In another life I did marketing communications. This was pre-Web 2.0 of course. In fact it was pre-Internet and even mostly pre-email so the rules have changed. But I can't believe they have changed so much that any PR agency would recommend announcing something on the Friday before the long Canada Day/Fourth of July(really long this year in the U.S.) weekend in North America or the day before the beginning of the two-month-long European holiday season. Therefore, not surprisingly--contrarian as a matter of a principle--the Free Software Foundation (FSF) today formally released the GNU General Public License Version 3 (GPL v.3) at its Boston headquarters.

Since there was nothing really new news in the license itself, the interest is in what others say. So far:
- Red Hat--We're thinking about it for the future
- Samba - "a great improvement over GPL v.2"
- Linux Foundation- nada

The press release was loaded with the usual FSF speak (treacherous instead of trusted as in trusted computing, restirictions instead of rights as in DRM, GNU/Linux specificity in contrast to just Linux, and so forth) and attacks against patents. But it was pretty tame overall.

I have commented here and at Research 2.0 (see link to right) in different ways about the fact that the attacks on whatever the opposite of free software is (priced software, proprietary software, closed software, or I think the FSF has settled on the term "non-free software") fall on deaf ears. Those suppliers that have traditionally priced their proprietary software are rapidly moving to software as a service anyways. Maybe not coincidentally, Microsoft Deputy Counsel Marshall Phelps was in Boston yesterday at the Red Herring East conference. One of his opinions was that the monolithic operating system is fading away along with the means by which it has been licensed. Of course, he was speaking about Windows but the same could be said of GNU/Linux.

So, happy Canada/Dominion Day, Fourth of July, Bastille Day, and so forth. When we all get back, I will talk about what Mr. Phelps has to say about patents.

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

There was a lot of open source software (OSS) under the covers of the new Internet B2B and B2C services introduced at Red Herring East this week in Boston. (There are a couple of more 2007 Red Herring conferences on tap in 2007 if you are interested.). And there was one out and out OSS firm—Apatar— that debuted at the meeting, which is a traditional venue to get a company story in front of movers and shakers in the venture capital world.

Apatar is bringing to market what it calls “Enterprise Data Mashups,” which are data integration software tools. The development model is OSS with what I think of as the Compiere twist (see the Talking with Compiere entry from March). Like Compiere, Apatar is probably more interested in getting the community to extend the product as opposed to playing with the core code. What Apatar wants most from its community is to add to the already over 100 pre-existing metadata maps it already has in its library. Also like Compiere, it doesn’t mind if the community includes power end users as well as traditional developers. Apatar says it is the first to release OSS for power users; I don’t agree with that but with the computer sophistication of hardcore analytical apps guys, it may be the first to make such an emphasis work. Wide end user participation is more likely to happen for Apatar than it is for Compiere.

The Apatar distribution model is primarily software as a service (SaaS) although it offers a MySQL-like two-tier licensing scheme as well. The target market is the millions of data integration projects that never get done because the return on investment is too low (or negative) using standard extraction/transform/load and integration software. Apatar lets users link information between files (Microsoft Excel spreadsheets, CSV/TXT files), databases (such as MySQL, Microsoft SQL, Oracle), applications (Salesforce.com, SugarCRM), and Web 2.0 destinations (Flickr, Amazon S3). Being able to work with RSS feeds is a nice functional plus.

We will be “talking with” Renat Khasanshyn, Apatar founder and CEO shortly. If you have some questions for him, let me know.

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

The Linux Foundation (LF), the nonprofit organization dedicated to accelerating the growth of Linux and funded by Google, H-P, IBM, Novell, Red Hat and others, has announced highlights from its June 13-15 Collaboration Summit held on Google’s campus. The meeting included kernel developers, distribution and system vendors, ISVs, end users and open source software (OSS) community project leaders. It was said to be an invitation-only event (which would be very un-OSS of them) but I think the only thing you needed to do to be invited was to join the LF.

Most key Linux developers are employed by the companies that fund LF. The rest of the OSS community is scattered all over the IT ecosystem. LF said that is was the first event of its kind where a cross-section of the Linux and the broader OSS communities met face-to-face. Many resulting press stories seemed to highlight industry turf battles and OSS political correctness but the group’s own press release highlighted key pieces of necessary nitty gritty.

For example:
• Representatives from the LF’s Accessibility workgroup explained the Linux model for writing accessible applications; accessibility of OSS to persons with disabilities was a key limitation with the first releases of the open document format (developed and promoted by many of the same companies sponsoring LF) and the group wants to avoid a repeat. This activity is near and dear to me because I severely tore a retina about two years ago.
• A collaborative problem solving session on device drivers was held. The session explained the Linux model and the newly formed program available for vendors where the Linux kernel community is offering all companies free Linux driver development.
• Specifically, key representatives from major printing vendors met with Linux community leaders to discuss a new Device Driver Kit that was made available to provide the tools and resources for printing manufacturers to easily support all Linux distributions with one driver package.
• Linux developers discussed the increasing need for efficient power management in Linux. As a result of these meetings, LF is organizing a “Green Linux” initiative.
• A newly created Linux Standards Base (LSB) Test Framework and Testing tools were demonstrated

This sort of cooperation is important because the OSS movement creates software differently than the software you’re used to deploying in your enterprise. Unlike with MVS and its successors and cousins, the other forks of Unix besides Linux, and Windows, with Linux and OSS the core operating software developers and utility/driver/test developers do not all work for the same company. Yet the group in general wants to avoid the forking that doomed earlier cooperative efforts around Unix such as the Open Software Foundation.

I will be holding a “Talking with…” session with Jim Zemlin, executive director of The Linux Foundation, next month to learn more.

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

Last night’s Microsoft press release about open source software (OSS) is not just about an IP agreement with faceless technology bureaucrats in Cambridge (Novell) or Seoul (Samsung). . Theoretically involved in the June 13, 2007 Microsoft announcement is Cathedral and Bazaar and World Domination author Eric S. Raymond (ESR), self proclaimed as one of the three most influential people in the OSS movement [(along with Richard Stallman (RMS) of the Free Software Foundation (FSF) and gnu.org and Linus Torvalds (just plain Linus)].

Last September, ESR joined Freespire as sort of a committer, a member of what Freespire calls its Community Leadership Board. Freespire is the development feed into Linspire (which in turn is a Debian-based version of Linux), in the same way Fedora is the development feed into Red Hat Enterprise Linux, and Ubuntu is the development feed into Canonical. Sorry for all these cross references but that’s the nature of the OSS movement.

Raymond may have done the Samba mambo and left Freespire over this the same way Jeremy Allison left Novell after its similar IP deal with Microsoft in November 2006 but I think not. He is still listed on the Freespire web site. And ESR has preached “open choice” and interoperability with Windows since the early drafts of World Domination in 2004, particularly concerned with the need to access codecs and other code that just plain doesn’t exist legally in the OSS world.

ESR is also revered as the OSS leader that widely distributed and publicized the Microsoft “Halloween memos,” documenting--in my opinion--standard competitive intelligence about Linux for internal use by Microsoft developers and executives. In OSS circles, the Halloween memos were seen as a Microsoft’s declaration of war against Linux. Therefore, in OSS circles, yesterday may come to be known as “Wednesday the 13th.”

The press release doesn’t even hint at this possibility but if the deal includes Linspire desktop Linux vouchers the way the Novell deal includes SUSE server Linux vouchers, it would cement Microsoft’s goal of giving its users (that is, almost everyone) true “open choice.” I discussed the concept of open choice relative to OSS here. Helping users put Linux on servers as the Novell deal does is one thing but helping them put it on desktops with or in place of Windows is the first concrete recognition of my theory that Microsoft is endgaming its lead product and textbook cash cow.

From a business perspective, this is the beginning of Microsoft’s counterattack on the soon-to-be finalized FSF General Public License version 3 (GPLv3). GPLv3 is now in “final call for comments” and this agreement provides a great way for Microsoft to explain the differences between open choice with OSS and rabid OSS as preached by the FSF. The way this option works has some interesting business and investor implications that I discuss over at research 2.0 (see link on upper right of this page).

Posted by dennisb in OSS Culture | Permalink | Comments (1) | TrackBacks (0)

I wrote the other day about the release of Apache Synapse version 1 and how it set the stage for a variety of open source software (OSS) products that can take advantage of its asynchronous brokering structure and service oriented architecture (SOA) framework. Of course it was easy to say that when I already knew that a couple of the Synapse committers would followup this week with a Synapse-based OSS product. Their UK-based company WS02, with Intel venture funds behind it, rolled out its Enterprise Service Bus (ESB) Monday June 11.

WSO2 says its ESB “can route messages with sub-millisecond overhead and can scale to manage thousands of simultaneous connections” because it uses an integrated XML-based registry of WSO2 design and a simple AXIS-based graphical interface on top of Synapse and not a lot non-Web-related framework underneath. WSO2’s products are optimized to handle web services, a characteristic that Progress’ Sonic (another committer organization) might suggest is limiting, but one that keeps the code lightweight and ideal for working across the web as opposed to across Intranets. WSO2 says throttling, load balancing and failover controls optimize availability and help to manage and maintain service-level agreement, a key ingredient in my definition of SOA.

The ESB is WSO2’s second offering, following its Web Services Application Server (WSAS) launched in 2006. Both are 100-percent open source, based on key standards, and work with J2EE, .NET, JMS, and HTTP/S-based systems, as well as Apache Axis and Axis2 endpoints. The product is available “free,” under the Apache license of course. WSO2 makes its money offering a range of service and support options that include consulting, custom development, and sponsorship of open source feature development; development and production support; and training. Additionally, the WSO2 “Oxygen Tank” is an interesting OSS portal concept that provides in-depth product information, tutorials, tools, forums, wikis and more.

I am not sure if the community is going to be more accepting of such vendor sponsored sites as opposed to sourceforge and like portals but it will be an interesting OSS side issue to follow.

I am a firm believer that web services are not the be all/end all but if they are all you need, why build them on an SOA platform that will weigh down your whole development effort? These guys are worth taking a look at; I hope to interview them in the future for the "Talking With..." series because of their long-term knowledge of the WS and other OSS community issues.

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

In less than two years from first being proposed as an open source Web service mediation framework, Apache Software Foundation (ASF) Synapse emerged from incubation status in January 2007 and moved to the step all projects aim for, Release 1, on June 8, 2007. Synapse is a lightweight XML and Web Services management and integration broker that can form the basis of a Service Oriented Architecture (SOA) and/or an Enterprise Service Bus (ESB) and/or another web services infrastructure offering. It provides transformation and routing, and supports loose coupling between services. Its components work together with Axis2 1.2, an implementation of the SOAP recommendation from the World Wide Web Consortium (W3C), and other open source software (OSS) projects, particularly the Axis Object Model (Axiom). Synapse remains a part of the Apache Web Services project, as compared to ActiveMQ, which more or less began as part of Geronimo but became its own top-level ASF project when it emerged from incubation. The Synapse hierarchy makes sense technically and in terms of the heritage of key Synapse movers and shakers.

I recalled a little debate about what Synapse was when first announced (sort of an integration broker, sort of an ESB, sort of a management product) so I went back and to this article by Coleen Frye. Dave Chappell, vice president and chief technology evangelist for Progress’ Sonic Software, was quoted by Colleen: “We decided we were not going to call Synapse an ESB. We talked about this in great length, and we're all clear we don't want it to take the place of an ESB or other SOA infrastructure products."

You can also note my snide remark at the time as well but I am happy to say the group delivered on its promise. It is not another ESB (nor a MOM) nor an SOA framework but can form the basis of either as well as other types of systems management or similar infrastructure software. Project leaders from WSO2 (founded by key leaders of the Apache Web Services Project), Sonic, Atos Origin, and IBM kept the project on track. Infravio dropped out along the way, possibly the effect of being acquired by a company (WebMethods) that was later itself acquired (by SoftwareAG). Iona became less involved as its OSS strategy evolved in different directions (e.g., moving Celtix over to Apache, acquiring LogicBlaze).

The major changes in Synapse since the 0.91 release are:
* Upgraded to use Apache Axis2 1.2
* Addition of the non-blocking HTTP and HTTPS transport
* Improved samples and documentation
* Enhanced support for WSDL endpoints and load-balancing, throttling and failover

The Apache Synapse code and binaries are available from the website

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

IDC has recently released one of the first 2006 sizings I have seen of the open source software (OSS) market. IDC also provides a forecast (naturally, it’s IDC!) that puts the OSS opportunity into perspective, raising questions that IT executives (vendors and users alike) and VCs need to ask themselves.

In a press release, IDC says “the market for standalone OSS is in a significant growth stage... Adoption of OSS will accelerate over the forecast period of 2007 through 2011 as barriers to adoption get knocked down. Growth in revenue, however, will lag behind the growth in distribution of open source software.” That last sentence is a key reminder from Matt Lawton, program director of IDC OSS Business Models research, that the rules in the software business may have changed.

First, here’s the data:
• Worldwide revenue from standalone open source software reached $1.8 billion in 2006.
• This revenue will reach $5.8 billion in 2011, representing a compound annual growth rate (CAGR) of 26% from 2006 to 2011.

IDC defines standalone OSS revenue as license and maintenance subscription revenue related to any type of OSS regardless of the operating system platform. Most of Red Hat's revenue qualifies as “standalone OSS” revenue for example. So does Covalent Apache-related subscription maintenance even if Apache is running on top of Windows. But the numbers do not include the related proprietary software revenue (that is, Windows in the Covalent example) and they do not include revenue from the sale of complementary services (e.g., the kinds of professional services and training that were a key part of the JBoss operation before it was acquired by Red Hat).

By comparison, a similar IBM-, Microsoft-, or even Sun-based software ecosystem sizing and forecast would be measured in the tens of billions, albeit growing more slowly. Be careful however because increasingly the IBM and Sun numbers will overlap with the OSS measurement. Some analysts have even hypothesized that the Microsoft and OSS ecosystems will someday intersect.

Here are some of the questions that suppliers and investors need to think about: How does the infrastructure potential within these IDC totals stack up against OSS applications? Applications traditionally account for 60%-70% of software spend but just the opposite seems to be happening with OSS, according to other IDC data released in May and discussed here. Is there any opportunity in the infrastructure sector given Red Hat’s and Apache’s head start and the support of the leading IT suppliers for Apache projects? What applications should a developer or investor concentrate on if he or she thinks the historical application/infrastructure split will emerge? And most important, how much will revenue lag adoption (see IDC quote above) and can a supplier risk an investment waiting for “a slower dime” when the whole 40 years of software market history has been based on “fast nickels.”

As for IT executives in enterprises and government, they need to ask whether the metrics and processes that they have traditionally used to run their IT shops are really changing. Is a maintenance contract for something as reliable as the Apache web server even necessary? A lot of JBoss application server users apparently felt they didn’t need one given Red Hat’s experience. How does this new development philosophy on the part of most IT suppliers fit with the emerging budget philosophy of paying for software as a service (SaaS)? And most important, can the enterprise truly “afford” to join the OSS movement, giving back as much as it gets?

Of course, those are just the questions. If you want to discuss some possible answers, post a comment or send an email. The IDC study, Worldwide Open Source Software Business Models 2007-2011 Forecast: A Preliminary View (IDC #206681) is available on the IDC web site.

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

May 30th Google announced that it is providing developers with an open source software (OSS) technology called Google Gears for creating offline web applications. There was a lot of immediate discussion about the benefits of offline applications, which anyone like me who labored in the old days of downloading Compuserve inboxes (or the real old days of waiting three days for your report printout to emerge from the "Data Processing Room" ) finds quite funny. At least those that tout the benefits of being offline have dropped the buzzword "disconnected."

But what this announcement is really about is Google enlisting the OSS community to bail it out of a big hole. The company's need arises because Google's growth rate, though still impressive, is falling simply because it is getting so large. At the same time, financial markets have anticipated a Google growth rate that continues to rise into infinity and Google executives would like to oblige its investors. Therefore they need some other products/services other than their cash-cow search engine.

That's where you in the OSS world come in. Google wants applications that will compete with Intuit, Microsoft, Oracle and SAP, which will be no mean feat. Although all of these other developers use OSS (including Microsoft believe it or not), they do not develop applications in the OSS culture .(Oracle has some non-applications software OSS projects in process.). If the gambit works, it will be the true test of whether OSS development is really quicker and of higher quality than "closed-source software" development.

As for the Google Gears technology itself, it's a browser extension purportedly tackling "... a key limitation of the browser in order to make it a stronger platform for deploying all types of applications and enabling a better user experience in the cloud," according to Eric Schmidt, Chief Executive Officer of Google. As I have written over at Research 2.0 (see button to right), the portal is a possible enabler of the SOA generation of applications. Google Gears works with all major browsers on all major platforms: Windows, Mac and Linux. It is more likely that a wider platform including a message bus, rules engine, integration server and other features will win out over portal servers as the major SOA enabler in the long run but Google's need is short term (as are most IT-investment-related needs unfortunately). Google needs a more complete apps portfolio in place before its advertising-revenue-based growth declines to normal ranges in the next 12-18 months.

Google says it will be working closely with all members of the web community to converge upon a standard so developers have one consistent API for offline functionality. Those statements make it a little unclear whether Google's effort is really OSS or a good-old-fashioned market-creating attempt to build a standard around its apps.

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