SOA: it's about people more than about technology

I took part in a panel session during the inaugural Oracle Enterprise Architects Club in London at the back end of last week, which focused on SOA. The Club is part of a wave of initiatives from the big IT vendors to get more engaged with that elusive group - "architects". This particular forum is about Oracle connecting with "architects" from its SI partners.

There were quite a few interesting questions but one in particular (from the near ubiquitous Steve "Java" Jones) stood out.

All the panellists (Miko Matsumura from Infravio and the OASIS SOA Blueprints TC; Steve Ross-Talbot from the W3C Web Services and WS Choreography WGs and various other places; and Anthony Reynolds from Oracle) had been talking about how the challenges of getting business value from SOA fundamentally concern how people (business and IT) work, rather than about implementing the enabling technology.

Steve's question was along these lines: "If all the big SOA challenges are about culture, organisation, etc, rather than about the enabling technology, what do you think this means for most technology vendors, who focus on selling products?". (He's followed this up on his blog).

Well, in answer to Steve and as a general point, this is hardly the first time that systems integrators and other consulting organisations have been the ones paid to take technology products and deliver real solutions - in any new technology-driven business investment wave (ERP, EAI are two recent examples) the capabilities of systems integrators have been the key to mainstream market development.

What's interesting in the context of SOA, of course, is that one of the primary underpinning ideas is to promote standardised approaches to technology delivery. When it comes to delivering the value of SOA initiatives there's therefore obviously interest in open approaches to intellectual property concerned with SOA implementation. In a world where suppliers are busy telling anyone who'll listen how open their technologies are, the same suppliers are naturally going to look a bit less-than-committed if the tools, methods and ideas they use for SOA implementation are proprietary and closed.

That's why the OASIS SOA Reference Model and SOA Adoption Blueprints TCs are so key to the long-term and mainstream acceptance of SOA. They're a reason to be positive about the the future of SOA, and I recommend anyone actively investigating SOA to check them out. posted by Neil Ward-Dutton @ 4:54 PM 0 comments permalink

Microsoft ups the ante with the Commission

On the day of the deadline originally imposed (Microsoft received a one month extension yesterday) in December's Statement of Objections from the European Commission concerning the usefulness of the technical documentation Microsoft had provided to help licensees of its Work Group Server Protocol Program (WSPP), the company announced that it is releasing the source code for Windows Server 2003 SP1 - Windows Server R2 is to be made available within a month - to licensees of the WSPP (in much the same way that the company has released source code to governments and other key customers under the Shared Source Initiative).

The announcement does not mean that Microsoft is open sourcing Windows. Rather, the company is making the code available under a reference license, allowing WSPP licensees to study and refer to the code in order to understand how the protocols covered in the 12,000 pages of documentation already released by Microsoft have been implemented. As Brad Smith, Microsoft's General Council put it, "The Windows Source code is the ultimate documentation of Windows Server technologies ... the DNA of the operating system". The fact that the source code is for reference purposes only is clearly demonstrated by the fact that the code will only be accessible via Microsoft's Code Center Premium (CCP), which provides smart-card-protected web based, read-only access to the source code. There is no offline use. The licensing of the source code also does not signal a change as far as Microsoft's view of distribution of its source code alongside open source products - it continues to emphasise a hybrid model - nor of its belief that the File & Print and User and Group Administration protocols which are at the heart of the EC's demands contain valuable intellectual property and innovation which warrants the royalties associated with WSPP.

In making this announcement Microsoft is clearly attempting to draw a line under the issue of documentation and focus, as Smith put it, on the "substance of the case" - not least because the European Court of First Instance will be hearing Microsoft's appeal against the EC's March 2004 decision in less than 3 months (April 24-28). Microsoft clearly believes, as Smith was at pains to point out, that things have moved on since the March 2004 decision and that the company is in a stronger position than two years. He noted the total lack of interest from consumers and OEMs alike in the Windows Media-less Windows XPN and pointed to the success of Apple with the iPod and iTunes as an indicator of the health of the competitive environment.

By going public with the announcement on the same day that it informed the Commission of its decision to do so, Microsoft is also making a point that it was dissatisfied with the EC's very public Statement of Objections in December.

Ultimately, then, I don't think this changes this as fundamentally as a headline "Microsoft Goes Beyond EU Decision by Offering Windows Source Code" would suggest. Rather, this is Microsoft upping the ante to eliminate the issue of documentation which it clearly believes is a minor irritant in the larger scheme of things - and it has nothing to do with the threat of 2 million Euro per day fines. The core issues of licensing of its intellectual property, particularly by the open source community, remain and will be the subject of continuing fervant debate with the EC, the Court of First Instance and more broadly. posted by Neil Macehiter @ 1:43 PM 0 comments permalink

Progress Software takes further action to bolster service infrastructure proposition

Progress Software today announced that it has agreed to acquire privately-held Actional for $32 million. Actional is one of the few pure-play web services management vendors left standing, with the likes of Adjoin, Confluent and Talking Blocks (amongst others) having been acquired by the major infrastructure and enterprise management vendors (CA, Oracle via Oblix and HP respectively).

By acquiring Actional, which merged with Westbridge Technology in October 2004, Progress adds web services management/monitoring and XML security capabilities to bolster the SOA infrastructure proposition of its Sonic enterprise service bus (as did the end-of-2005 acquisition of NEON Systems for legacy integration). This is essential for Progress if it is to build on the early momentum it has established in largely project/departmental level SOA deployments and move into more strategic, enterprise-scale deployments where it will come up against the likes of BEA, IBM and Oracle with their own ESB propositions. Actional, by virtue of its partnership with Microsoft, also strengthens Sonic's .NET interoperability story.

Progress certainly seems to be getting good value for its $32 million, reflecting the fact that many of the pure play web services/SOA specialists (and more importantly their investors) are struggling to grow their businesses in the face of competition from the major infrastructure and management vendors that are increasingly building or acquiring equivalent capabilities.

Whilst Progress may be getting good value for money, its not going to come without some effort. Although the integration of the management and security capabilities with the Sonic ESB should be comparatively straightforward, significant effort is going to be required to provide customers with a common policy/configuration framework for control of the integrated solution.

I can't help thinking this announcement was welcomed by the team over at Amberpoint. With yet another (and arguably the most significant) of its pure-play competitors acquired, it leaves more room for Amberpoint to play the best-of-breed card and, perhaps more importantly, to strike partnerships with SOA infrastructure vendors, such as BEA and Tibco, who are competitors to Sonic. posted by Neil Macehiter @ 4:52 PM 0 comments permalink

On OSS as prior art, tagging coding semantics, and folksonomies

This will probably be my most buzzword-filled blog post title of the year!

On Jan 10 IBM announced that in collaboration with the US Patent and Trademark Office (USPTO), and the global open source software development community, it is initiating three projects to "improve patent quality":

• the creation of an Open Patent Review process, which will allow experts in industry to contribute to the review of pending patent applications


• a project to make it easier to use the body of available Open Source Software as prior art in patent applications


• the creation of a Patent Quality Index, which will be used to help tune the patent approval process and (hopefully) incent people to draft better patent applications.

All of these three initiatives are interesting, but it's the second (the project to make OSS searchable as prior art) that I think is particularly so from a technology point of view.

The idea is that IBM will work with OSDL, Novell, Red Hat, SourceForge and others to develop a system that "stores source code in an electronically searchable format [that] satisfies legal requirements to qualify as prior art". The core of this scheme will be a taxonomy that can be used by developers, patent examiners and others to "describe and help locate relevant source code and documentation".

This is interesting for a number of reasons.

First off, it's a fascinating potential application of tagging. Of course annotation has been applied to source code for a long time (both automatically, in the compilation process; and latterly in languages like C#, Java and of course in vendor-specific SQL hints). But this proposal is about annotation of the semantics and patterns of the underlying design behind source code, by people on a grand scale. It's hugely ambitious.

There is a mass of questions which spring to mind but a couple of key ones are: who will create the taxonomy? What kind of process will be involved in its creation? As things like the Dewey Decimal System show, taxonomies can be great for "putting information in" to a system; but lousy at helping regular punters get information out.

Given the organic, open and participatory nature of the open source software lifecycle, it strikes me that a folksonomy might work better than a top-down imposed taxonomy. Although (as Clay Shirky points out) we are likely to have the kind of "expert catalogers and expert users" that make a formal ontology useable, what we don't have (I believe) is the kind of "stable and restricted entities, and clear edges" which would be required. Moreover we don't have "coordinated users".

However the challenge with a folksonomy is that the community-generated categorisation scheme – whatever it might be – would have to evolve to a relatively stable point before it could be useful in something like patent prior art searching.

Perhaps we need some professional tag gardeners to help chart a middle ground between the "folk" and the "tax"...? posted by Neil Ward-Dutton @ 10:14 PM 1 comments permalink

Some predictions for 2006

At around this time, it seems to be pretty much mandatory for IT analyst firms to offer predictions for the coming year. Although at MWD we like to try and do things a little differently, we couldn't stop ourselves. So below is our take on some of the trends, topics and technologies that we think will have an important bearing on our own area of interest – the challenge of IT-business alignment – throughout 2006.

[Note: this is quite a long entry.]

Application architecture will become more confusing, and more important

The shift in focus of automation within enterprises – away from automation of established, mature day-to-day operations, and towards supporting management and strategy-setting activities and those processes which reach outside the organization – is driving change in the ways in which people, processes and information have to interact. Change is occurring not only in the "process support" layer, via ideas like service oriented architecture (SOA) and new approaches to business process management (BPM). It's also occurring in the "information management" layer, via master data management (MDM) and information integration technologies; and in the "user experience" layer, via the emergence of a variety of "rich client" web-based technologies and the increasingly blurred lines between user interface and document format technologies.

These three changes will start to heal three fractures that have long leached value from enterprise IT: the fracture between the management of structured and unstructured information; the fracture between the automation of structured and unstructured business processes; and the fracture between the desktop and the web user experiences. At the heart of this fundamental "healing process" is one core technology: XML.

In SOA, there will be a much greater focus on business value

In 2005, SOA was a very hot topic – and it will continue to be so in 2006. However as organisations start to grapple with significant SOA deployments and go beyond pilot projects focusing on using service-oriented concepts to improve application integration, they will start to need much more help in ensuring that their initiatives aren't just cool, but deliver real business value. Too much industry focus has so far been on the "bottom up" aspects of SOA – principally, on standards and platforms for helping organisations use web services technologies to improve integration. Not enough attention has been paid to the "top down" aspects of SOA which really drive business value – how can a group of services best be designed so that they represent real business requirements? How can the lifecycles of services best be managed to ensure that business expectations are met? And so on.

2006 will see much more attention being paid to initiatives like the OASIS SOA Adoption Blueprints Technical Committee, which aim to bring together and formalise some of the SOA best practice knowledge that exists in the industry.

User experiences and information access will be key areas for innovation and disruption

2006 will see continued change in the ways that software functionality, structured and unstructured data are all presented to "users" (whether they be people or other programs). More and more mainstream software products and online services will take advantage of technology advances to offer semantically-rich yet open interfaces that blend the advantages of tagged document markup, rich graphics and interactivity, platform-neutrality and automated deployment to create experiences that erase the historical technology boundaries that have separated desktop applications and online services; and rich UIs and structured documents. The result will be challenging for a number of IT industry incumbents, not least Microsoft, which have built their franchises on being able to control the user experience and delivery channel for software applications.

We saw with the emergence of the web and its associated technologies – HTTP, XML, the browser – how it shaped the expectations of employees and customers alike, driving enterprises to invest in browser-based delivery mechanisms, enterprise search technology and so forth. As the web moves into its next phase – Web 2.0 anyone? – enterprises will come under similar pressure to deliver similar experiences. Many of their traditional suppliers are likely to be found wanting.

Suppliers will continue to shift away from traditional product licensing

2005 has seen a number of changes, in both technology and sourcing, which have challenged vendors' enterprise software licensing models. Virtualisation, multi-core processor architectures, open source subscription licensing and software-as-a-service delivery models have lead customers to question the applicability of the traditional, per CPU/user/server perpetual licensing model. In parallel, many IT organisations are looking to position themselves as service providers to their business stakeholders.

Vendors, such as IBM, Oracle and Microsoft, responded to the technology changes in 2005 with tweaks to their existing models. However, changes in sourcing strategies pose a far greater challenge. Microsoft's struggle to engender significant support for its Software Assurance subscription model is a case in point. The vast majority of enterprise software suppliers' business models and sales and marketing strategies have been finely tuned to the one-off, up-front license sale with annual support and maintenance. Vendors such as Sun, which admittedly have less to lose, have taken bold steps with their per-employee licensing model but this an exception, rather than the rule. IBM, with its Gluecode acquisition, and Microsoft, with Office Live, are taking tentative steps, particularly in terms of meeting the needs of the SMB segment, but this is only the start. 2006 will see enterprises of all sizes applying pressure on their suppliers to provide more predictable licensing models, which reflect value as it is delivered, rather than potential value when it is installed.

IT supplier acquisitions will continue at the same rate as in 2005

2005 has seen a number of mergers and acquisitions between mid-sized and large IT suppliers; and a whole host of acquisitions of small specialist suppliers. Throughout the year we've tracked Adobe-Macromedia; CA-Concord; BEA-Plumtree; Attachmate-WRQ; Oracle-Siebel; Sun-SeeBeyond; IBM-Micromuse; and many others.

2006 will deliver a similar number of invited marriages and surprise elopements. As the client-server and web-based application platforms and technologies on which today's IT giants made their fortunes passes into the territory of late adopters and laggards, those customers prefer the advantages of "one throat to choke" over the potential innovative power of smaller, quicker suppliers. Of course, at the same time, the emergence of a new web-based "software as service" application architecture (as described above) is catalysing a new wave of small innovators: many of these, too, will make tasty snacks for the likes of IBM, Microsoft, CA, Oracle, SAP, BEA, Compuware, Mercury and other billion dollar-plus suppliers.

Virtualisation will become more sophisticated and will deliver more strategic value

Virtualisation has been a significant focus for enterprises and public sector organisations in 2005. Whether explicitly, in terms of server and storage virtualisation from the likes of EMC, HP, IBM and Microsoft, or implicitly, in terms of service-oriented approaches to application development and integration, organisations have been turning to virtualisation in its various guises to grapple with the complexity and heterogeneity of their IT infrastructure. However, virtualisation remains a largely siloed proposition.

2006 will see vendors which span those silos begin to enunciate integrated virtualisation propositions, which position virtualisation as an architectural approach to the delivery of IT services. To have any credence, these propositions will have to be backed up with architectural advice and guidance, together with the automation, management and monitoring solutions, required to harness the virtualised data centre. This will pose challenges for the likes of Oracle and SAP, who will face increasing pressure to ensure that their application solutions are capable of exploiting this virtualised environment. Fusion and ESA are a step in the right direction, but there is further work to do to facilitate application provisioning, management and monitoring, together with changes to their licensing models.

IT process automation and improvement will drive spending and attention

There has been a significant focus in enterprises and public sector organisations in 2005 on IT process automation and improvement. Interest in frameworks like ITIL and COBIT are just two symptoms of this. These concerns will attract even more attention in 2006, as IT spend management and value justification add fuel to the drive for formalisation that has in many cases been initiated by regulatory and corporate policy compliance requirements. Leading organisations will look beyond the scope of ITIL to consider how all their key IT management processes can be addressed in a systematic way.

Related to this, software and application quality issues will move higher up development shops' lists of priorities as experience of web-based application technologies continues to mature and the challenges of operating complicated, multi-tier, distributed applications in heterogeneous environments become even more apparent – and as organisations getting to grips with SOA look to manage services throughout their lifecycles.

Once more, "open" will be a buzzword to watch out for and be wary of

2005 was the year when the big IT industry incumbents – particularly IBM, SAP and Oracle – started to realise the potential opportunities and threats presented by the open source software movement, and react to attempt to engage with development communities in new ways. IBM's conversion has been the most high-profile and thorough, but "openness" is finding its way into more and more vendors' marketing pitches.

As was the true when "open systems" were all the rage in the late 80s, 2006 is going to be a year in which organisations will have to carefully scrutinise vendors' claims to openness. The value of a great developer platform and developer programme has long been known (just look back at the value Microsoft got from all those developers building VB controls in the 90s) – and it's too easy to make a straightforward developer programme and make it sound like an open community development effort. There will be lots of bandwagon jumping in 2006, as IT suppliers of all stripes try and convince their customers and prospects – and commercial software developers – that supporting open innovation models is a "tick list" feature and nothing more. 2006 will show that this is far from the case.

The identity management market will remain a hot-bed of activity

2005 saw significant activity in the identity management market, driven by a combination of factors, particularly regulatory compliance, SOA approaches, growing concerns about identity theft and national identity card initiatives. The leading infrastructure, application and enterprise management vendors responded, with the likes of BMC, HP and Oracle, following the likes of CA, IBM and Sun, and acquiring identity management specialists in 2005 to bolster their capabilities in the enterprise market. At the same time, we saw ongoing development in grass roots identity initiatives such as Light-Weight Identity (LID), OpenID and Yadis (Yet Another Decentralised Identity Interoperability System!), together with offerings from vendors such as Sxip Identity, focussing on user-centric identity. The Liberty Alliance continued its work on federated identity standards, converging its work with that of OASIS with SAML 2.0, and branched out into other areas, such as strong authentication and social networking. IBM and Microsoft, meanwhile, continued to evolve the WS-* identity specifications, submitting WS-Trust, WS-SecureConversation and WS-Security Policy to OASIS.

2006 will see no let-up. Acquisitions will continue in the enterprise identity market (with SAP entering the fray), as the major vendors look to fill out their "suites", in areas such as federation, strong authentication, role-based access control and virtual directory capabilities. These vendors will continue to position their suites as more comprehensive, integrated and so forth, whilst feverishly working behind the scenes to ensure that the technology lives up to the promise. The sterling work initiated by Kim Cameron of Microsoft with the "identity metasystem" will continue with the growing realisation of the need to bridge the gap between enterprise and user-centric identity. The dual paths for federated identity standards will (sadly) persist, with the WS-* family receiving a boost as Microsoft customers begin to adopt Active Directory Federation Services as part of the latest Windows Server 2003 service pack and InfoCard materialises as part of Vista towards the end of the year. IBM has a foot in both camps, which together with the "identity metasystem", holds some promise for convergence/interoperability. posted by Neil Ward-Dutton @ 2:08 PM 0 comments permalink

Mercury finally makes a SOA play...but communication is an issue

So Mercury has announced its acquisition of "SOA governance" tools player Systinet for $105 million cash.

I can't say I'm surprised that Systinet has been snapped up by someone: the area where it (and others) play is a cornerstone technology area if service-oriented software propositions are going to really work in the enterprise, and it's been striking that small specialists like Systinet have managed to go through 2005, building nice businesses, without being snapped up.

But I (and I suspect, many many others) would have put money on the company being bought by one of the big platform vendors - IBM, BEA, Oracle, ... particularly as the latter pair both embed Systinet's technology in their service infrastructure offerings. Not Mercury.

The reason is that for a while now Mercury has seemed to be de-emphasising its interest in the software development lifecycle, and focusing more time and money on IT process management issues around portfolio management, service management and high-level IT governance; and application performance management. The software development lifecycle related tools which made the company's name (well, actually the company's previous brand name: Mercury Interactive) are still there, but Mercury has been working hard - to a significant extent, through acquisitions (Kintana, Appilog, Performant, etc) - to move well beyond them.

When you look beyond the PR and marketing at what Mercury actually offers, though, it's clear that managing the quality of the output from SOA initiatives would certainly be something that Mercury would like to do for its customers. And until now a SOA play has been conspicuous by its absence.

What's surprising to me is Mercury's timing, given that just a couple of days before the company disclosed that its failure to restate prior earnings had resulted in its delisting from NASDAQ. It's possible that the company had little choice in this regard, but I would be surprised if there wasn't a little latitude, and Mercury couldn't have delayed the Systinet announcement a little. It could be, of course, that Mercury saw an opportunity to give its shareholders some good news. But in these matters, it pays to tread carefully.

The problem is one of pure PR, and Systinet's arch rival Infravio has wasted no time in making hay. Miko Matsumura (Infravio's VP of Marketing and Technology Standards) told me:

"with Systinet consumed by Mercury, a company recently de-listed from NASDAQ for accounting improprieties, the field is open for Infravio to score some big wins."

This is quote-worthy stuff (and I've just quoted it of course ;-) despite the fact that Mercury wasn't actually delisted for impropriety, it was delisted for its failure to meet a deadline. Mercury needs to get to grips with this ASAP.

I think Mercury's challenge here is part of a broader issue of engaging with the industry and with its markets. It's interesting that the normally well-informed RedMonk boys think that Mercury is a testing tools company. The truth is quite different. The Mercury story isn't particularly smooth at the moment, but it is broader than testing.

Finally, I'm not sure if either of the following snippets of information have any bearing on the sniping by Infravio:

• One of Infravio's board members, Ido Sarig, was previously the Mercury VP of Technology Strategy


• Miko Matsumura used to be Systinet's VP of Product Marketing.

Maybe a case of sour grapes? posted by Neil Ward-Dutton @ 11:02 AM 2 comments permalink