John Freivalds, JFA, Inc.
“I think there is a good degree of recognition that Open Source gives a project or organization, an instant global reach given that developers and users with open source foreign communities are not limited to code and licensing provided by the vendor.”
Linux Today, December 16, 2008
These words are music to the ears of at least four groups that are trying to provide open source to the IT marketplace with the first functioning global translation management system (TMS), a “globalization linux” if you will. These groups are: FOLT, which was founded in 2005 by ten language service providers; Andrä Ag, which was founded in 1999 and in 2002 released its first translation management product; Tiny Tm, which in 2008 released a product to complement the existing translation workflow; and GlobalSight.
I contacted Bill Price about what any open source globalization software tool should be. He holds the impossible titles of Manager and CTO-Localization Services and CTO Documentation Delivery and Publication for the French telecommunications conglomerate Alcatel-Lucent.
“Open source software tools should be part of the next wave of advancement for the globalization and localization industry. While there are issues and limitations when it comes to integrating open source tools with existing tools suites, there are a lot of synergies as well. It does require an open-eye approach. Open source tools tend to be developed for an English-based world which limits language compatibility (increased use of UTF-8 will take care of this). Open source also requires programmers who understand the overall objectives of using open source software tools to develop more universal software solutions in much less time. The advantages of using open source software tools include speed to develop creative solutions for customers and reducing overall costs. End customers are more and more often demanding not only faster solutions, but also reduced royalty fees, reduced right to use fees, and reduced licensing cost. Open source software fits into this nicely and presents increasing possibility for reuse across multiple projects.”
The GlobalSight Case History
When GlobalSight hit the market 10 years ago, Upside magazine called it “one of the top ten private companies in the world.” Red Herring magazine was equally effusive by citing GlobalSight as one of “ten companies to watch.” And this praise came when GlobalSight was hardly a year old! During the decade it has been around, it has gone through US $50 million dollars of investment, three changes of ownership, many more of management, and several business models. So what happened? Corporate IT departments are interested in many factors when choosing software, and don’t rely only on what is written in trade journals. Competition happened in the form of other TMS software vendors—globalization service providers offered TMS as part of their services and potential buyers were wary of having to change their systems to accommodate the system—not to mention the license and service fees. And the real cynics said the venture capitalists who funded GlobalSight merely wanted to do an IPO and cash out. In 2005 John Yunker, an analyst with Byte Level Research, noted “I don’t believe that GlobalSight’s product is the reason for the software vendor’s floundering. The software has a future provided it is well priced, well managed, and “plugged in” to all other CMS tool (not a simple task).”
Industry consultant Don DePalma of Commonsense Advisory put the initial problems of GlobalSight and other TMS software firms this way, “There were efforts to create service deliverables … which essentially became free proof of concepts for larger companies…. In one sense, it was the wrong technology at the wrong time in the market. In many ways, TMS experienced the same kind of market rethinking that other internet spanned companies went through.”
Welocalize, which acquired GlobalSight in an earlier acquisition, fashions itself as an “integrated globalization service provider of enterprise content and applications to international markets,” has acquired a number of firms in recent months. No big deal here, since globalization service providers are always acquiring other firms; some corporate trivia here, the average life span of a corporation is just 12 years.
Welocalize has come to realize that the best way to bring the GlobalSight promise to reality and justify the US $250 million invested in the globalization marketplace is to make it open source. But the marketplace has many questions about GlobalSight, not the least of which are: does it make sense, is it doable, is it economical, and is it really free?
To get an idea of what the Open Source Initiative will look like, I spoke to many people related to Welocalize: members on its steering committee, potential users, and a commercial competitor. In an interview, Smith Yewell the CEO of Welocalize said, “Look. Two heads are better than one. Sure we can continue to improve GlobalSight by ourselves but we have come to the conclusion that there are enough people interested in its success that open source was the way to go.” By way of comparison to show what the open source business model can do, a 2001 study of Red Hat Linux 7.1 found that 30 million lines of code had been developed. If this were done by the conventional proprietary software model, it would have cost US $1.8 billion.
The first step was to form an Industry Steering Committee which was done in July 2008. Representatives from AOL, Autodesk, Cisco, NetAPP, Sun Microsystems, TIBCO, EMC, and IBM joined the committee. The first meeting was in August 2008, at which members considered strategy for open source licensing, governance, product development, and community building. The near term goals were to evaluate the current product feature set and industry landscape and select courses of action. You can follow the progress of all this at www.Globalsight.com.
There was need to re-architect a number of functions with open source equivalents including Workflow, Data base, Object Relationship Mapping, Middleware, and Directory Management but all this has been done. For example, it re-tooled the Oracle database layer for MySQL.This process is ongoing and Gary Prioste VP Technology of Welocalize told me that an open source roadmap will be posted on the web site “and available for download, along with accompanying product specifications and deployment requirement.” As this article goes to press, Beta Testing has begun.
To get the concept going, Welocalize presented the GlobalSight Open Source Initiative at such conferences as Localization World in Madison, Wisconsin and the recent LISA Forum in Dublin. In Madison, Ben Sargent of Common Sense Advisory reported that “two arguments underpin the current LSP (language service provider) around Open Source TMS. First and foremost is the argument that protects the user from predatory M&A by SDL or other power players in the market. Second, open source carries an attractive price: free.”
There is no doubt that Welocalize’s actions have changed the globalization software landscape, particularly from conflicts of interest arising from acquisitions. Across Software, for example, joined several translation automation providers and announced their independence from any LSP. To wit: If Across Software merges with or is acquired by an LSP within five years of the purchase, the company will refund alliance fees paid through 2009. Since much of the concern over continuity and conflict of interest has been directed at SDL, I thought I would ask my friend Terry Lawlor of SDL his thoughts about being cast as the “devil incarnate.” He told me to wait and “expect some interesting perspectives direct from Lucifer!” So Terry arranged a telephone call between me and Sophie Hurst, Director of Product Marketing at SDL, which in recent years acquired two major globalization software firms, Trados and Idiom. Both are included in an overall service package called “SDL Global Information Management which also includes its own SDL Translation Management System™ (SDL TMS™).” Both were fully aware of the Open Source Initiative but Sophie Hurst brought up these points:
“Implementing an open source solution may limit or remove the license cost, but there are costs which must be taken into consideration when looking at such a solution…You need to cover the costs of development to bring features in line with your requirements, and there are implementation and consultancy costs that must be factored in. Typically in an enterprise software sale, the license cost is a small percentage of the overall system implementation cost. What is most important is that the solution is based on your business requirements, not whether it is open source or not.”
So what now? Although Welocalize hopes to earn some money by offering consulting services showing customers how best to use GlobalSight, its future success really comes down to passion. Can the GlobalSight Initiative sustain the passion of its managers and the steering committee? At the start of 2009, almost 400 members have signed on, and, on January 12th, Welocalize announced 132 application downloads in the first week and 828 downloads of all materials.
Don DePalma of Common Sense Advisory likes the early results but points out “users of open source and off the shelf software both need to explore bigger issues such as the five year road map for the product. Since enterprise applications run for five or more years, you need to know that the software will be there. … The bottom-line on all of this is whether the issue is cost or comfort level, you still have to choose which issues make the biggest difference to you.”