06.25.08
Posted in .NET, Java, SOA & Web Services at 7:10 pm by Tony Baer
In a deal announced early this morning, Progress has made an offer to buy Iona, another middleware company located just a few exists up Boston’s Route 128. At first blush, the deal sounds awfully redundant: Progress invented the ESB, and Iona’s growth business was supposed to be an ESB. My colleagues Dana Gardner and Joe McKendrick have already weighed in on the deal.
In fact, Iona CTO Eric Newcomer stated the case quite succinctly in his blog. “It’s a bit like two former rivals of the basketball l court, each with different strengths and skills, finally getting put on the same team.” More to the point, Newcomer described how, at a SOA forum several years ago where both companies presented. “For the first couple of hours we took turns saying exactly the same things about SOA, application architecture, and the unnecessary complexity of Java EE application servers. Then we each took a turn describing how our respective products met the same requirements, and served exactly the same segment of the industry (i.e. SOA infrastructure) with different approaches.”
There are some differences, notably that Artix has some mainframe and Microsoft connectivity that could complement Progress’s Sonic ESB. But the more important place to keep your eye focused is on the open source FUSE side of Iona’s (soon to be Progress’s) business, as it makes a play for lighter weight containers that Sonic doesn’t focus on. It could also make a good match for Progress’s legacy OpenEdge customer base, which consists of SMBs who like things simple.
Permalink
06.23.08
Posted in .NET, Application Development, Java, Open Source, e-Commerce at 8:40 am by Tony Baer
We had an interesting conversation with Peter Cooper-Ellis, the guy who ran product management at BEA from the time it acquired WebLogic, who’s now taking on a similar role with SpringSource. Obviously, in the wake of the Oracle acquisition, it’s not surprising that Cooper-Ellis jumped ship.
But in making the jump to SpringSource, Cooper-Ellis has come full circle. As BEA was digesting its WebLogic acquisition, Cooper-Ellis was there when the Java stack was being built. Now at SpringSource, with its Eclipse Equinox OSGi-based server container, he’s now part of an open source company that’s helping deconstruct it. So we explored some history with him and compared notes. To summarize, Cooper-Ellis sees a bit of history repeating again today: a decade ago, it was drive for a unified middle tier stack to make the web interactive, and today, it’s the goal of having a dynamic lightweight stack that uses simpler constructs. In other words, a technology framework that actually delivers on the old promise of Internet time.
Let’s rewind the tape a bit. In the 90s, BEA (originally called Independence Technology) was formed to make a business in middleware. It thought its business would come from transaction monitors, but that only addressed a tiny portion of the market with transaction loads huge enough to justify buying another layer of software. Instead the killer app for middleware occurred with the appserver. When the web caught on, there was demand to add the kind of data-driven interactivity that became real with client/server. BEA bought WebLogic, a company that gambled (and won) a bet that EJBs would become standard (which it did with J2EE in 1999).
The good news was that J2EE (later joined by rival .NET) provided the standard middle tier that made e-commerce bloom (if you’re going to sell something, you need a database behind your online ordering system). The bad news was that J2EE was obese, proving overkill for anybody who wasn’t an Amazon, eBay, online banking, or travel reservations site – it was engineered for transaction-intensive, highly distributed data centric websites. Not surprisingly, the complexity of J2EE subsequently spawned a backlash for Plain Old Java Objects (POJOs), supported by a growing array of open source frameworks made famous by then Burton Group analyst Richard Monson-Haefel in 2004 as the Rebel Frameworks. More recently, there has been surging interest in dynamic scripting languages that let you manipulate data using higher-level constructs.
But so far, all these technologies were about development, not run time. That’s where Eclipse Equinox comes in. Leveraging the OSGi component model that Eclipse embraced for the IDE, Equinox extends the idea to run time. Like Java, OSGi was conceived for different purposes (Java was for set-top boxes, while OSGi was the framework for provisioning services in the smart, networked home) – you could consider both as fraternal twins reunited at adolescence. Eclipse reincarnated OSGi as the dynamic service bundle, first for its IDE (where developers could swap different vendor plug-ins at will), and more recently as a new run time.
That’s where Cooper-Ellis views OSGi as giving Java appservers a second wind. In place of installing the entire Java EE stack, OSGi lets you provision only the features you need at run time. So if you add distributed nodes, take the JMS plug-in; if traffic spikes, hot deploy clustering support, and so on. The advantage is that if you don’t need these or other bundles, you could run the server on a very tiny footprint of code, which reduces overhead and potentially makes it blazingly fast.
That was what BEA was trying to do with the micro-Service Architecture (mSA) that it announced roughly 18 months before Oracle swooped in, and how it built WebLogic Event Server, its complex event streaming engine. The product only used Java EE features such as managing availability, security, and user management; dispensed with much of the rest of the stack; and supported development with POJOs, which included support of the Spring framework.
OSGi/Eclipse Equinox is part of the same return to simplicity that originally spawned popularity of POJOs and the rebel frameworks. Beyond the Java space, it’s the same trend that has driven popularity of dynamic scripting languages as faster means to developing the relatively straightforward data-driven apps that are the mainstream of the web, and it’s also the driving force behind Ajax (which is easy enough that casual developers, business analysts, or web designers can grow dangerous). Each of these has catches and limitations, but they are evidence that for the rest of us, the 80/20 rule lives when it comes to developing for the web.
Permalink
06.19.08
Posted in Application Development, Application Lifecycle Management (ALM), IT Infrastructure, ITIL at 5:04 pm by Tony Baer
Reporting from HP Software Universe, colleague Dana Gardner provided an interesting account of a keynote delivered by BTO software unit general manager (and Opsware alum) Ben Horowitz on the synergies between the application lifecycle and the data center lifecycle. It’s a topic that’s not just tailor-made for HP Software (whose acquisitions include Mercury, covering the software lifecycle, and Opsware, which complements the former OpenView in data center operations), but also a hot button for us as well. Traditionally, the IT organization has been heavily silo’ed: not only are there walls between different players in the software group (e.g., architects, developers, QA), but also between software and operations. So while software development folks are supposed to performance and integration test code, when it comes to migrating code to production, the process has been one of throwing production-ready code over the wall to operations.
While there has always been a disconnect between software development and the data center, the gap has become even more glaring with emergence of SOA and its promises for enabling enterprise agility. That is, if you can make services so flexible that you can swap pieces out (like selecting a different weather forecasting service for transportation routing), and make them responsive to the business through enforcement of service contracts, how can you deliver when you can’t control whether the underlying IT infrastructure can handle the load and provide response times that comply with the contract. Significantly, none of the tools that handle run-time SOA governance have trap doors that automatically re-provision capacity. In an era of risk aversion, the last thing that data center stewards want is software developers hijacking iron. When we spoke with Tim Hall, product director for HP’s SOA Center after the product was released, he told us that “Dynamic flexing of resources is a nice idea that won’t sit well with the operations people.”
Gardner reported HP’s Horowitz describing the roles that the business, security specialists, IT operations, and QA (note that developers were omitted) play in the transition form design to run time.
What makes HP’s proposition thinkable is that there is an emerging awareness to enforce process management mentality on IT operations. While most organizations observe software development lifecycle processes in the breach, there is a consciousness that developing software should encompass collecting and validating requirements, developing or mapping to an architecture, generating code, and testing. With some of the newer agile methodologies, many of these steps are performed concurrently and in smaller chunks, but they are still supposed to be performed. What’s new is the IT operations side; significantly, notably where the latest version of the ITIL framework takes a lifecycle view of the management and delivery of IT services. There are some parallels with the SDLC: Service Strategy has a logical fit with Requirements; Service Design fits with well with Development (although likeness to architecture or design may not be apparent); Service Transition deals with operations and incidents, which is not addressed in the SDLC; while Continual Service Improvement relates well to the maintenance and upgrade part of the SDLC.
While HP’s Horowitz might not have been speaking about ITIL v3 literally in his keynote – and while not all data center organizations are gung ho about ITIL itself – there is growing awareness inside the data center that you can’t just run operations by reflex anymore.
Permalink
06.13.08
Posted in .NET, Agile Development, Application Development, Java, Middleware, Open Source, Rich Internet Apps., SOA & Web Services, Technology Market Trends, Web 2.0 Apps at 12:39 am by Tony Baer
A conversation this week with database veteran Jnan Dash reminded us of one salient fact regarding computing, and more specifically, software platforms. That there never was and never will be a single grand unifying platform that totally flattens the playing field and eradicates all differences.
Dash should know, having been part of the teams that developed DB2, and after that, Oracle, who currently keeps off the street by advising tools companies that have gotten past startup phase. For now, his gig is advising Curl, developers of a self-contained language for Rich Internet Applications (RIA) combining declarative GUI and OO business logic inside the same language, which had the misfortune of emerging before its time (the term RIA had yet to be coined).
Curl provides an answer to unifying one piece of the process – developing the rich front end. But it’s a far cry from the false euphoria over “Write once, run anywhere” that emerged during Web 1.0, where the train of thought was on a single language (Java or later, C#) for logic on a single mid-tier back end and a universal HTML, HTTP, and TCP/IP stack for connectivity to the front end. Of course, not all web browsers were fully W3C compliant, and in the end, bandwidth killed the idea of Java applets (the original vision for RIA) and disputes between Sun and Microsoft gave rise to a Java/.NET duopoly on the back end. But the end result was not only a dumbed down thin client that was little more than a green screen with a pretty face, but also a dumbed down IDE market, as the Java/.NET duopoly effectively made development tooling commodity. Frankly, it made the tools market quite boring.
That’s in marked contrast to the swirl of competition that characterized the 4GL client/server era a few years before, where emergence of two key standards (SQL databases and Windows clients) provided a standard enough target that spawned a vibrant market of competing languages and IDEs that rapidly pushed innovation. Competition between VB, SQL Windows, PowerBuilder, Delphi and others spawned a race for ease of use, a secondary market for visual controls, simplified database connectivity, and birth of ideas like model-driven development and unified software development lifecycles.
What’s ironic is that today, roughly a decade later, we’re still trying to get to many of those goals. Significantly, as technology grew commodity, most of the innovation shifted to process methodology (witness the birth of the Agile Manifesto back in 2001).
While agile methodologies are continuing to evolve, we sense that the pendulum of innovation is shifting back to technology. In a talk on scaling agile at the Rational Software Development Conference last week, Scott Ambler told agile folks to, in effect, grow up and embrace some more traditional methods – like perform some modeling before you start – if you’re trying agile on an enterprise scale.
More to the point, the combined impacts of emergence of Web 2.0, emergence of open source, and a desire to simplify development such as what former Burton analyst Richard Monson-Haefel (who’s now an evangelist with Curl) termed the J2EE rebel frameworks spawned a new diversity of technology approaches and architectures.
Quoted in an article by John Waters, XML co-inventor and Sun director of web technologies Tim Bray recently acknowledged some of the new diversity in programming languages. “Until a few years ago, the entire world was either Java or .NET… And now all of a sudden, we have an explosion of new languages. We are at a very exciting inflection point where any new language with a good design basis has a chance of becoming a major player in the software development scene.”
Beyond languages, a partial list of innovations might include:
• A variety of open source frameworks like Spring or Hibernate that are abstracting (and simplifying use of) Java EE constructs and promoting a back to basics movement with Plain Old Java Objects (POJOs) and rethinking of the appserver tier;
• Emergence of mashups as a new path for accessible development and integration;
• Competition between frameworks and approaches for integrating design and development of Internet apps too rich for Ajax;
• Emergence of RESTful style development as simpler alternatives for data-driven SOA; and in turn,
• New competition for what we used to call component-based development; e.g., whether components should be formed at the programming language level (EJB, .NET) vs. web services level (Service Component Architecture, or SCA).
In short, there are no pat answers to developing or composing applications; it’s no longer simply, choose vanilla or chocolate for the back end, and using generic IDEs for churning out logic and presentation. In other words, competition has returned to software development technologies and architectural approaches, making the marketplace interesting once again.
Permalink
06.12.08
Posted in Application Development, Rich Internet Apps., SOA & Web Services, Web 2.0 Apps at 4:06 pm by Tony Baer
How often have you heard vendors extol their products as being so simple that people from the business side can take charge and configure their reports, manage their portals, or if you listen to all the enterprise mashup providers, that business people can assemble neat little personalized disposable apps without having to call on IT? We’ve seen our share of easy-to-use end-user tools that look pretty impressive, and at times have drank the Kool-Aid ourselves.
Reviewing the proceeding of a panel session at this past week’s Enterprise 2.0 conference, prolific BPM blogger Sandy Kemsley gave us a fresh shot of common sense. Commenting on a panel session that covered how mashups could consume data from basic, ubiquitous sources such as Atom/RSS feeds, SOAP, RESTful services, etc., Kemsley reminded us that when you put together mashups, the processes is akin to piecing together what we’d term jigsaw puzzle: you have to know something about how the pieces fit together. She stated that you need to consider the interfaces, and concluded, “Realistically, business users still can’t do mashups, in spite of what the vendors tell you…”
She stated that dragging and dropping is, literally, only the tip of the iceberg, as you need to know how those pieces may interact (isn’t that the point of doing a mashup?). Otherwise, if you just stick a two, three or more silo’ed data sources on your screen that don’t interact, you’re simply putting together a portal page, which may be OK in and of itself. It’s the difference between a dynamic mashup of a Google Map which shows the location of the sales leads that you overlaid atop it, or just a Google Map with a static table that doesn’t show where on the map those leads are. And, as we wrote after a conversation with Informatica’s Ashutosh Kulkarni a few months back, issues such as architectural integrity, customer privacy protection, or access control may not necessarily be forefront on the end user’s mind.
Admittedly, enterprise mashup providers like Serena and IBM remind us that their offerings provide protected sandboxes within which business users can mash safely vetted assets to limit or eliminate the possibility of data breaches. Clearly, mashups have the potential to make disposable applications more accessible to the rest of us. Just don’t forget to get some adult supervision.
Permalink
06.11.08
Posted in .NET, Application Development, Application Lifecycle Management (ALM), BPM, SOA & Web Services, Standards Development at 7:28 pm by Tony Baer
While initial reviews of Bill Gates’ final Microsoft keynote in his chief architect role were rather underwhelming, partner in crime Jack Vaughan caught one interesting detail that was latched onto by Michael Meehan this week: that Microsoft’s forthcoming Microsoft SOA strategy would indeed embrace UML as its core modeling language.
Previously, Microsoft had been noncommittal and was promoting the idea of Domain Specific Languages (DSLs) as replacement for UML. The backdrop to all this is that, while UML received practically universal vendor support when it was proposed a decade ago, version 2.0 took UML down a much weightier, more controversial path. That is, while the first version modeled the logical structure of a software application, UML 2.0 added artifacts relating to how an application is deployed at runtime. Or as IBM Rational distinguished engineer (and UML 2.0 task force co-lead) Bran Selic explained in a Computerworld article back in 2004, “The other one that I would single out as being a major influence is the ability to model very large heterogeneous systems so that one could describe the architectures of these systems quite succinctly.”
The result was that some UML 1.x stalwarts felt that the language had grown too complex and veered away from the spirit of earlier versions, which didn’t factor physical architectures. Dissidents either searched for alternate, lighter weight ways of modeling an application or simply stuck to the diagrams contained in UML 1.2, which largely remained intact as a subset of 2.0.
Ironically, Microsoft was one of the very parties that signed off on UML 1.0, and shortly thereafter opened a partnership with Rational that bundled a lite version of Rational ROSE that included a couple basic UML diagrams; it later developed its own tool that supported more UML 1.2 diagrams when it released Visual Studio.NET in 2002 (later that year, it passed up the idea of acquiring Rational, clearing the way for IBM).
Fast forward several years to Microsoft’s ramp up of team development with Visual Studio Team System (VSTS), and UML was MIA. At the time, Microsoft stated it preferred lighter alternative, which is when it began mentioning the idea of using DSLs. But as Team System took several years to evolve (and initially focused more on testing), the need for modeling wasn’t critical. But with disclosure last year of Oslo, Microsoft’s future product strategy for model-driven, service-oriented development, it had to plant its stake in the sand as to what modeling language(s) would get supported.
Last week at TechEd, Microsoft committed to releasing a community technology preview of three Oslo pillars: development tools, repository, and a new declarative modeling language that Mary Jo Foley has reported is code named “D.” So, what to make of Gates’ commitment (probably his last Microsoft product decision) to UML?
Our take is that Microsoft is pursuing a dual-pronged strategy: so-called D will be a higher level modeling language that would be its own “better” answer to BPMN, which would be targeted at business analysts, while UML is for developers working on how to represent business processes in software programmatically.
Permalink
06.10.08
Posted in IT Governance, IT Value/ROI, IT/Business Alignment, SaaS (Software as a Service) at 12:45 am by Tony Baer
The old adage of the shoemaker’s son’s going barefoot has long been associated with IT organizations because they have often been the last to adopt the kind of automated solutions to run their shops that they have implemented for the rest of the enterprise.
Of course, packaging up what IT does is kind of difficult because, compared to line organizations, technology group processes and activities are so diverse. On one hand, they are regarded as a digital utility that is responsible for keeping digital lights on; not surprisingly, that is a dominant impression given that, according to a wide range of surveys, IT infrastructure and software maintenance easily consumes 70% or more (depending on whose survey you read). As to what’s left over, that’s where IT is supposed to play project delivery mode, where it delivers the technologies that are supposed to support business innovation. In that role, IT juggles several roles including systems integration, application development, and project management. And, if IT is to function properly, it is supposed to govern itself.
With ITIL emerging as a -– fill in the blank -– process enabler or yet another layer of red tape, the goal has been to make the utility side of IT a more consistent business with repeatable processes that should improve service levels and reduce cost. While adherence to the ITIL framework, or the broader discipline of IT Service Management, is supposed to make the business feel that it is getting more value for its IT dollars (better service can translate to competitive edge, especially if you heavily leverage the web as a channel for dealing with business partners or customers), the brass ring is supposed to come from the way that IT overtly supports business innovation in project delivery. But like any business, there is the need to manage your investments, a concern that has driven emergence of Project Portfolio Management as a means for IT organizations to evaluate how well projects are meeting their budgets, schedules, and defined goals; in some ways, it’s tempting to label PPM as ERP for IT, as it’s intended as an application for planning where IT should direct its resources.
Five years ago, Mercury’s (now part of HP) acquisition of Kintana, which was intended to grow the company out of its development/testing niche, began putting PPM on the map. That was followed in the next few years with each of the major development tools players acquiring their ways into this space.
Of course, the devil’s in the details. PPM is not a simple answer to a complex problem, as it requires decent data from IT projects that could encompass, not only accomplished milestones from a project management system, but feedback or reports from quality management or requirements analysis tools to ensure that the software – no matter how mission-critical – isn’t getting bogged down with insurmountable defects or veering away from intended scope. Significantly, at least one major player – IBM – is rethinking its whole PPM approach, with the result being that it will likely split its future solutions into separate project, portfolio, and program management streams.
Against this backdrop, Innotas has carved a different path. Founded by veterans of Kintana, the goal of the founders is to make the new generation a kinder, gentler PPM for the rest of us. Delivered as a SaaS offering, the company not surprisingly differentiates itself using the Siebel/Salesforce.com metaphor (and in fact is a marketing partner on Salesforce’s App Exchange). While we haven’t yet caught a demo to attest that Innotas PPM really is simpler, the company did grow 4x to a hundred customers last year, expects to double this year (recession notwithstanding), and just received a modest $6 million shot of C funding to do the usual late round expand sales & marketing.
The dilemma of governance is that lacking bounds, you can often wind up spinning wheels just to get the last detail, whether you actually need it or not. Not surprisingly for a top-down solution, and one that’s aimed at IT, the PPM market has been fairly limited. Significantly, in the press release, Innotas referred to the size of the SaaS rather than the PPM market to promote its upside potential. While that begs the question, there’s always the classic China market argument of a relatively empty market just waiting to be educated. In this case, Innotas points to the fact that barely 300 of the Fortune 1000 have bought PPM tools so far, and that doesn’t even count the 50,000 midmarket companies that have yet to be penetrated.
Our comeback is that like any midmarket solution, the burden is on the vendor to make the case that midmarket companies, with their more modest IT software portfolios, have resource management problems that are complex enough to warrant such a solution. Nonetheless, the PPM market is in need of solutions that can at least give you an 80% answer, because most organizations don’t have the time or resource to maintain fully staffed program management offices whose mission is to perform exactly that.
Permalink
06.06.08
Posted in .NET, Application Development at 6:16 pm by Tony Baer
While we were down at the Rational Software Development Conference in Orlando earlier this week, we were a bit wistful that we didn’t have the time to make it cross town to Microsoft TechEd, where Gates was delivering his last keynote in his official Microsoft capacity at a Microsoft event.
Later that day, we caught up with one of our colleagues who made the trek, who told us that we didn’t miss anything. Our friend and uber .NET developer Andrew Brust, who was also present, confirmed as much but then said that he (and about a dozen other inner circle .NET consultants) had the chance to dine with Gates afterward, and the conversation was far more engaging. It confirmed our impressions that at this point, Gates’ mind has checked out in favor of a broader mission to stamp out hunger and disease. Call him the Andrew Carnegie of our generation.
You can read Brust’s account of his lunch with Gates here.
Permalink
06.05.08
Posted in Business Intelligence, Data Management, Database, Enterprise Integration, Middleware at 3:21 am by Tony Baer
During a day of briefings before analysts, we gained a glimpse at Informatica’s mid-long term product roadmap. The theme was reflected in a recent poll that the company conducted across a sampling of its installed base: while 90% of its customers use Informatica’s tools (principally PowerCenter) for its well known use as a data transformation and migration tool for data warehouses, almost all of them are also using its tools for other purposes such as feeding operational data stores, providing data services, or performing integration of unstructured data for B2B documents.
The not-so-dirty secret behind all this is, of course, that in an era where blades, storage, and bandwidth are almost like free beer, that the demarcation between classic business intelligence (which trends historical data) and business activity monitoring/corporate performance management (which tends to be more real-time, is evaporating. Adam Wilson, product manager for the PowerCenter and related products, displayed a chart showing Informatica extending its reach beyond classic ETL running up to, but stopping just short of full transactional two-phase commit.
In other words, if you’re talking data federation or operational integration of data from multiple sources, Informatica says it is or will soon be in that space. A major chunk of that was its announcement this week of a real-time edition of PowerCenter, which adds data streaming to its bag of tricks. In other words, you’ll have a spectrum of options, from classic batch-driven ETL through near real-time trickle feeds, to more real-time data streaming. It means that Informatica is potentially butting heads with EAI players like Tibco and IBM (on another front; thanks to its acquisition if Ascential, IBM was already its main competitor for the more classic data integration piece).
Data streaming is considered to be a more efficient means for transferring large volumes of data compared to conventional I/O processes, and in fact, it’s the foundation for business, or complex event processing. The operable notion is that, when you have large torrents of events and need some way of deciphering trends before they splash you in the face – such as scanning of RFID data to detect if contraband is slipping through. It’s a market that virtually every database and EAI platform provider, plus a bevy of startups, are pursuing – and it’s one where a new technical society is being formed. As Dana Gardner speculates, it’s an area that should spawn a wide variety of intelligent services, not only for financial services companies (where processes like trading are assumed to be masses of complex events, but also consumer applications.
The question is, where will Informatica play in all this? The company is in the midst of an aggressive partnering strategy that is seeing its wares become the official data integration and migration engine for Wipro’s outsourced services, not to mention starter packages offered by Teradata, and extensive alliances with Accenture and HP/EDS. So it’s clear that the company won’t become an application or database provider, but as it adds a streaming capability to its data integration offerings, it will dictate some careful butterfly strokes to figure out where those dips in the stream will lead.
Permalink
Posted in Database, Open Source at 3:04 am by Tony Baer
On the heels of securing a new round of $10 million venture financing barely a couple months ago, EnterpriseDB has deepened its management bench by naming Red Hat’s former head of North American sales to take over as CEO. As you might recall, EnterpriseDB is the company commercializing the open source Postgres SQL database, focusing on the Oracle market.
Ed Boyajian, the former Red Hat exec, succeeds founder Andy Astor, who is concentrating on partner development. Excluding the fact that today, open source has become accepted by the enterprise mainstream, EnterpriseDB is roughly where Red Hat was when Boyajian, came onboard 6 years ago. The company was developing a model for commercializing Linux, which culminated in the dual-track Fedora open source and Red Hat Enterprise Linux commercial model, and while it’s premature for Boyajian to tell us his plans for EnterpriseDB (he’s formally joining the staff next week), it’s not brain surgery to conclude that the company will likely travel a similar track.
Behind the headlines, what’s significant is that Astor will now spend full-time on partnerships. That’s potentially pivotal given that IBM was one of the backers of EnterpriseDB’s last round of funding. Draw your own conclusions.
Permalink
« Previous entries Next Page » Next Page »