consumed - instantaneously. This is, of course, not the case.

My Japanese investment is an example of this: The Economist reported the opportunities in Japan several weeks before the mainstream newspapers started telling their readers to invest there. Those few weeks were when I made a lot of my gains.

Now... this case was good luck. But it hits on an important point. I am a very quick reader and can extract the key points from a mass of text with a skim read. This allows me to invest a small amount of time each day skimming internal IBM blogs, newsgroups and intranet sites to gain a disproportionately large amount of information to help me with my job.

Looking externally, the same is true... the economics, business, technical and news sites I read consume less of my time than one would think. My question is: does this give me an advantage that I can exploit - either for my investments or for my career?

When we are asked to think about our strengths, "reading comprehension" isn't usually at the top of anybody's list but perhaps it really can be used as a strength. I think it certainly gives me an information advantage in multiple situations... presumably this is true for lots of other people... I wonder if there are techniques one can use (or career paths to follow) to get the maximum benefit from this ability?


Why Gendal World?

I don't think I ever explained why this blog is called "Gendal World".

No... it's not some horrific (and misspelled) "Lord of the Rings" reference. It's actually very simple. Gendal, for reasons that my parents have never satisfactorily explained, is my middle name.

As a child, I was told that Gendal was the name of a Welsh great uncle. When several Welsh people told me they had never heard the word before, my parents said that perhaps it is Cornish.

Who knows.....

It has several advantages..... for one, I can almost always get the user id I want on any given system (except for DNS... I used to own www.gendal.com but let it lapse...)

Thursday, December 01, 2005

The Deployment Descriptor on Steroids...

Well.... the thing that I had been hoping to see for some time now happened yesterday... Oracle, IBM, Sybase, Siebel, BEA, SAP and IONA announced support for the Service Component Architecture (SCA). SCA is at the heart of WebSphere Process Server and so I've been exposed to this technology for some time now.

Why do we need SCA? What problem does it solve? What's wrong with WS-*? JAX-RPC? J2EE? etc, etc, etc.

The links above are worth looking at for the "official answer" - and the reasons are good. However, the motivation for SCA that I give is usually as follows. I'm not claiming this is the real reason it was invented or that this is necessarily the most important aspect but I've found it useful for my own internal thinking.

So here goes....

Let's imagine we're starting a new development project. We've heard all about SOA... perhaps there is even an existing SOA initiative in our organisation. Perhaps several key services have already been exposed (and advertised) as reusable services. Perhaps there is even some governance in place around them. Excellent.

We start to design our system and we realise we can use these services in our application. That is: we will be building a composite application that contains both new code and invocations of existing code... we will orchestrate the execution of services hosted elsewhere if you like.

Two immediate technical questions arise. 1) How do we invoke these services? 2) How do we build our own application such that it isn't monolithic but can, in turn, provide services to other users in the enterprise?

Today, the answer to the first question will vary based on who you speak to. But the answer usually comes down to some API (JAX-RPC? JMS? EJB calls? whatever...) and some meta-language for describing the service (WSDL?). This is just fine... as long as you're happy to treat these services as end-points.

But what happens when we look at our own piece of development work? It is likely that we will be trying to build a modular solution which consists of various "chunks" of functionality and we may even wish to regard these pieces of functionality as services in themselves. That is: we will use this new functionality but it's sufficiently coarse-grained to make parts of it suitable for exposing to other users in time.

The big problem, however, is that these components are not just end-points.... they are components that provide both a service and which have dependencies. This is a problem since we're writing this application! It's up to us to satisfy these dependencies.

What SCA does is allow us to write our applications as we always have done if we choose (Session Beans, chunks of java, whatever makes sense for each component) and attach a descriptor to each component in the application which contains several pieces of information: the interface provided by this component and the interfaces it needs to call in order to perform its work. This descriptor contains other information but these two aspects (the interfaces and references) are the key.

They are key because it means we now have a consistent way to build and package components that provide and consume services. Even better, since we now have a useful descriptor for each component, we can create tooling to facilitate the building of such applications - the descriptor (the ".component file") tells the tooling all it needs to know to allow users to "wire" components together.

Now, a secondary benefit of this is that we now have a unified API for invoking services (a component merely says to the runtime "give me the service which implements this interface that I am dependent on - and wired to" and the runtime is responsible for taking care of the invocation, regardless of the actual technology in use). However, this benefit really is secondary.

So, to my mind, the big step forward with SCA is not the "extra layer" or the "unified API". Rather, it's the fact that we now have way of describing components and for wiring them together to build composite applications that is consistent whether we're programming "in the large" or "in the small" - and which extends service-orientation from the end-points into the heart of an application in a unified manner.

Of course, not everybody seems to think this is a good idea :-)