In a recent issue of the Globalization Insider, “Isaac Bickerstaff” paid LISA and standards (and TMX in particular), some rather back-handed praise. He claims that the present state of the industry and standardization allows him to manipulate his customers into further dependence on his company, while he claims to support openness and standardization.

So what’s wrong with this picture? One must wonder what to make of Bickerstaff: are vendors really this dishonest, and if they are, what are we to make of a dishonest vendor who is honest about his dishonesty? And while Bickerstaff is obviously satiric in his intent, how serious is he about this strategy?

Where are we in reality? Is the industry what Bickerstaff describes it as, and if it is, what are we to do? I think this all comes back to the question of TMX, the one effort in this industry that has resulted in a real standard (complete with implementations and certification). Critics of TMX like to point out the problems that remain with the standard (problems which no one has ever tried to conceal, by the way). The standard is not yet perfect, but it is a usable standard, and it has been implemented. Now there is even certification of a TMX product (SDLX Translation Suite 2003), and other products are lined up for certification in the near future. So, even if there are still problems, TMX has made considerable progress since work on the standard began in 1997.

According to Winston Bumpus, Director of Open Technologies and Standards for Novell, standards have a six-stage life-cycle, consisting of the following:

1. Draft
2. Reference implementation
3. Approved specification
4. Implementation phase (standard is implemented in real products)
5. Revision phase (after implementation experience)
6. Certification phase

TMX has reached the last stage in this cycle - a point most standards never make it to. So TMX has really come a long way. Six years of hard and often tedious work has paid off.

I think it is worthwhile to compare TMX to another standard that has tremendous implications for the GILT industry - Unicode. Unicode has been around over twice as long as TMX, and is aimed at solving a problem that has been around for a very long time. Unicode has the potential to impact tremendous numbers of computer users around the world on a daily basis. Unicode is much more important than TMX in the grand scheme of things, but how much impact has Unicode really had on the average GILT user to this point? To be blunt, not a whole lot - by and large it hasn’t changed how we work or what we do, and its impact on the tools we use to actually produce the files we deliver to our clients has been essentially non-existent in most cases. (If you don’t believe this, just try using a Unicode-rich font in your latest Quark XPress file…) In spite of all of this, no one is calling Unicode a failure. We realize that translating a standard like Unicode into practical applications and ironing out the bugs takes time and effort, and we are willing to wait and work for the results. It’s taken a while, but the sorry state of Unicode implementation in DTP applications is finally changing - see my article on Unicode implementations in end-user applications in this issue for more information.

So why are so many of us ready to bag TMX and throw away what OSCAR has accomplished with far less consideration and patience than we have given Unicode? There is an old story about a bucket full of crabs on a dock within easy reach of freedom: the crabs never escape because they can’t stop trying to keep the other crabs from escaping the bucket. Each one is more interested in keeping the other crabs from gaining freedom than in getting free himself. In the end all the crabs get hauled off and cooked, even though they all could easily have escaped if their attitude had been different. TMX represents a way out of at least one of the buckets we find ourselves in (tool and vendor dependence), yet failure to take advantage of TMX ensures that we will all stay in the bucket. If all of us really wanted out and were willing to cooperate, we would get out.

Another problem, and perhaps the most critical, is that too many consumers of GILT services claim to want or need TMX, but are unwilling to do anything to actually make sure TMX meets their needs. If GILT consumers really want TMX to succeed they need to get involved with OSCAR and demand that their tools providers support TMX. Yet so far there hasn’t been a rush to join OSCAR or walk away from vendors who don’t really support TMX. Those few clients (like SAP and JD Edwards) who have participated in OSCAR have provided some of the best input that OSCAR has received. Sadly, for far too many GILT consumers, TMX is simply a check box on a vendor questionnaire - “Can you provide TMX? Yes? OK,” and that’s the end of it - TMX is never mentioned again, files are never actually requested, and the consumer has no idea what to do with them in any event. We need to hold vendors and tools developers to a higher standard than this - and clients have the power to do so. Ask that your vendors use TMX-certified products, and ask for TMX files. When their clients ask for TMX the vendors will provide it.

At the recent LISA Forum Europe in London Jonathan Bowring of Canon was asked how important TMX was to him. He answered that it was very important. If TMX is important to you, please make sure that you let your tool and service vendors know this. The vendors in OSCAR have really done a wonderful job of pushing TMX along, but clients need to make sure that they demand TMX-certified products and that they are educated about TMX. I realize that GILT consumers have other needs and are trying to stay afloat themselves, but TMX is a life preserver for consumers - if they can’t be bothered to grab onto it, who can they blame when they sink?

If this sounds like a rant, perhaps it is. OSCAR has put a real certifiable standard on the table, and all many of us can do is sit around and complain that it isn’t perfect, that we don’t like this or that about it, and that we want something else. Fine, ignore TMX and go back to being stuck with your same old vendor and same old tools, with no way to change. Like it or not, TMX is the only real alternative to the way we’ve done business for the last decade.

On a positive note, let’s consider Unicode again. It may have taken a while, but Unicode has been implemented in all sorts of applications now, and the number is growing. Chances are that if you aren’t already using Unicode for some of your work, you will be using it very soon, and it will impact you. TMX will progress the same way since standards implementation takes time: you aren’t using it today, but you will be within the next few years, and its impact will grow as the actual implementations improve and become more ubiquitous.

And now for something completely different…

Two issues ago we carried an article on the progress of LISA’s Terminology Special Interest Group. In that article we announced the forthcoming release of the SIG’s report, Terminology Management - A Comparative Study of Costs, Data Categories, Tools, and Organizational Structure. This report is now available.

In this issue we are happy to present an interview with Suzanne Hamlin of Cisco Systems (premium content) entitled Don’t Try to Boil the Ocean in the First Pass. Hamlin provides a candid view of what it’s like to guide one of the fastest moving high-tech companies in the world as it’s trying to learn how to go global. She also shares some very good advice for small- and medium-sized companies that are attempting to set up globalization in the right way from the beginning.

For those of you who missed out on the recent LISA Forum Europe in London, we have some highlights of the Forum.

Finally we present a review of FontLab 4.5 (premium content), a tool that many of our readers may be interested in for production of international fonts. FontLab allows users to build Unicode-rich OpenType fonts that can be used in the increasing number of Unicode-aware applications on the market, and represents a solid (if somewhat quirky) contribution to the bag of tools available to GILT professionals.