Many people much younger than me know the name of the Lone Ranger’s horse even though the TV show went off the air in 1957 (according to Wikipedia). The concept of the silver bullet (as in “there is no silver bullet”) and the masked man’s accompanying call to action live on everywhere. But few remember the name of the Ranger’s faithful companion Tonto’s faithful companion.

What’s that have to do with open source software (OSS)? Stick with me here.

My recent blogpost on FiveRuns genned up a little interest and illustrated to me the vibrancy and potentially disruptive nature of the Ruby/Rails ecosystem to the relatively stable open source development model. (Note I said “relatively stable.”) An example is information I received subsequent to the FiveRuns post from Derek Haynes of Highgroove Studios, which has been stealth marketing an open source server monitoring and reporting application for the last few months.

On April 15, Highgroove “went public” with its project. If you already know that its name is Scout, you are already a kemo sabe to Ruby and Rails. (I could go on about the fact that Tonto was supposedly an Apache but I would really be going off the rails with that one.)

Per Derek’s announcement,

“The Scout client... is a normal Ruby gem, open for development, and distributed under the MIT and/or Ruby License.”

Members of the community write Scout Plugins which are also completely open.
Derek says:

“In fact, they are surrounded and fostered by a community that encourages branching, fixes, and general openness.”

Highgroove maintains the system and the Scout Server software, where data is aggregated and users collect information about their account, is not open-source. Highgroove maintains the server, and worries about security, uptime and all those good things.

The Scout client is lean and the plugins are whatever you want them to be. There are already over a dozen, such as for permission checking, feedburner statistics, disk usage and so forth.

Hayes and his group, which has been developing in Rails since 2005, plus Scout partners (e.g., Rails Machine) are pushing the community process because it believes:

“No single organization can keep up with innovations in the Rails ecosystem and the entire industry.”

No lone rangers here.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

No, let me rephrase that headline. Alfresco wants to make its software THE enterprise content management (ECM) commodity. Scratch the "open source;" Alfresco's open source development model and terms and conditions are a means to an end by which Alfresco becomes a factor in the entire ECM market. Scratch the "a;" it implies Alfresco would be happy sharing the limelight with another supplier.

I met up with Ian Howells, chief marketing officer at Alfresco, recently. The immediate reason for the meeting was to get his input to my Industry-Oriented Open Source Software (OSS) research, an abstract of which has been posted in the features section on ebizQ (Gold Club membership required but there is no cost).

But I took the opportunity to ask Ian two of my standard “Talking to… questions:” How did you get to the open source movement? Where do you think the movement goes next? (My third standard question concerns the company; for more about Alfresco itself, see my Talking to… interview with John Newton posted here, also in the ebizQ features section with all my monthly articles.)

How did you get here? Ian followed an interesting path to OSS and formed his ideas about OSS marketing from that experience. He worked for Ingres, Documentum, and SeeBeyond (now part of Sun). Particularly at Ingres, he saw the way standards solidified the database market and how the standards ended up pointing to an eventual market winner (at least so far it's Oracle but Sun/MySQL would like to change that). In Ian’s opinion, Ingres had a better database product back in the 1980s but Oracle took advantage of the standards situation by pushing SQL, making it a virtual commodity. Ian sees a similar pattern emerging in ECM, where—he believes—both Alfresco and Microsoft (with Sharepoint) will commoditize ECM.

Where are we going? As you can see in the Industry-Oriented OSS article referenced above, Alfresco leaves it to the Alfresco channel and the community to write the ECM applications. OpenQuote, mentioned in the article, is an example. That makes Microsoft Alfresco’s biggest competitor according to Ian Howell because it has such a channel advantage, and so much experience marketing commodity technology. He points out that traditionally a first mover in a market did not have an advantage (and often had a disadvantage). Software markets usually consisted of dozens of competitors. That group was eventually whittled down to two or three. Often first movers ran out of gas before having a chance to make the cut.

However, Ian thinks the open source movement has led to a change in software market dynamics such that the leading open source project in any given category can become a commodity quickly. With open source, Ian believes, a different pattern is emerging. First movers such as JBoss, Linux (and he hopes Alfresco) tend to dominate going forward.

I agree with him but probably because of a different thought process: most open source players are not forming a market themselves but coming in as competitors in existing markets in almost all cases I can think of (web server software is an exception but even now that functionality has merged back into application server sector or back within the operating system with which it works). So Alfresco is not competing on the basis of its open source terms and conditions against other open source ECM products but against the traditional ECM guys (EMC/Documentum, IBM/Filenet, etc.).

I think Ian and I get to the same conclusion by different roads. That's where he says Alfresco and Microsoft have a great advantage because both products are opening up an entirely new ECM opportunity, increasing the number of ECM seats in an enterprise by making ECM dramatically easier to install and use.

And giving Alfresco (and Microsoft) partners a great base on which to build industry-specific ECM applications. Read the artcle if you get a chance; it's published in two parts.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

In recent posts on this blog and across the blogosphere, there is much discussion about the increasing penetration of open source into enterprises. The illustration above, from a year-end open source software (OSS) report available here on ebizQ, explains the three different strains of open source, and why, when analysts talk about open source taking over the world, it all depends on what the meaning of open source is. The illustration uses middleware as examples but a similar timeline of open source history can be drawn with most types of infrastructure software as examples.

Open source has sort of been around since the beginning of the information technology (IT) industry. Of course that didn't really matter 50 years ago because most software was also bundled with, leased with and only ran on the hardware it came with. After IBM was forced to unbundle its software by a consent agreement with the U.S. Department of Justice around 1970 (sound familiar all you twenty-something Microbashers), a split in software terms and conditions (Ts&Cs) became pronounced. There were academic developments (on the right of the chart) with one set of Ts&Cs while most software remained effectively bundled (left hand side) even though it was priced and licensed separately. The GNU utilities are an example of the former. In the middle, independent software vendors (ISVs), a new type of company, created software products by writing their own middleware and kernels (e.g., SAP ABAP and Basis) in order to more easily port their software from one closed platform to another. (Many other ISVs, most of whose names are long forgotten, wrote their software exclusively for one platform such as the System/3, AS/400, or NetWare.)

Then the left and right were joined (literally and figuratively) by a third strain of products that are released with some open source characteristics but under Ts&Cs that do not require that any software that uses it also be redistributable or its source opened (the best example being the original Apache). This means the products can be used and forked and otherwise played with by anyone in a way similar to GNU General Public License (GPL) products. But it also means that the new strain of OSS products can be built into closed products (which can't be played with). Both strains can be used to build a maintenance support and professional IT services business exactly the same as the original ISVs have run for 30 years. Sometimes these are closed source products that have the code quite "embedded" (e.g., Apache in vanilla WebSphere AS since the mid 1990s) and sometimes the products are simply an "enterprise edition" of the community project (e.g., WebSphere's version of Apache Gluecode).

From a market analysis perspective, in order to determine how OSS might "change the market," the major question is what strain of open source are we talking about.
1. If Academic/GNU/GPL/community-edition type software can be used effectively by enterprises without accompanying service (for example, as Red Hat discovered with JBoss, some of this software doesn't require a lot of maintenance), enterprise spending on IT will actually contract (at least comparing apples to apples) and it can be truly said that open source has taken over enterprises.
2. If it turns out that the community edition strain of OSS requires as much support as the more traditional strain of commercial software, so that enterprises typically buy the "enterprise edition," the IT market will evolve the same as it has for the last 30 years. The major difference will be that today IT vendors recognize about 33% of your payments to them as licenses, whereas -- if scenario 2 is the case -- they will recognize all your payments as services revenues. You'll still pay the same.

Separately I will look at the penetration of open source into cloud computing.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

UPDATED 4/4/2008: I talked to Ian Howells of Alfresco, the U.K-based enterprise content management software maker the other day as I promised to do when I wrote this post back in February.

More important, here's a special apology to anyone who spent time in a computer-acronym dictionary trying to figure out what an IHOP did. It stands for a U.S. restaurant chain, the International House of Pancakes. Now I am hoping everyone around the world knows what a pancake is. You must because the IHOPs fly all your flags as part of their decor.

The second edition of the Alfresco Software Open Source Barometer was released on February 12 and its claims and some misleading possible non-Alfresco interpretations rapidly spread across the open source blogosphere. The most interesting information I found in blog posts about the survey was that IHOP gives away free pancakes on National Pancake day. That was useful data.

But don’t necessarily blame Alfresco for any misinterpretations. I have cautioned before about putting too much stock in web-input, “opt-in” (Alfresco’s term) select-audience, apparently non-weighted statistics such as these. Alfresco itself recognizes this is an issue and describes the methodology in the second slide in the presentation on the survey that it made at the JBoss user conference.

The Alfresco survey results actually come from two different sources and the “n” and methodology of the most interesting finding is different than the more widely publicized findings based on the “Barometer Survey.” Alfresco’s separate research is based on what Alfresco calls the “Group Deployment Survey.” It confirms a long-time statistic in the IT industry, which predates the modern era of open source software (OSS). That finding: Developers like to evaluate on Windows and deploy to Unix/Linux.

Overall in my research experience that statistic is moving in Windows’ favor over time as more and more Windows servers have become available. But the press is reporting that Alfresco found the opposite. That would truly be interesting but I cannot find a time-series comparison for the Alfresco “Group Deployment Survey” data so I cannot analyze the claim. And to be fair, what Alfresco is quoted as saying is that

“In the 2006 survey, for example, 41 percent of respondents reported that they deployed on Linux. A year later, that number jumped to 51 percent.”

If the switch is from UNIX, that would be consistent with my experience over the years.

Most important, the other publicized findings (Ubuntu beating Red Hat, MySQL beating Oracle, Ubuntu and Red Hat beating Suse because of the Microsoft-Novell patent agreement) can only be applied to people that sign up for the Alfresco Community. These are not necessarily even Alfresco “users” as is widely reported. North America is grossly underrepresented in those that sign up for the Alfresco community (presumably because it is a UK company) as are Oracle database users (which years of research confirm equals 75% of the universe).

It would be interesting to see a cross-tab of just the Oracle database users and just the North American users and see if they differ greatly from the net of the rest.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

Attention all open source software (OSS) projects and organizations: I am researching the next in a series of OSS-related middleware research articles for ebizQ.

This month we are looking for open-source software and projects specific to an event processing--or complex event processing (CEP)--middleware.

The article is tentatively scheduled for release in mid-May 2008. It will be similar to recent ebizQ reports on open source business process management software and ERP open source software (OSS) applications (ebizQ Gold Club membership required but there is no charge to join). Your company’s or project's product(s) may be mentioned based on my secondary research but if you would like to formally participate, please download and return the attached 1-page survey form by Friday April 25, 2008 to dennis@ebizq.net.

Download file

If you do not offer such software but have a partner that uses your OSS product to enable event processing or CEP, pass this on (and let me know your partner’s company or project name). The partner can be a systems integrator or other type of services provider. OSS service providers, et me know what you are doing as well although the survey form might not be approrpiate. Just describe your activity in an email to dennis@ebizQ.net. Open source software delivered as a service (SaaS) will also be covered.

Note that as the survey indicates, software products will be covered in the report if they use OSS (e.g., bundle in an OSS application server product such as JBoss) even if they are not “sold” as OSS themselves and no matter how they are monetized.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

I met up with Jonathan Marsh of WSO2 last week; I introduced you to WSO2 in mid 2007 via a Talking to… interview with Paul Freemantle. Jonathan’s role is Director of Architecture for Mashup Technologies but he prefers to call himself a “product designer for people.” He joined the UK-based open source software (OSS) organization in 2007 from that alleged OSS nemesis, Microsoft.

But no Microbashing in this interview. Jonathan simply felt OSS—and WSO2 in particular—provided him personally a better way to “design for people.” This is partially because he didn’t want to move to the company’s headquarters location, something that is highly encouraged in any large non-OSS development organization. He also believes OSS’ community aspect is especially well suited for people-oriented software design.

Although new to OSS, Jonathan is not new to community-based activities. At Microsoft, he had worked on the Worldwide Web Consortium (W3C) as Microsoft representative, and was co-chair of the WSDL 2.0 group. Microsoft’s involvement in standards activity such as WSDL illustrates its understanding of their importance. Jonathan sees a lot of similarities between standards and OSS in terms of there being community, processes, and a wide variety of contributors. (Truth in advertising: I told Jonathan of my objections to all de jure standards groups but I agree with the similarities analogy.)

One difference he pointed out: the end product of standards communities is a specification and a test suite for 50-100 vendors. The end product for a successful OSS project is thousands or millions of users. Another thing Jonathan’s Microsoft experience brings to WSO2 is an understanding of what it takes to bring to market a product that might be used by millions of users.

If you want to see the beginnings of how that might work out, look at mooshup.com where the community is getting off the ground. The site is based on the WSO2 mashup server also. The server is available under the Apache license and can be downloaded from mooshup.com and the WS02 site. Consistent with the business strategy for its middleware (web server plus ESB plus new registry), WSO2 will provide service and support of course.

Looking at the functionality, Jonathan positions the mashup server as a natural extension of WSO2's middleware to “service composition.” In his product design for people, he wanted to do something more friendly and lightweight than the Business Process Execution Language (BPEL). I think of it as a poor man’s business process management BPM (watch for my upcoming blog posts on BPM by the way; my initial feature article for ebizQ on BPM is available here for ebizQ Gold Club members—no cost to join). Given his standards background, the mashup server ties into the WSDL standard; like the way a web server works.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

UPDATED 3/6/2008 to better explain why open source helps users if it turns out they don't like what Buytaert/Acquia does with it.

Dries Buytaert is the creator and project lead of the popular Drupal open source content management software that lets an individual or an entire community of users publish, manage and organize a wide variety of content on a website. 800 friends and associates heard his state-of-the-project keynote March 3 at Drupalcon, which is co-located from March 3-6 with the AiiM show in Boston.

I give him high marks for clarity and courage in his presentation.

The over 800 attendees at the Drupalcon conference are mostly designers, developers and webmasters that both use and contribute to the Drupal open source software (OSS) project. Dries said he envisioned a future where designers, developers and webmasters are no longer needed.

He was preceded on the dais by an executive of the John S. and James L. Knight Foundation, which “invests in journalism excellence worldwide” (along with other charitable efforts). The executive had just announced the next stage in the ongoing Knight News Challenge, a program of grants likely to help the Drupal community. The foundation descends from (but is no longer related to) the Knight-Ridder publishing empire. Buytaert said in his speech that he foresaw a future without publishers too.

He provided strong technical and market proof points in delivering the information that the audience presumably would not want to hear. But it was well received. He made it clear that he was talking ‘future,” some point after the Release 7 currently in process and scheduled for a November 2008 code freeze is completed.

Of course, one of the major benefits of OSS is that anyone that does not buy into the Buytaert vision if Drupal can take a fork and move forward in another direction. Related to that possibility, in a press release issued at the conference, Dries also put some substance around Acquia, a company he co-founded in December 2007 with funding from North Bridge Venture Partners, Sigma Partners and O’Reilly AlphaTech Ventures. The company plans to commercialize Drupal along the lines of other for-profit ventures tied to community projects (e.g., Red Hat Enterprise Linux to Fedora, IBM WebSphere Community Edition to Geronimo, Iona Fusion to Apache Software Foundation ServiceMix and ActiveMQ). Dries’ co-founder is Jay Batson, previously founding CEO of open source VOIP software company Pingtel. They have been joined by Jeff Whatcott (formerly of Adobe/Macromedia/Allaire) who will run marketing.

(As an aside, AiiM is now known as the Enterprise Content Management association but we greybeards will always know it as the Association for Information and Image Management.)

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

On Feb 13, Red Hat announced it wanted to “own” 50% of middleware deployments by 2015.

On February 14, Red Hat put some meat on that lofty goal by formally announcing the previously pre-announced JBoss SOA platform and three new open source software (OSS) community projects. Two of the new projects are tied to Red Hat’s acquisition of Metamatrix and the company’s long time relationship with Hyperic respectively.

The third project, code-named “Black tie” because it is a direct shot at BEA’s Tuxedo base, is based on the Arjuna technology which Red Hat acquired in late 2005. (Arjuna had been spun out of HP when it exited the middleware market in 2002, dropping the Bluestone middleware line of which Arjuna had been a part.) That idea makes a lot sense given that many Tuxedo users—who have already been moved from AT&T to Novell to BEA over the years—are now going to be moved to Oracle.

These are the Red Hat middleware products/projects—along with their future versions—that it is going to take to displace long-term installed middleware products such as Tuxedo as well as IBM WebSphere, MQ Series and CICS, BEA/Oracle’s WebLogic and Plumtree, Oracle’s many heritage and acquired (non-BEA) middleware products, TIBCO Rendezvous and subsequent products as well as its BPM line up, Iona CORBA-based heritage middleware, SoftwareAG/WebMethods products, Adobe/Macromedia, Sun/SeeBeyond, and so forth.

Of course the Sun/SeeBeyond example highlights the nature of Red Hat’s challenge: all of these companies are adopting open source terms and conditions—and development models—rapidly, thereby eliminating one of the major reasons a user might want to consider kicking out the long litany of past and present middleware legends in favor of JBoss.

One advantage Red Hat has over many of the more established middleware players is that it has practically built its SOA platform from scratch (with the Arjuna exception noted above). Many of the others have a 40-year technology amalgamation that shares very little in common other than a brand name. Without mentioning their name, Red Hat’s Pierre Fricke rightly criticized Oracle for having both an SOA platform and an event-driven architecture platform. “Why are two needed?” he implied. He could have added: And a BPM platform? And a collaboration platform? And a you-name-it platform? Oracle has all that stuff because of its many acquisitions.

The JBoss Enterprise SOA Platform also supports mix and match a little better than the established players although I do not expect that advantage to last very long. Red Hat announced that active JBoss SOA middleware partners with which its users could mix and match include Active Endpoints, Amberpoint, Information Builders and iWay Software, SeeWhy, SOA Software and Vitria Technology. In particular, Red Hat talked about an interesting BPM partnership with Vitria. This may breathe some new life into one of the original integration middleware providers.

The Metamatrix technology mentioned above will drive a project called DNA related to SOA governance. Red Hat and Hyperic have released into the OSS community a jointly developed management platform project, named RHQ.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

As discussed here last month, one of the recent knocks against open source software (OSS) has been that it simply reverse engineers proprietary technologies at some time period after the proprietary technology becomes popular, in effect inhibiting further innovation and “mopping up” what’s left of a market demand. There are examples of such OSS projects. But it’s not a universal truth.

As I indicated in a recent article, the original Apache HTTP server came on the scene just about coincident with the proprietary Netscape equivalent. “Apache” did not play a mopping-up-old-technology role (although it did clean clock Netscape pretty much) and for a time it did create a separate new middleware market.

The Apache Software Foundation (ASF) Tuscany project also seems equally well timed to current market need. On February 5, 2008, the ASF Tuscany team rolled out the 1.1 release of its Java Service Component Architecture (SCA) project. It’s a runtime environment based on the SCA set of specifications being standardized by OASIS as part of the Open Composite Services Architecture (Open CSA).

The overall effort—both the projects and the standards—are aimed at simplifying SOA application development.

SCA first came out in November 2005 when IBM, BEA and others announced that they were working on it together. It goes hand in hand with the Service Data Objects (SDO) standard focused on uniform access to data residing in multiple locations and formats. In total some 17 suppliers also including SAP, Oracle/BEA, Progress and TIBCO that preach SOA as a design approach for their own product development took it where they wanted to go: developing standards to cut their development costs to increase their profits and/or allow them to reduce cost of ownership and build up the IT market (see more research at my investment research outlet, Research 2.0—button at right).

What it means is that these companies—as they have been trying to since the 1980s via Atlas, OMG, OSF, Xopen and so forth—agreed to put gas pedals under the right foot, agreed that red means stop in brake lights, agreed to simplify translation between metric and imperial measures, and so forth (while at the same time agreeing that they will always disagree on which side of the road to drive on, which side to put the steering wheel on, and so forth).

Then in March 2007, the suppliers, having reached their goals, turned their standards efforts over to OASIS. Enter Apache. The Tuscany project team (admittedly many of whom work for the same companies that originally collaborated on the standard) began working on an OSS instance that met the standard. The Tuscany SCA Java 1.1 release this week adds a number of features to the November 2007 release including a JMS binding, improved policy support and an implementation extension for representing client-side Javascript applications as SCA components.

Tuscany is still an incubation effort at ASF, sponsored by the Apache Web services PMC.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

I have written here and here (see "moving up the stack") about the importance of applications to the growth of the open source software (OSS) movement and the importance of services (or components if you prefer) to applications.

Black Duck combines the trends in an announcement on January 28.

The company calls it the “Black Duck Code Center” and its part of Black Duck’s line up of products that address life cycle management and application development in an era when pieces of code are “flying in” (pardon the pun) from all over under a variety of terms and conditions (Ts&Cs), including OSS Ts&Cs. Black Duck anticipates general availability by March 31.

One issue Black Duck is addressing is like but not exactly the same as the issue discussed recently on the Open Source Initiative (OSI) license-discuss list and with John Roets. OSI members and Roets were discussing a trend where company management, especially legal departments, were not allowing more software into enterprises if it was licensed under the GNU General Public License (GPL) because of some of the license’s Ts&Cs. Black Duck says it is seeing situations where enterprises are not allowing any OSS into the shop--or are worried about its dissemination within the enterprise--for similar reasons.

Black Duck’s functionality is like but not exactly the same as the governance software HP announced the week of January 21 and the OpenLogic Discover software. Black Duck goes broader and deeper than the others based on my quick look. For example, their “search” software does not care whether it’s looking for OSS or not and the knowledge base that results can have other than OSS poured into it.

One key point however: Black Duck itself does not OSS its product at this time.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

Beginning in 2000 or so, brand name enterprise application integration (EAI) and later business process management (BPM) products were supposed to be all about connecting clients, prospects, suppliers, manufacturers, etc. all along the supply chain. But IT staffs tended to use them for internal integration instead, in places where less complex technology would work just as well. It's like the old saw that casual computer users use only 10% of the functionality of Notes, Office and similar personal-productivity tools.

I was reminded of this tendency of IT folks to overcomplicate their lives when I was researching ebizQ's latest article on open source software (OSS). As part of the research I talked to John Roets of createtank, a consultancy that has open sourced an interesting little framework named "elemenope." Interesting little name too.

We're not talking IBM SystemView or Sun's Distributed Object Environment here. In fact, just the opposite: keep it simple if simple is all you need is Roets' motto. Developed around the same time as these more comprehensive but now long forgotten frameworks (and before SOAP), elemenope follows the same design philosophy in terms of abstracting away from the underlying technology. It includes:
• A repository of connectors (might be XML RPC, might be SOAP call)
• A little engine that abstracts the transport
“Forced abstraction” keeps users from writing spaghetti code every time they add a new application or data source. There's “one pathway no matter what the logic is,” says John. And the next time you need to integrate something, you don’t have to go find the guy that wrote the original service.

Just to achieve this simplicity for its consulting practice, createtank wrote elemenope in the mid 1990s. They OSS'd it under the GNU General Public License (GPL) in 2003 and also under the Apache license in 2006. (More on that in my next blog post; perceived legal problems with GPL are a big issue over on the Open Source Initiative license-discuss list this week.)

By the way, there's more detail on elemenope as well as activeBPEL, Apache ODE, Intalio, Jitterbit, jBPM, Open EAI, Shark and Sun in the feature article on OSS referenced above. It's is available over there to the left of my bald head. (You need to join the ebizQ Gold Club but there is no charge.) The premise of the research is the possibility that OSS integration server software might change the dynamics of external EAI/BPM use in the next decade in the same way that OSS application/web server software (see Open Source at the Core) changed "internal" application integration functionality in the last decade. Let me know what you think of the theory.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

It’s services oriented architecture (SOA) “framework” time in the open source software (OSS) community. For more on how the two buzzwords—SOA and OSS—interrelate, and even what they really mean, see my related feature article from November 2007. The new ingredient this year is the buzzword, framework. OSS-heritage players in the software market such as Mulesource, Red Hat (RHT) and WSO2 are trying to catch up with the proprietary-heritage OSS hybrids such as IBM (IBM), Oracle (ORCL), TIBCO (TIBX) and BEA (BEAS) in the SOA “platform” game. Proprietary-heritage OSS hybrid Iona made a similar announcement on December 12, 2007.

(Just for the record, our research shows users don’t care a wit about whether a company is an OSS pure play, an OSS-heritage OSS hybrid, or a proprietary-heritage OSS hybrid. But since the community blogosphere seems to like to keep track of such OSS virginity issues, we’ll play along.)

On January 14, WSO2 announced the debut of the WSO2 Web Services Framework for Ruby (WSF/Ruby), an OSS framework for providing and consuming Web services in the Ruby object-oriented programming language. On January 15, MuleSource announced a new subscription-only “enterprise-edition” packaging of the Mule Enterprise Service Bus (ESB) along with open-source governance and monitoring tools into an integration platform. (The announcement of an enterprise edition formally pushes Mulesource into the hybrid category decreasing the number of OSS pure plays further.) Given Iona’s December announcement of FUSE HQ, a solution based on OSS Hyperic that provides systems management and monitoring capabilities across all products in the FUSE family, we expect Red Hat to come in batting clean up any day now. Red Hat had previously promised a late fall rollout of its JBoss-centric SOA Platform when it announced JBoss Enterprise Application Platform 4.2 in July 2007. It appears the leading OSS provider has instead waited in order to make the best feature/function response possible to the smaller organizations.

Everyone now perceives having a platform to be preferable to just offering some of the parts, such as an enterprise service bus (ESB), a registry/repository, or toolkit. Of course, as I said in my review of OSS in SOA, and as I said about the advent of the enterprise service bus in 2002, and as I said when proprietary vendors announced similar concepts 10 years ago using terms like Internet Application Components, S-O-A won’t matter much until we get some S’s.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

I caught up with Bryan Che, product manager at Red Hat, to get some background about Red Hat’s Enterprise Messaging/Real-time/Grid (MRG) concept, the distributed computing framework pre-announced by Red Hat December 4, 2007 as a follow-on to November’s Linux Automation announcement.

Linux Automation—based on Red Hat Enterprise Linux (RHEL) 5.1 and its virtualization features—provides for seamless application provisioning/management/monitoring across both physical and virtual servers as well as appliances and “the cloud.” MRG, when available in 2008, will sit on top of Linux Automation and provide similar provisioning/management/monitoring for a grid. The Messaging and Real-time functionality will be supported separately as well (and all of the above is available already in respective OSS communities).

With this announcement, Red Hat is competing as directly as ever against Oracle’s Fusion middleware as well as Oracle’s grid computing support, its RHEL clone, and the recently announced Oracle VM; IBM WebSphere MQSeries, WebSphere Real-time—built on the same Linux extension as Red Hat Real-time—and IBM grid capabilities; as well as (at least for the Messaging component), products from BEA, SoftwareAG, Sun and Tibco. The real-time component appears to be based on development work performed jointly with IBM; the IBM version was made generally available in August 2006.

Of course, I was primarily interested in the OSS aspects of the announcement. To steal a line from Murat Aksu of Zenoss in another context, “This is not your grandfather’s open source software (OSS).” I wondered
(1) why the messaging component community was over at Apache rather than in Red Hat’s own JBoss.org and
(2) where all the other pieces were coming from and how they were being licensed.

The answer to the first question is a matter of both timing and strategy. Some of the work started before Red Hat acquired JBoss but also it makes sense that the messaging component not be as dependent on one middleware stack as most JBoss.org work is, at least implicitly.

As for the answer to the second question, the answer is too long for a blog post. It involves Apache, JBoss, the University of Wisconsin, IBM as noted above, and many more communities and license terms and conditions (Ts&Cs). The complexity of putting the framework together simply from a Ts&Cs perspective demonstrates the advantages that a profit-motivated provider such as Red Hat brings to OSS.

It also demonstrates why OSS is more a development model than a business model as discussed here.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

Following up on the initial Red Hat Enteprise Linux (RHEL) 5 announcement in March 2007, Red Hat filled out its stack strategy on November 7. The company extended the previously anticipated release of RHEL 5.1 to a concept called Linux Automation. Linux Automation provides for seamless application provisioning/management/monitoring across both physical and virtual servers as well as appliances and “the cloud.” The virtual server support was first made available with RHEL 5 and Red Hat says it has deployed to 18,000 servers in under six months.

Red Hat Enterprise Linux 5.1 is immediately available to customers via Red Hat Network, Red Hat's own management automation platform. To support the broader Linux Automation concept, Red Hat will also make an Appliance Operating System available in 2008 and also said it struck a deal with Amazon making RHEL the first commerical operating environments supported on the Amazon Enterprise Cloud 2 (EC2). The latter is what the two companies are calling a “private beta” at this point. Other “clouds” can be supported also.

So whether a user or ISV is deploying to physical dedicated servers, virtual servers or the cloud, the application runs and can be managed the same way. When the appliance product is rolled out, it will integrate JBoss and other Red Hat products as well. The appliance of course will be moved to market via ISVs, who incorporate their own solution or solutions into the appliance. The cloud option is relevant to those that want to make their software available as a service (SaaS)

More interesting than this ambitious any-app/any-where/any-time strategy is an ambitious objective/goal. Red Hat predicts that this stack strategy will let it double its market share, and that it will end up driving over half the world’s servers by 2015. I'm not sure if that is an installed-based or shipped-that-year metric but the mothballing of a lot of Digital UNIX, AIX and HP-UX systems will at least support the latter. In addition Red Hat sees its products becoming a mainstay on the mainframe as well. Also contributing, there will be a lot of justifiable double counting as stacks like Red Hat's (and so far it's the only real game in town) forms the underpinning of virtualized Windows and VMware servers.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

The Apache Struts group has announced that Struts 2.0.11 is a "General Availability" release, which is the highest quality grade of Apache Softwre Foundation (ASF) open source software (OSS) release. But don’t make the mistake I made and think that 2.0.11 is a next-generation replacement for various Struts 1.x.x releases. They are separate animals, like donkeys are parents of but different than mules (not sure the Struts group will like that analogy, nor will Mulesource). Both are popular Java web frameworks along with MyFaces, Tapestry and Wicket.

Struts 1.x.x is a control layer based on standard technologies like Java Servlets, JavaBeans, and other technology. It helps users create an extensible development environment based on published standards and the proven Model/View design pattern. Struts 2.x.x is more comprehensive and designed to streamline the full development cycle, from building, to deploying, to maintaining applications over time. Apache Struts 2 was originally known as WebWork 2. After working independently for several years, the WebWork and Struts communities joined forces to create Struts2. WebWork continues to deliver patch releases. There is also still a robust and vibrant community of developers using Struts 1 in production, according to its site, and it expects that thousands of teams will continue to base new projects on Struts 1, and continue to support existing projects, for "many, many years to come."

Of course, if you are starting a new project, and have your choice of frameworks, this might be a good time to consider whether you would like to continue to use Struts 1 or whether it's time to try Struts 2. The Struts group says there are "five reasons to migrate:
• "Your Struts 1 application is ready for its own version 2.
• "Your Struts 1 application is still under development or in the planning stages.
• "You inherited a Struts 1 application that needs work, and you don't understand how it works.
• "You would like to integrate a few Java Server Faces (JSF) components into your application.
• "The charming quirks of Struts 1 have become downright annoyances, and your team is ready for an elegant yet familiar solution. "

But the community also suggests five reasons not to migrate so check out the Struts home page.

At least the term Struts is now only the umbrella for two frameworks. According to one of the founders of the community, it also used to include Apache Shale, a web framework based off JSF. Although the Apache Struts project has two major versions, both are action (as opposed to component) based. Just remember, JSF is a "component" framework. If you are more into JSF, Struts 2 supports JSF components.

The 2.0.x series of the Apache Struts framework has a minimum requirement of the following specification versions: Servlet API 2.4, JSP API 2.0, Java 5, Backported Java 4 JARs are also available.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

I often get the question whether open source software (OSS) is “ready for prime time?” I typically use Google as an example and recite the litany of key line-of-business processes that Google has that are dependent on various types of OSS. Last week, an old-fashioned proof point was provided by the venerable Transaction Processing Performance Council, keeper of the holy grail of TPC benchmarks, the successor to the even more venerable credit/debit benchmark. Think of TPC as the people who run the ACID tests (as in are your transactions “atomic, consistent, isolated and durable?”).

Through the TPC, IBM announced that it has received what it claimed to be the highest TPC-C performance result ever achieved by a 4-processor server. The IBM System x3850 M2 was running DB2 with Red Hat Enterprise Linux 5 Advanced Platform and 64-bit IBM DB2 9.5. The configuration achieved 516,752 tpmC on the benchmark. IBM System x3850 M2 with the Quad-Core Intel Xeon Processor X7350 2.93 Ghz (4 processors/16 core/16 threads) has an availability of March 14, 2008 (although the Red Hat part of the mix is available now).

Of course like MPG stickers on cars, TPC benchmarks come with caveats. As the TPC says “You must understand what the benchmark is intended to measure, before you can understand throughput. Throughput, in TPC terms, is a measure of maximum sustained system performance. In TPC-C, throughput is defined as how many New-Order transactions per minute a system generates while the system is executing four other transactions types (Payment, Order-Status, Delivery, and Stock-Level). All five TPC-C transactions have a certain user response time requirement, with the New-Order transaction response time set at 5 seconds.” Therefore, for the 516,752 tpmC number cited by IBM, a system is generating over a half-million New-Order transactions per minute while fulfilling the rest of the TPC-C transaction mix workload

Now the thing this OSS agnostic and open choice advocate admires is that this high-performance transaction load is monitored not through the equally venerable IBM Customer Information Control System (CICS), not through ATT/Novell/BEA Tuxedo, but through Microsoft Com+ (within Windows Server 2003 Web Edition) sitting out at the configuration’s 16 theoretical desktops. Whatever it takes should be the motto of IT folks.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

I promised some time back to look more at open source software (OSS) development tools.

So far, I have failed in that objective because of all the interest in
-- Microsoft "going OSS" (in my IT investement research analysis, I have even hypothesized about Microsoft acquiring Red Hat and Steve Ballmer accomodated my hypothesis at the Web 2.0 conference today by saying "anything's possible")
-- so-called "Linux patent battles" (that have nothing to do with OSS)
-- the need for more applications in the OSS space to keep the OSS ball rolling.

So a press release that crossed the wire today from Evans Data broght me back to more important stuff such as what do software developers use and why. Evans' recent survey says the Eclipse OSS development framework is an up and coming means to do service oriented architecture (SOA)-based development.

Actually what the survey found was that the Genuitec's MyEclipse IDE was gaining in popularity, and because MyEclipse supports primarily OSS tools... ispo facto and all the Latin stuff. To be clear, MyEclipse is not OSS; you need to pay a whole $50 a year to subscribe to it. But here are all the OSS distributions included with MyEclipse. Not surprisingly, Evans says, Visual Studio is still the most popular integrated development environment (IDE) and IBM's Rational, Oracle's JDeveloper and BEA's Weblogic Workshop are popular SOA developement tools as well. All have OSS tie-ins but are not OSS themselves. Genuitec, Oracle and of course IBM are on the Eclipse Foundation board.

All this development talk also reminds that I will be presenting a discussion of OSS in SOA on October 30 here on ebizQ.net along with many other great sessions as part of ebizQ's annual SOA In Action virtual tradeshow. To register, click here.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

Per a post I put up back in April, I said that one of the biggest issues in blogging about and researching open source software (OSS) is keeping track of the products, communities, and commercial backers. At the time I said the scorecard was timed to the beginning of the baseball season so now here is a revision timed to the beginning of the baseball playoffs.

There's pretty much a one to one to one relationship among the products, communities and backers but that relationship is not always obvious. Even for the simplest connections. For example, it's probably intuitive that the Mozilla portal is built by the Mozilla Foundation which is backed by the Mozilla Corporation. But it is probably not obvious that whole string really unravels all the way back through AOL to the original Netscape browser. As for something more convoluted but not unusual: the Sleepycat OSS database project is really a derivative of the Berkeley database that forked out of UNIX at System V time but the community is now backed by Oracle, which purchased Sleepycat in 2005.

Some OSS has a community but no commercial backing (such projects don't usually last too long). Others have commerical backing but no community. Sometimes, that is intentional and the product is really freeware rather than actual OSS. But in other cases, it is is because the project is just getting off the ground.

So to keep the players straight I have updated that scorecard. It is available as a free download by clicking where it says "Download file" immediately below. The taxonomy behind the scorecard is explained in more detail in this ebizQ.net feature article released in July. The article includes a short version of the OSS scorecard.

Let me know if there are other OSS projects you would like to see added--yours or something you want me to research.

Download file

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

IBM has resurrected the old Lotus brand Symphony to launch its Sun OpenOffice-based open source office suite. Although termed an office suite and hyping collaboration on its own web site, neither Symphony nor OpenOffice has an email component the way Microsoft Office comes with the Outlook email client or the way DEC All-in-One and Data General CEO and all the real original albeit proprietary office suites worked (which is why Ozzie wrote Lotus Notes, right?). It's an interesting choice of brandname since the original was chosen by John Dvorak in 2004 as one of the 10 "worst software disasters" in industry history. Symphony also includes some of same tools that are inside Lotus Notes 8, which can be used to extend a business process or to create dynamic composite applications. The product supports Microsoft and Adobe formats as well as the Sun OpenOffice format.

So unlike when IBM entered the open source software (OSS) application server market with its purchase of Gluecode in 2005, IBM is not competing with itself with this OSS move. You can use any email client and server and IBM hopes you will chose Lotus Notes and Domino. Of course, the documents will not appear in the email client in an integrated fashion the way Word is usually used as the text editor within Outlook. This appears to be an effort by IBM to bolster its OSS bona fides, particularly related to the Sun OpenOffice Open Document Format (ODF) format. In fact, you can now call it IBM/Sun Open Office because IBM (according to press reports) said that it will contribute 35 developers to the Sun OpenOffice project.

And this move by IBM illustrates how important applications are to the eventual success of OSS. Despite all the fits of pique in the OSS community over Linux vs. Windows, the thing that matters to real people is how they read their email, type their personal and business correspondence, pay their bills at home, plan their production lines at work, and so forth. Real people could care less if there is Linux in their Tivo or Symbian in their cell phone or Windows in the cash register at the corner convenience store.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

On Thursday August 9, ebizQ presented Mark Driver of Gartner discussing the relationship of open source software (OSS) to services oriented architecture (SOA). You can still listen to the webinar but you missed the chance to participate. So the purpose of this post is to provide a reprise of the key issues raised in the session. By the way, Red Hat sponsored the event and some of the questions were answered by Burr Sutter, Red Hat's director of product management, who was also on the panel.

Attendees wanted the panel to talk about the difference between OSS and open standards. One felt that participating in the SOA standards movement would provide his or her enterprise stronger investment protection than access to particular code, or the right to distribute it. Good point and it represents the different sides of the same coin.
-- OSS guarantees enterprises access to the code and the right to upgrade/improve that code even if the company or community that provided it goes out of business or disbands. For more details about our view of the OSS world, see the Features section under this tab and this taxonomy document in particular
-- While standards can provide a similar advantage, IT history shows us that standards are more about least-common-denominator functionality, not usually what an enterprise wants if the code is providing competitive advantage. But SOA standards are critical to intercompany activity such as supply chain management or CRM among business partners. If that is more important to your IT strategy than internal code that supports competitive advantage, working with the standards movement would be a better way to support IT community.

There tends to be such a standards movement in every industry. Of course, if you have the resources (who does these days), you can do both.

Another question was, “What's the driver for business to adopt OSS-based SOA?” These are two different questions really (and then I’ll try to put the answers together in the next paragraph). The drivers for the architecture are the ability to build utility-like computing for your enterprise or organization with benefits such as service level agreements, grid infrastructure for protection against “power outages,” unit-based metering of compute power and more. The drivers for the OSS development model are the protection against a company or organization disbanding as described above, faster bug fixes and other quality assurance facets, and less vendor lock-in.

In short, SOA is an architecture and OSS is a development model and the two together complement each other with benefits that do not overlap.

The $64,000 question was “Mission critical apps are often considered too important to put on OSS platforms. What are some of the arguments against that view? As I said in the session, the flip answer is “Google.” Both Burt and Mark went into more detail. I don’t believe anyone is peddling the canard that OSS is not ready for primetime anymore but if you hear it, consider all the companies and organizations that have already crossed over from a proprietary UNIX distribution to Linux. There are similar examples up and down the stack.

Overall it was a great session with another half dozen questions like the above. To “attend in arrears” and at least hear the answers directly, click on the Webinars tab on the upper left and then click on the Archived Webinars link. Look for the 8/9/2007 webinar titled “Optimizing Open Source and SOA Strategies.” You’ll have to sign up for the ebizQ Gold Club if you have not already; it’s free.

And if you still have a question you wished you’d been there to ask, do not hesitate to email me at dennis at ebizq-dot-net. And save the dates October 30-31 for ebizQ's two-day virtual conference on SOA In Action where we'll talk move about the convergence of OSS and SOA.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

One of the key tenets of the open source software (OSS) culture is the belief that working with a diverse developer set provides better quality software than working with a homogenous group of developers at a single company. That does not mean that you don’t still need a process to help assure that theoretically better quality and there’s the rub.

I was recently pointed at the web site, openQA.org, and hooked up with Patrick Lightbody, founder and prime mover of openQA. Patrick is also QA solutions product manager at Gomez, which provides on-demand services to measure and manage website and web application performance and the customer web experience from design and development, through deployment to production.

Quality assurance (QA) processes are as old as the software development process itself. Patrick points out that QA has the same kind of second-class-citizen characteristic in OSS development that it has always had in the proprietary software world. Except the problem is worse because there is no boss to say, “You will do the QA.” OpenQA aims at helping OSS communities and companies that by definition lack the benefit of “the boss” deciding who will do the QA.

OpenQA combines corporate sponsorship and a community approach much like an Apache for QA. It includes projects such as Selenium and its offshoots for rich dynamic web testing and smaller projects such as Watir, which stands for web application testing in Ruby. OpenQA software is distributed under an Apache-type license. Selenium was originally a Thoughtworks project designed to support its consulting practice. A few other contributors added Firefox plug-ins as recorders, some autocomplete features, breakpoints etc. to Selenium leading to Selenium IDE. Even some BEA QA specialists added some code. Given Patrick’s day job, not surprisingly the Gomez Reality Check on-demand service is OSS-based and the code has been placed back into the OpenQA community as Selenium RC.

There are other standalone OSS QA projects such as the PushToTest tools (pushtotest.com) distributed OSS under the Gnu General Public License version 2 (GPL V2), and the WebLoad community founded by established QA software provider, Radview. WebLoad is also licensed under the GPL.

It used to be in analyzing the health of the manufacturing industry in the U.S., you looked at the tool and die makers as leading indicators. I am thinking that the uptake of OpenQA and these other QA-oriented OSS projects may be a leading overall indicator of OSS success as the business model continues to mature in synch with Web 2.0 application development. If OSS is to be adopted by the large enterprises that spend the vast percentage of IT dollars, those enterprises are going to want assurances.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

I have shared my opinion on this blog (and provided more detailed research over in the Features section of the ebizQ Open Source tab) about the importance of building up the open source software (OSS) applications catalog. I'm talking industrial-strength enterprise applications. If OSS is to continue to grow as a development model for the largest IT suppliers and as an entree into the IT market for smaller suppliers that need a minimal barrier to entry, that growth will come in the applications/solutions area of the market.

That's because surely there will not be another OSS operating system and the OSS middleware stack is consolidating almost as rapidly as the so-called proprietary middleware stack. I say "so-called proprietary" because much of the stack from the large middleware suppliers is already OSS. And consolidation from a marketing perspective means there is no money to be made.

That leaves the solutions/applications space as the area of OSS growth potential.

Microsoft realizes that solutions will be the battlefield for the next 10 years and is gearing up to out-Eclipse Eclipse with the release of Longhorn as its next-generation platform and the next generation of Visual Studio. Microsoft plans for its partners--the melding of the Business Division partners that work with Great Plains/Navision/etc. and the group known as Windows Solutions Providers--to develop multiple solutions in every one of the theoretical 9,999 industry niches represented by the North American Industry Code (NAIC) at the four-digit level.

Some of these Microsoft solutions (or components or services if you prefer) will be standalone business intelligence and transactional applications. Others will be add-ons to Office and Dynamics. Some will be somewhat monolithic and others will be mash-ups. Some will be licensed in the sense we have come to know for the last 20 years and some will be available as a service (SaaS). Some of the SaaS offerings will be monetized by advertising, some by paying for related content (e.g., tax tables) and others by support arrangements. But all will be seamlessly architecturally compatible and interoperable. All 15,000-plus of them.

The OSS world by design and culture does not have such a unified launching pad for the thousands of industry-centric solutions that the market demands and the OSS community could provide. It has the LAMP stack but as I have illustrated elsewhere that stack is as often WAMP or LAOP. The Open Solutions Alliance (OSA) could be the change agent that puts OSS on equal footing with the Microsoft ecosystem and helps preserve the open choice that many of us would like to see.

More than a LAMP stack or replacement, OSS needs a WebSphere or NetWeaver equivalent, a stack with more muscle than a low-level middleware layer and database access on which to base solutions. The OSA recognizes that need but is not quite ready to step up to the plate to provide it. But this week at LinuxWorld the OSA did come out into the on-deck circle. It introduced the first interoperability scheme in its ambitious Interoperability Roadmap. The OSA’s debut "prototype" includes specific standards and best practices for delivering a Common Customer View (CCV) across open applications. And the OSA has initiated projects in business process management, identity management, service level agreement action, and in other application middleware categories.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

On Thursday August 9, 2007 at noon eastern U.S time ebizQ will present Mark Driver of Gartner discussing the relationship of open source software (OSS) to services oriented architecture (SOA). Everyone's invited and encouraged to participate with questions. Red Hat is sponsoring the event and Burr Sutter, Red Hat's director of product management, will also be on the panel.

I have written of and commented on the investment implications of SOA at Research 2.0 and elsewhere (the link is just one example; for more, google "Dennis Byron" filtered by "SOA").

Of course, I dig down into OSS here at ebizQ.

Mark ties the two together in a way users, investors and suppliers will want to hear. Why didn't I think of that?

To attend, click on the Webinars tab on the upper left and then push the button with Mark's picture beside it and the title "Optimizing Open Source and SOA Strategie" If you can't make it on Thursday, you miss out on asking a real industry expert your most pressing questions but the webinar will replay here on ebizQ for the foreseeable future.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

Construction- and architectural-industry analysts don’t spend a lot of time on hammers and saws; IT-industry analysts don’t spend a lot of time on integrated development environments and frameworks. Yet if you like to design and build physical things with wood and metal as I do, you quickly realize why you have to have both claw and ball-peen hammers and a wide selection of teeth-per-inch in your saw blades. That is a weird way of saying that I have not paid enough attention to open source software (OSS) development tools in this blog space. This is primarily because shortly after I trained as a COBOL programmer in 1969/1970, I became an IT marketeer and never did anything really productive.

That is going to change in a couple of ways this month. Maybe you have noticed that Open Source now gets its own tab right above these words and on the ebizQ.net home page. Along with SOA, BPM, Security and so forth, Open Source coverage will be expanded and part of that expansion will be a good hard look at the tools available to build OSS applications and deployment software, such as operating systems and middleware. In fact, I plan to give OSS tools the same “weight” of my attention as I have given packaged applications and middleware/Linux as I continue to blog under this new format and also add feature articles that go into more depth than a blog post can.

So the timing of this Red Hat story last week and my change in emphasis was interesting. When I talked to Tim Yeaton of Red Hat back in May, he suggested I get up to speed on Red Hat’s relationship with Exadel in order to better understand where they were going to help developers. There was speculation in the press at the time that Red Hat would be acquiring some of its solutions partners but he said look at Exadel to see Red Hat's direction.

I had that meeting last week just as Red Hat was expanding its developer-centric efforts. JBoss.org, which is to middleware at Red Hat what Fedora is to operating software, created the JBoss Tools project, a step toward delivering Red Hat Developer Studio, an entirely OSS integrated development environment (IDE), later this summer. Exadel, an OSS professional services firm, had contributed Eclipse plug-ins, Ajax4jsf and RichFaces in March 2007 to get the ball rolling on the OSS IDE. (If you can’t wait, starting last week, developers can download plug-ins from JBoss Tools to compile their own IDE with tooling for Java EE and Ajax development.)

Not being a developer guy, I would have thought Eclipse already delivered on that need but Red Hat believes that pieces of Eclipse above "core Eclipse" are not OSS and that effectively that makes Eclipse just a framework rather than an IDE. (In addition to a definition debate between framework and IDE, note from this reference in a recent post that we will also be following a growing debate about what is truly OSS, what is hybrid OSS—and whether that matters—and what is out and out freeware masquerading as OSS.)

One thing I will be looking for and that my marketing development/investment research background says should be a natural is a connection between the Red Hat Developer Studio and the Red Hat Exchange, Red Hat’s growing list of solutions provider partners. If Exchange members use the Studio, it is going to be easier for other community members to add on to and expand the Exchange catalog. If the OSS movement is going to continue its momentum, it needs to keep pace with activity at Microsoft, Oracle, SAP and elsewhere to make thousands of OSS minisolutions available to the community (also called objects, applets, enterprise application components, services, and so forth over the years).

Red Hat says stay tuned. We ask you to stay tuned as well as ebizQ.net turns up the volume on OSS. Click that XML or RSS button to the right right now.

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

Investors have to make a decision whether to consider Google a marketing/advertising agency or a media company. But as an open source software (OSS) user you probably care more that they are technologists at heart. And among the biggest users of OSS to boot. Look no further than CEO Eric Schmidt discussing the LAMP stack with financial analysts back on March 5 and 6 and this week when Alan Eustace and Jeff Huber, engineering VPs at Google, spoke to an “adoring crowd” at a Goldman Sachs investor conference. Just letting two engineering VPs loose on Wall St. is signal enough but they made it clear, saying: “We are a technology company” and “Google is a software company that should be measured against its peers in delivering user experience,” when asked some question about return on investment metrics.

Therefore it is no surprise that next week, on May 31, 2007, Google offices in ten countries will host Google Developer Day, a global virtual event featuring workshops, keynotes and breakout discussions on Google's APIs and developer tools. Focusing on the theme "Building Blocks for Better Web Applications," Google Developer Day will explore innovative uses of Google developer products to create and enhance applications, and integrate with Google services.

Google Developer Day will take place at Google offices and offsite locations in Mountain View, Sao Paulo, Madrid, London, Paris, Hamburg, Moscow, Beijing, Tokyo and Sydney. The proceedings will begin in Sydney on May 31, 9 a.m. AEST and end 27 hours later in Mountain View on May 31, 7 p.m. PDT. Confirmed presenters include Guido Van Rossum, Google software engineer and creator of the Python programming language (Beijing); Chris DiBona, Google open source programs manager (London); Mark Stahl, Google data APIs tech lead (Madrid); Bruce Johnson and Joel Webber, co-creators of the Google Web Toolkit (Mountain View); Bret Taylor, group product manager for Google developer products (Mountain View); Lars Rasmussen, Google Maps senior engineer (Sydney); and Greg Stein, Google engineering manager and chairman of the Apache Software Foundation (Tokyo).

To reach developers everywhere, Google will offer live streaming webcasts from its Mountain View office and provide a YouTube channel with videos of Google Developer Day sessions at other sites. For registration and additional information, see the Google Developer Day website: http://www.google.com/events/developerday

Posted by dennisb in OSS Development | Permalink | Comments (0) | TrackBacks (0)

Most people “love” lists. The psychology of lists has been proven somewhere (no URL provided). On the other hand, analysts love surveys. So I spent a little time digging into the factoids and finer points of portions of a user survey released this week by open source enterprise service bus (ESB) supplier MuleSource. I know from my years at Datapro and IDC that IT users like IT research too. (Perhaps they are normal and like lists as well). For IT staffers, the idea is to see what other IT guys are doing. Unless they are out and out contrarians, like analysts, there is benefit in knowing which technologies peers are betting on.

Take a look at Mulesource’s data yourself. The company gives you its opinion on what the results mean, and I give you mine below. In this world of open source, research and analysis are becoming open source as well. You might also want to look at my opinion on Linux Leveling Off, posted at Research 2.0. It’s based on IDC quarterly data released to the public this week.

A couple of cautions: Only 34% of the respondents were from the U.S., making the mix a little suspect as compared to overall software usage (the U.S. accounts for about 45% of spending, according to publicly released IDC data). Mulesource didn’t publish the survey instrument so I don’t understand some of the tables provided out of context. I also didn’t see any indication of a statistical significance percentage but I suspect it is “wide.” Most important, this is a survey of Mule users so what they say they are doing with Mule cannot in any way be inferred to Sonic, Fiorano, Logicblaze, or other ESB users. But unless Mule ESB users are Vulcans or something, the non-Mule factoids below probably have some universality to them.

According to the release, of Mule ESB users:
• Almost 30% already have their ESB in production; I find this percentage high for any relatively new technology (the ESB in general, not just MuleSource’s) based on my research into the adoption of client/server, ERP, ecommerce, CRM, and so forth over the years.
• About half are “community” oriented, in that they give back custom add-ins to the market; this statistic is probably what is driving the other numbers indicating higher than expected (by me) early adoption. That would be an interesting change in behavior caused by the OSS movement if my theory is correct.
• A major Mule usage is custom integration with legacy apps according to Mulesource’s analysis but I find it hard to reach that conclusion without the survey instrument and without the statistical significance information; clearly is it an important use along with building (presumably new) self-service apps and portals
• Mule users are overwhelmingly JAVA guys, claiming a large adoption of service oriented architecture (SOA) principles
Not coincidentally the group reports big Spring application development framework usage, another possible statistical anomaly.

What I find most useful in such supplier-sponsored and technology-restricted research findings are the factoids separate from the restriction (that is, Mule usage in this case). As I implied above, unless there is some weird correlation between Mule usage and otherwise aberrant behavior, these findings should be more representative of the IT universe as a whole:
• BEA WebLogic, IBM WebSphere and Red Hat JBoss were statistically tied (whatever the precision of the sample) as preferred application servers; Oracle AS trailed badly. I suspect this finding is based on the company not getting a representative mix of database users in their surveying methodology; if they did not have about 66% of their sample running Oracle RDBMS, they did not have a good sample of the IT universe
• 50% are using integration techniques (web services specifically) to connect to customers and suppliers, not just to connect within their own enterprise
• A large percentage of respondents are using MQ Series as transport (although JDBC was leading transport/protocol); in a related question, as many respondents said they were using MSMQ as using TIBCO for actual queuing; Oracle AQ scored well too.
• Following conventional wisdom, there was a high use of Active MQ and JBoss messaging for queuing. But defying conventional wisdom, 60% of these presumably avid OSS fans are working with Windows

Mulesource claims to be the “the leading provider of open source infrastructure and integration software.”
Be careful however. The company uses the now debased JBoss-driven metric of downloads to characterize its market acceptance. I call it the “free AOL CD in the mail” metric. Once JBoss numbers became “public” after its acquisition by Red Hat,