Wednesday, July 25, 2007

SOA adds broth to technology acronym soup

SOA consultant Todd Biske, who did a great job on a panel I was honored to moderate earlier this week, has some good thoughts on the definitions and roles of IT categories and SOA.

His blog posting on "acronym soup" is worth a read.

And thanks to the other members of "The Future of SOA" panel at The Open Group conference -- Tony Baer, Beth Gold-Bernstein and Eric Knorr -- for their excellent contributions.

We'll make the discussion available as a BriefingsDirect podcast and transcript. I'll blog on it here when it's up.

Tuesday, July 24, 2007

Red Hat takes SOA modeling tool concept to open source development

In an attempt to add grease to the skids of SOA adoption, Red Hat has agreed to take a patented concept for a services modeling approach and usher it through an open source community development process before eventually adding it to the JBoss Enterprise SOA Platform.

Red Hat announced today that SOA expert and author Thomas Erl's concept for a services modeling and analysis tool will form the basis for a new JBoss.org project, the results of which will become a LGPL product that will be then be supported under the JBoss SOA offerings.

The tool-in-progress is designed to fill a gap in the market for a top-down, platform-agnostic way for non-developers to assess a business process and determine the required services that would comprise the process within a SOA environment. Erl said the tool will be a first in terms of its utility.

The way in which the tool incubates and arrives may also be unique. I'm assuming that Erl was compensated in some way by Red Hat for his intellectual property in defining the tool. And Red Hat will guide the community development process, and reap the rewards of a commercial open source model when it charges for subscription support and training for the tool and its associated SOA platform.

Perhaps more of us should be patenting ideas for software products and services, and then pitching them to Red Hat. I've always liked the idea of a cottage industry for enterprise software.

Anyway, the modeling tool as defined to date targets technical analysts so they can create a business-centric design of services, or containers of services, that combine to provide a business process or processes. The tool helps conceptualize and refine an inventory or blueprint of services, to then set the stage for the actual creation or acquisition of the requisite services, explained Erl. Managing multiple inventories also becomes easier via the tool's design.

Erl hopes the tool fosters better collaboration between analysts and SOA architects. It should help coordinate their group planning for services development and to provide a framework for iterative or Agile back-and-forth services creation with an end business process in mind.

The goal is to address the needs in the market today, to get the ball rolling on modeling for SOA, and also to be easily extensible to work within other SOA modeling initiatives -- such as UML-based work by OMG and others -- as they mature.

Community development may also create interfaces or plug ins for the service inventories to map to UDDI generally, as well as lead to specific connectors to registries and repositories.

Erl also published today a new book, "SOA: Principles of Service Design," which better fleshes out his concepts and methods on service engineering.

Monday, July 23, 2007

HP's Opsware buy highlights burgeoning role of management

I'm sitting at an engaging Open Group conference on SOA and its deep impact across IT and enterprises. A common theme, not surprisingly, is the issue of control ... some of the biggest fears about SOA is that it may be too good, too inclusive -- and may consequently spin out of control.

Managing complexity is at the heart of making SOA successful, but it's a double-edged sword. SOA can both slay and spawn complexity. Who wants to swap out integration complexity for services-driven operational and governance complexity?

Sure, SOA can bring more agility to the business processes that span companies, industries and commercial relationships. But at what cost if it's done poorly? If we gain agility but suffer chaos, breakdowns, lack of control and even more higher-abstraction complexity -- you'll see closed and brittle silos of IT assets remain very popular.

At the same time that SOA needs to be controlled well to be productive, we're seeing massive re-centralization of IT resources -- especially hardware, storage and networking infrastructure. Again, there's a tension here between efficiency and the dangers of having all your eggs in one basket. Going from nine datacenters to two makes a lot of sense, unless one of the two is out of control.

It seems to me that HP is evolving and inserting itself into the role of arbiter between control and creativity on many essential levels. We saw this in the Mercury acquisition last year, and we're seeing a affirmation of this role for HP in today's Opsware announcement.

IBM wants to help you keep your company special. HP, it seems, wants to let you know how to keep it stable.

The largest enterprises will thrive or stumble on their ability to gain agility and creativity, along with operational efficiency. They must go hand-in-hand. But true management comes with holistic control -- that which spans operational and systems management, heterogeneous datacenters management, services governance, IT governance -- and ultimately business governance. There needs to be semantic continuity to how all these management layers relate, and that usually means standards.

I was recently quoted in an article on IT and SOA management, and I believe that the distinctions between layers of management need bridging via standards and expertise.

This type of widely and deeply inclusive management and control has been missing in the IT industry. HP is demonstrating by its organic and acquisitions growth that this will be one of the essential pillars of functionality it will provide for the IT infrastructures of the future. Vigorous commercial competition and aggressive open communities involvement will be essential for making IT and automated business management conceptually and practically coordinated.

I hope we see more open source and community-based initiatives from HP to augment and accelerate its recent acquisitions activities. Opsware will be a powerful asset for HP for many years to come.

Disclosure: HP is a sponsor of BriefingsDirect podcasts.

SOA will fold into general enterprise architecture, Linthicum tells Open Group crowd

While enterprise architects (EAs) and services oriented architecture (SOA) architects are often at loggerheads today, the two will working hand in hand toward the same goals in a matter of a few years.

"In five years I don't think there will be SOA ... it's all going to fold back into enterprise architecture," said Dave Linthicum, CEO of Linthicum Group, in a keynote address today at The Open Group's Enterprise Architecture Practitioner's Conference in Austin, Texas. "SOA is a subpattern of EA."

[UPDATE: Todd Biske is also blogging from the event. ... Many slides from the speakers here.]

In effect, Linthicum said, SOA is just good EA. The goals for each are ultimately the same: To get better at building agile IT architectures and to make change the number one requirement for IT.

But that's not what you'll find on the street. In many cases those planning SOAs are not in synch with those that are keeping the trains running on time, so to speak, inside enterprise datacenters. Linthicum pointed out that there are currently "two worlds out there," enterprise architects and SOA architects, with one working up from the existing IT landscape and the other working down, respectively, from the larger concepts of agility, reuse and orchestration of service points.

"There's not a lot of synergy, and even some fighting," said Linthicum. "The EA guys don't get full implications of SOA, and SOA guys don't get how SOA meshes with existing enterprise methods and standards."

Into this Babel, many if not most CEOs think that IT is holding them back. The business leaders want to change automated business processes much more quickly.

So many business leaders are open to SOA. They cotton to the idea of IT easily adapting and becoming agile, they encourage reuse, they want independent change management. The idea is to "orchestrate" rather than integrate, and to "configure" rather than develop, said Linthicum.

At the same time, CEOs need for all the parts that are currently running to keep running. Hence the need for deeper understanding and cooperation between the EA crowd and SOA crowd.

So what are the next steps to make EA and SOA act in concert? How can the will of the organization at large be cultivated to support the $7 million to $10 million needed for even a medium-sized business to meaningfully implement SOA?

Linthicum recommends that IT leaders see beyond the SOA hype, to encourage enterprise architects to become advocates for positive change that embraces SOA principles and methods. He also says that SOA must play well with and embrace such mega trends such as SaaS, Web 2.0, application modernization, datacenter consolidation, and semantic data management.

"See the emerging web as a resource," Linthicum extolled the crowd of some 300 IT leaders and architects. "The lines are blurring between enterprise apps and the web."

Conceptually IT professionals are on board about SOA ... how to get there is the rub, he said. There remain too many "bad practices," such as selecting technology before knowing SOA needs, not sticking to EA best practices, not creating a strong business case for SOA, using wrong people, and suffering from a lack of influence and political strength in the organization.

"Don't select technology too early, don't get caught up in the hype ... look to the data issues and semantics first ... Keep your vendors working for you. The smarter you are the more successful they can help you be," said Linthicum.

Sunday, July 22, 2007

The macro economics side of new marketing efficiencies

The Wall Street Journal on July 2 contained an excellent story about how economists view U.S. businesses' prospects for the coming months. The story explains how businesses are facing higher energy, food, and labor costs while needing to hire more workers -- all of which mutes general productivity.

The story says:
If they can't pull off a resurgence in productivity, businesses face a tough choice: Raise prices or live with reduced profit margins. Judging from their outlook for corporate profits, the forecasters believe that many ... will choose to split the difference.
That means raising their prices some and living with lower profits, too. This is quite different from the most previous business climate where consistently rising productivity and tame inflation allowed for ongoing record profits.

However, there is another aspect to this somewhat bleak assessment. The next business cycle will demand that many companies focus on efficient top-line growth so that they can maintain profits, even as productivity slackens.

Many businesses may be constrained in how they can grow revenues, perhaps because they only supply a static region or supply a shrinking customer base. Most businesses, however, can find new ways to sell their goods and services in more places -- especially by better use of the Internet. That's because the Internet is and will continue to be a marketer's most powerful tool.

If the past business climate was about productivity as a means to profits, and the Internet was a benefit, then the next business cycle is about revenue growth and finding new markets as the means to fiscal health. And that means that the Internet becomes indispensable. The better you know how to leverage the Internet for your business the better off you will be as an individual -- for your current employer or the next.

Interestingly, the need to find efficient ways to increase the market for goods and services comes just as advertising -- a traditional way to grow revenues -- is in transition. Many businesses are re-evaluating how they advertise, spurred on by Google, viral marketing on the Web, and better use of community outreach and online communication with customers, partners, and prospects.

In observing IT vendors in how they reach markets in novel ways, I now see four major thrusts (and a further diminishing role for traditional advertising). The four major go-to-market avenues are:
  • Traditional inside efforts. This means creating a compelling web site, a great sales force (inside, outside, direct and channel), strong ecommerce applications, downloads and other online distribution means, and super customer support. This will not change.
  • Traditional outside efforts. This means advertising through new and old media, marketing promotions and events, email and direct marketing, and PR/AR/IR. This section may well see resources shifted to the two newer categories ...
  • Viral. This means creating content and conversation, blogs/podcasts/videocasts, or reacting to content and conversation, such that online awareness is understanding is generated about your goods, services, and image, via the social networking effects on the Web. This is and should continue to grow significantly, probably funded by the previous ad budgets.
  • Search. Through all of a business's efforts, they should focus on making their values and knowledge easily accessed via Web, and more discretely searched through the keywords and phrases that best bind it to their users and communities. This will be the more effective way to grow the top line revenues for many companies for some time. Again, look for traditional ad budgets to fuel this arena, too.
So if you are associated with a business they sees the landscape as the country's leading economists do, and you recognize you must grow both current and new markets -- to spur more revenue and business volume -- and that you must do it efficiently, do yourself a favor. Right now pretend that you are a customer or prospect of what you sell.

Now, go to Google or another major web search engine. Search on some terms that might come to your mind as you begin a research or informational journey on what your business supplies. Focus on the problem that your online prospect will have, and the solution you bring to them. Does a search on one lead to the other? Does a question about to fix what you fix actually point to how to evaluate and/or acquire that fix? Right now, online?

It should. What you should see there in the top search results on the left, or organic side, are the fruits of all your marketing efforts:
  • Your website, your product and service descriptions, pricing and how you beat the competition in value, the means to contact a sales rep, a click-through to purchase option, or more direct help.
  • The freely available trustworthy informational assets on the problem-solution set that defines your business value, including conversations, media write-ups, third-party endorsements, interviews, and blogs ... anything that your communities generate about you.
This is how new revenues and market opportunities will be born most productively. When the going gets tough, a business's online marketing gets going.

Like many, you will also buy search-based advertising, based on those essential keywords, that create more ways for those seeking you out as a business to reach your website, product information, sales and support, and solutions. But when you want the most bang for the fewer bucks, the organic results pay best and longest. Invest in them now.

What's also interesting is that the investment, and -- more importantly -- the return on the investment you and your clients make in organic search results will remain strong in nearly any macro business environment. That's right, whether the business cycle is in profit growth mode, revenue growth mode, recession, depression, boom times or flat -- your best ticket to keeping the accountants happy is bringing in leads and sales organically, via search-inspired research and inquiry.

So my prognosis is that the ways in which the Internet can assist companies have evolved well during the past 10 years, but the coming business climate -- no matter what climate it is -- is where Internet marketing will be needed the most. And the future, no matter what the world economy is up to, will also deliver the most value and power through the reach of the World Wide Web.

Saturday, July 21, 2007

Optaros opens OS projects catalog, JBoss Rules 4.0 debuts

With more than 140,000 open-source projects floating around, finding the right one for your company is a daunting task. For IT managers, who usually don't have the vantage point that developers enjoy, the search can also be fraught with danger. Choosing or backing the wrong project can have far-reaching operational, legal and financial consequences.

Now, Opatros Inc. has a helping hand. They have morphed their popular Open Source Catalogue into an online version that provides a searchable and interactive source for finding the open-source projects that best meet a company's business and technology needs.

The Optaros Enterprise Open Source Directory (EOS) eliminates the need to download and print the hefty catalogue that Optaros introduced in November 2006. That version was downloaded over 10,000 times, however, indicating an industry need for this information.

The online directory gives viewers a short description of the project, includes an Optaros rating -- one to five stars -- as well as a community rating, comments from community members, and a link to the project Web site.

The ferment in the open-source community is one of the driving forces behind the move to an online, interactive product, according to Bruno von Rotz, Optaros's EOS Directory executive sponsor.

Von Rotz explained, "Since the launch of the Open Source Catalogue six months ago, some of the leaders have further expanded their competitive positioning. Additionally, the ratings of more than 70 percent of all the projects listed have changed, and we have added and removed approximately 45 enterprise-ready projects. The open source community is developing so rapidly that an online resource is the best approach to capture changes and stay relevant."

In other open-source news, business rules are moving into the mainstream with the release of JBoss Rules 4.0, which promises faster and leaner performance, a more powerful scripting language, point-and-click rules editing, and -- most important -- access to non-programming IT workers. JBoss Rules is the RedHat product that combines the Drools project with a JBoss subscription.

JBoss blogger Pierre Fricke says the new release "lays the foundation for bringing rules-based solutions into Simple, Open and Affordable SOA deployments. . ." Would that be SOASOA or just SOA-squared?

Tony Baer at Computer Business Review sees the new offering as a move by JBoss to position its new version "as a lightweight, more accessible alternative to spending tens of hundreds of thousands of dollars on more complex alternatives."

Some of the benefits of Rules 4.0 include:
  • More expressive and powerful declarative business action scripting language
  • New Guided Rules Editor with point-and-click functionality
  • Visual modeling technology to declaratively model execution paths of related rules
  • Multi-application support (for stateful and stateless processing)
  • Hibernate readiness, and
  • Business Rules Management System (Technology Preview), a web-based, AJAX-enhanced, collaborative rule authoring, versioning and management system to help non-programming IT workers interactively author and/or modify rules that are then automatically versioned.
In other JBoss community news, ICEsoft Technologies has announced what it's promoting as a "new and improved" version of ICEfaces, the company's flagship Ajax development environment.

Version 1.6.0 now offers deeper integration with JBoss Seam, the Web 2.0 application framework.

According to ICEsoft, ICEfaces extends JavaServer Faces (JSF) and eliminates the need for low-level JavaScript development. In the new release, SeamGen enhancements support the rapid generation of functional Seam + ICEfaces applications. It also contains source code examples.

An earlier e-newsletter from ICEsoft reported that the new version contained more than 180 bug fixes, as well as several enhancements.

ICEsoft claims a developer base of 12,000 and says ICEfaces has been downloaded more than 150,000 times. The latest version can be downloaded from icefaces.org.

Wednesday, July 18, 2007

Where are the best SOA synergies now that Rogue Wave is independent?

Either today or tomorrow the completion of Battery Ventures' buyout of Rogue Wave Software will prompt some strategic thinking about where this high-performance C++ platform supplier best fits in.

As a bit of background, Rogue Wave was a part of publicly traded Quovadx, which found itself in a bit of hot water with the SEC. That now appears to be resolved, and the Integration Solutions Division and Rogue Wave units of Quovadx are being scooped up in private equity buyout by Battery Ventures. The Rouge Wave unit will operate as a stand-alone unit within a Battery subsidiary, Quartzite Holdings. The third Quovadx unit, its CareScience division, was spun off to healthcare alliance Premier in March.

What's more, word has it that Rogue Wave Executive Vice President and President Cory Isaacson will be leaving the company entirely, replaced by new CEO Bart Foster (CEO of former Battery portfolio company DecisionPoint Software).

So now that Rogue Wave is un-entangled, as it were, and part of a private equity firm that, if history is an indication, might well want to find a good home for its newly adopted firm ... someday, someway -- where might Rogue Wave land?

Well, consider that there are hordes of C++ applications scattered around the world. Consider that these applications need to accomplish two things for health and economy: modernization of platform and service-enablement. The Java applications world is undergoing just such a transition, as are myriad mainframe applications. Not too many C++ applications are being newly written specifically for SOA; Java and C# are pretty popular there.

But the larger SOA infrastructure vendors might just want a high-performance, multi-core-optimized C++ complement to their Java, .NET and mainframes modernization and/or SOA-enablement strategies. Rogue Wave also has some marquee customers in financial services, telco, software, and government. Large professional services companies that are also enjoying growth synergies by bundling modernization and SOA initiatives would also no doubt value a strong C++ component to these efforts in the market.

As frosting on this tasty cake, Rogue Wave is also tightly involved in the Apache Tuscany SCA Native (aka C++) incubation project, and Rogue Wave will likely produce a commercial platform that combines its Hydra SCA 3.5 (releases in April) with Tuscany SCA Native (aka C++) community-developer and open source code. Rogue Wave is a contributing member of the Open SOA Collaboration. So there's an open source path and commercial-open source business model potential here too.

And which large IT vendor might want to 1) augment a SOA strategy with strong C++ platform and modernization offerings, 2) avail itself of a commercial-open source market opportunity, 3) augment Java and mainframe SOA migration initiatives, and 4) make application modernization and SOA transformation professional services hay?

Why, Novell, of course. Actually, I can think of a couple of others. Rogue Wave bears watching and waiting over the coming months.

Monday, July 16, 2007

IBM broadens its data reach for SOA with DataMirror buy

Further banking on SOA as a consolidation strategy accelerator for enterprises, IBM on Monday announced its intention to acquire DataMirror.

The Markham, Ontario-based DataMirror provides real-time data capture and delivery across a broad range of data sources — including IBM's own DB2, Oracle, Sybase and Microsoft SQL Server — and allows altered data from such sources to be fed into the popular IBM Information Server. The goal: real-time integration and delivery of data deltas across a variety of sources to then reach a variety of applications and services.

The approximately $161 million acquisition, if approved, will thereby extend the reach of IBM's Information Server and give businesses faster access to data for making business decisions, responding to market demands, and rapidly identifying new business opportunities. The acquisition continues a fast-paced (for IBM) buying spree, almost as active as Oracle's over the past several years, although IBM tends to buy smaller companies that augment its strategies, rather than buy its way into new businesses wholesale.

Data and access to data is the lifeblood of any agile big businesses. According to IBM, the DataMirror technology captures data changes as soon as they occur and delivers the changes and new data either to other business processes or the enterprise service bus (ESB). This makes it an integral part of an SOA strategy.

IBM Information Server currently processes data from DB2 databases, but what DataMirror brings to the party is its technology works with a wide variety of other databases, sitting on top of the other processes, and can capture the data without slowing the performance of those operations.

So the inclusion of the DataMirror technology into its "Information on Demand portfolio" will allow Big Blue to cast a bigger umbrella over more types of real-time data for distribution and processes-updating capabilities. That means a more comprehensive SOA infrastructure capabilities set that spans data services with transactional and presentation services.

IBM said the acquisition will move their information-on-demand strategy closer to the master data management "holy grail" of providing one view of the customer -- the total and correct view.

This acquisition also therefore further amplifies the vision that management of meta data about data (aka master data management) is the keystone of the future for enterprise software infrastructure vendors. By making data free yet coordinated, IBM can position it's Information Server and IBM's Dynamic Warehousing offerings as the best-of-breed data management environments for gathering, distributing, and also analyzing real-time business activities. The value continues to move up an abstraction from the core RDBs.

As an example of how the real-time capture and delivery can aid business development, DataMirror pointed to a telecom application, in which a client company can detect when customers are running low on pre-purchased minutes, allowing the company to contact the customers in advance about purchasing more time, thereby increasing sales and preventing customers from experiencing a loss of service.

You have to wonder whether IBM will also extend what DataMirror does into open source databases -- from mySQL to Ingres to perhaps even SaaS applications and/or repositories. Such a move could allow IBM to become the all-data-for-all-purposes-oriented leader (more "open" than Microsoft or Oracle) -- while still protecting its DB2 franchise (for now). Being inclusive at the data management level is more important than protecting an installed base, right?

Among the other benefits of the combined platform will be:
  • Dynamic data warehousing
  • Real-time analytics
  • Faster disaster recovery
  • Production and e-business integration
  • Real-time event detection
DataMirror currently has about 2,200 customers, about 60 percent of whom are also IBM customers, as well as 15,000 licenses. The company and its 220 employees will be integrated into IBM's Information and Platform Solutions business unit, and IBM said it intends to retain DataMirror's Toronto-area development center focused on heterogeneous data capture.

The acquisition requires approval from DataMirror's stockholders, as well as regulators, and the deal is expected to close in the late third quarter of this year.