Tool Box Newsletter Logo

 A computer newsletter for translation professionals


Issue 13-10-228
(the two hundred twenty-eighth edition)  
Contents
1.What happens in Böblingen . . . (Premium Content)
2. TDAta
3. XTRF (Premium Content)
4. Windows 8.1
5. On Good Terms with Terminologists
6. New Password for the Tool Box Newsletter Archive
The Last Word on the Tool Box
Numerology

No one knows better than us that words are far from easy to pin down. But numbers are a lot easier, right?

I was amused by numbers this week when I looked at recent releases of translation-related software programs.

  • You would have had to be in another world if you missed the release of Trados Studio 2014 a couple of weeks ago. Fair enough: it's close to the end of the year, so it's no over-exaggeration to name it for the next year. Plus, looking at the many (positive) changes in the new version, it's certainly justified to make this a major new release, complete with a price for the upgrade. You can find a good review right here, and you can also look at what I wrote in the August edition of this newsletter.
  • Star released its last "full version" (of Star Transit NXT) in 2008. Less than a month ago it released its free Service Pack 7. Now, the service packs that are typically released for other products contain a bunch of bug fixes and some minor new features. Looking at this "service pack" (and most of the six preceding ones), though, this is far from a mere bug fix, and the new features are anything but minor. Surely this could have warranted a full release.
  • Also, just this week, Kilgray released memoQ 2013 Release 2. To the outside onlooker this creative numbering scheme might sound like a "redone" version of a previously botched version, but it's far from it. The original 2013 release was neither botched nor are the new features minor. This, too, could have warranted a full new release. (You'll get a good taste of that release right here.)

You tell me...

Of course, this whole numbers-and-version game becomes increasingly insignificant when some software development companies -- such as Kilgray and Star -- require maintenance contracts to access their new versions, but it still makes me chuckle.

Chuckling was not likely to be an activity indulged in by the folks at the Vatican when they discovered that a commemorative coin they issued this year bore a misspelling of the name of Jesus.

Lesus coin

But on the other hand, it was no worse than this Chilean coin issued in 2008. (Don't see it? Iook closely!)

Chiie coin  

At least only two specimens of the Vatican coin were actually purchased before the error was discovered, but thousands of the Chilean one are in circulation. (Oh, and if you ever want to make very good money by using your skills as a proofreader, work for a mint, misprint a coin, have your spouse buy it on the first day of issue, and wait a few years before you sell it.)

Finally, so I never have to bother you again with coins in future newsletters, some of you might remember that I recently wrote about the rejected Slovakian two-euro coin depicting two translators (Cyril and Methodius). Well, it was published after all. And here is my well-manicured hand proudly displaying it:

Translators on coin

1. What happens in B�blingen . . . (Premium Content)

True story: Ever since I was little I asked myself why anyone would call a town "B�blingen." I don't know how that name sounds to you, but even as a six-year-old the word sounded weird to me. My deep thoughts about said town were inspired because my aunt and uncle lived there for a while, but what I could not have known then was that this town would also be the birthplace of not one, not two, but three cornerstones of the technology that most of us use today.

B�blingen is home to the biggest IBM research lab outside the US, and it was there that an internal IBM team developed the translation environment tool IBM TM/2, and two translation service providers for IBM -- Trados and Star -- developed MultiTerm/Workbench and Star Transit/TermStar. This all happened between 1992 and 1994.

IBM stopped marketing its product in 2002 and semi-revived it in 2010 as the now again defunct open-source product OpenTM2. Of course, we all know what Trados did, but what happened to Star Transit? I'd bet anything that many of you have not even heard of the product, let alone seen or used it, especially if you are in the US, South America, or Japan. And yet it's alive and, as far as I can tell, well.

The "as-far-as-I-can-tell" disclaimer has its reasons. Unlike Trados, the makers of Star Transit have always maintained their emphasis on the translation service part of their business and trusted that their product will essentially sell itself. But now they have come to the "realization that it takes more than the best product to be a market leader." I'm quoting Star AG's CEO Josef Zibung here, with whom I had a long chat about the product and its philosophy.

Of course, there are many more TEnTs than Trados and Transit on the market today, but these two products inhabit two opposite ends of the spectrum representing what it means to develop and sell technology. Both have emerged with strong and mature products, but they reached that goal in very different manners and certainly with widely divergent market shares.

On a functional or feature level, Transit and Trados had two main differences: First, from the get-go, Transit has used one independent translation interface for all formats (pre-Studio Trados, of course, used four or five different interfaces for the many difference formats). And second, while Trados (and virtually all other tools) uses a classic translation memory, Transit has always used a system of reference files so that the translator is able to see the context of each segment; this means that, aside from the many translated file pairs, you also have an index of sorts that connects the files and the segments therein. I know that both Trados and Transit would protest this "oversimplification" of things, but this is what it really comes down to.

Transit does not release many "versions" -- Transit 2.7, the much-loved and very stable version, was introduced in the late nineties, followed by an ill-fated and faulty successor (Transit 3) that was quickly replaced with Transit XV in 2001 and with Transit NXT in 2008. These are long stretches without new payable versions for a development company, especially because the development never stopped (though for the higher-priced version you will have to pay for maintenance contracts). Shortly after NXT's release I wrote a review, but here is a list of the major features that have been added in the form of service packs since then:

  • Ongoing support for all new versions of InDesign, FrameMaker, MS Office, OpenOffice/LibreOffice, QuarkXPress, Quicksilver, and AutoCAD
  • Added support for all new versions of Windows and Windows Server
  • Added database systems for TermStar, the terminology component (MySQL, MS SQL Server)
  • Batch processing of QuarkXPress files
  • Language pivoting for the terminology and reference material
  • Preview for Office documents, synchronized PDF preview
  • Integration of OpenOffice and MS Office spell-checking
  • Support for XLIFF as an export and translation format
  • Morphological support for 15 European languages (incl. English, French, German, Italian, Spanish, Czech, Dutch, Polish, Portuguese, Russian, and Swedish)
  • Added MT interfaces (both for Google Translate/iTranslate4.eu as well as customer-furnished MT systems)
  • Support for subtitling files with synchronized playback during translation
  • Introduction of a parallel translation memory system with the TM-Container

And then, of course, improved usability, stability, speed, etc. No matter how skeptical you might be, that's a lot of improvements through service packs.

The most fundamental change in the list above is probably the introduction of a true TM system that now runs parallel to the reference file-based system. The reason behind the introduction of that system is easier transfer and remote access to resources. The emphasis of the Star representatives who continued to send me information on this was that it differentiates itself from "other TMs" by maintaining all the document structure information, so any document can be regenerated just on the basis of the content in the "TM-Container," and the user has all context-based information that she usually has access to in Transit.

I have actually used Transit a lot -- in its older XV version -- and really like it. For this article I had a chance to look at the NXT version again. It certainly attempts to convey a much more modern feel -- with a ribbon and all -- and it mostly succeeds (I say "mostly" because it still has a "workshop feel" to it, but that does not have to be a bad thing). It's by no means a system that you can sit down with and immediately know what to do with -- there are incredibly many options to use and fine-tune the system. But if you take the time to learn it properly, I have a hard time imagining anyone not liking it. While some processes are slow (like the initial processing of large termbases or reference materials), the system is very quick and responsive when it comes to the regular translation work, the termbase is wonderfully integrated into the workflow, and the automatic matching on the target language level when no source match is found (called "dual fuzzy") is truly unique and very powerful.

I hope Zibung's and Star's "awakening" in regard to marketing and publicity will continue to have an impact and that this tool will reclaim a viable spot in the market, not just in Europe but elsewhere, too.

ADVERTISEMENT

The Swiss Post-Editing Score™

Automatically monitor your translation workforce, in real time. 

Contact [email protected] to take part in the free beta test starting this month.

Modulo Language Automation LLC, Switzerland

2. TDAta

Don't think that I like to give presentations only because I like to talk. I like to listen to the groups I'm presenting to just as much -- no joke!

One thing I recently realized is that there is one terminology resource that is heavily underused by translators, the repository of the TAUS Data Association.

Here is what I wrote about them in the past:

The TDA is an association of mostly large corporate translation buyers who originally came together to pool their translation memory data to better train their machine translation engines. To make statistical machine translation work, you need a lot of high-quality data, and even industry giants like Oracle, Microsoft, and Adobe did not have enough data on their own to get the results they hoped to achieve.

So, rather than semi-secretively pooling their data to train their MT engines, they decided to open the data up to the public -- not as translation memory data, mind you, but as a terminology resource. (If you want to get to the data as TM data, you can become a TDA member, contribute your own data, and download some other data for your own use.)

The terminology search site is open for everyone (you only have to register -- and sell the birthright to your first-born -- just kidding) and it really is very helpful -- provided that you work in the subject areas that are covered by the data donors, which are still mainly to be found in the IT industry as well as the public sector (EU and UN data).

The cool thing is that you can prefilter the data according to language pair (duh), subject matter, and client. Once you enter a term and the filtering criteria, you will get all hits that contain your term with some metadata (origin, etc.), information about the original translation direction, as well as a field that allows you to enter problems. At the very top you will also get a list of "computed translations" based on probability data derived from the underlying corpus.

Since I wrote this, the size of the database has grown significantly. The total number of words in the repository is presently more than 55 billion (yes, that's a "b"), and the total number of language pairs is 2200. Additionally, you can also use the so-called "Matrix TM" data, which essentially increases the number of matches by using a pivot language (if there is a match of a segment between language 1 and 2 as well as language 2 and language 3, then that might also be a match between language 1 and language 3).

It really is a very high-quality resource (though I noticed that some of the data contributed by some of the freelancers is of very poor quality...) that you would be silly not to use if it matches your areas of translation. You can access the data right on the website, you can use a free little Java-based widget (that word again!), or you can even use tools like IntelliWebSearch for it (see my description of how to do that in the 182nd edition of this newsletter in the archives for Premium subscribers).

3. XTRF (Premium Content)

In all my talks with the folks from XTRF, I never once asked what their unwieldy product name actually is an acronym for. How about if we make one up? Say, "EXTRemely Faceted"? It may not sound too clever, but it does describe the system -- lots of facets, including a lot of features to work through and understand.

There are a number of industry-specific workflow and transaction systems for LSPs out there, but there are only two that every LSP will surely stumble upon when looking at what's available: the German system Plunet and the Polish system XTRF. Both had their origin as internal systems for a translation company, and in both cases they were split off from the originating company to focus solely on the development and marketing of the software product.

Andrzej Nedoma, the son of Lido-Lang's founder, was the driving force behind XTRF, and his background in management is clearly reflected in the design of XTRF.

LSPs can use XTRF in two different flavors -- well, since XTRF is very customizable, there can be an almost unlimited number of flavors, but there are two different usage models. You can use it as a cloud-based SaaS application, hosted by XTRF (through the German hosting service Hetzner), or you can install it in your own premises. Typically, operations director Tomasz Mr�z told me, customers with up to five licenses (licenses usually are needed for project managers, accountants, and sales personnel) tend to use the cloud offering, whereas larger LSPs tend to choose the on-site installation -- with the trend going toward the cloud offering.

Pricing mostly depends on the number of licenses (about 80 euros per license per month.)

Presently the system consists of two distinct but interconnected parts: the client portal and the project management interface. I love the client portal. It consists of a widget-based interface. I know that many might still be suffering from the ill-attempted attempt at "-dgets" with the poorly implemented Windows Sidebar in Windows Vista and 7 and its "gadgets," but XTRF has a widget-based client interface view that is beautiful and very customizable. You choose what you want your client to see (progress of the project, YTD spending, fee overview, etc.), all of which will be presented as individual overview panes ("widgets") on the screen that allow for further access to the respective feature and more detailed view by the client. The idea and the hope is to always bring the client to that portal since this greatly streamlines project management for the translation provider and results in lower costs. One way to achieve this is to make it easy to use and attractive, and the other way is to promise cost savings to the client by automating the process of submitting new projects (at least for smallish kinds of projects) and thus foregoing project management efforts and potentially associated fees. XTRF makes both possible.

Tomasz was pretty honest in saying that most XTRF customers are still far from the goal of having all or even most of their clients using XTRF's client portal, but he feels they're on the right path.

The other big part of the system is where project managers interact with the system. Like most other workflow systems, this includes guiding your project from the (optional) quote through to project creation, which includes file and vendor management, vendor communication, and project tracking all the way to the invoicing process. XTRF presents a powerful system that is quite customizable and seems to leave very little room for improvement, were it not for the user interface that is not particularly transparent. I especially missed (optional) wizards that would lead a project manager to the desired result.

Case in point: the powerful report section hosted by Google Libraries. Aside from a number of pre-configured reports about productivity, there is essentially no limit as to what you can query out of your data. It took Tomasz less than a minute to show me how to configure a meaningful report: "You just have to know how to do it," he said ... and, boy, I wouldn't have.

Another of those hurdles is the vendor portal, the access point to the system for freelance translators. Rather than making this the default way of downloading and delivering projects translators are encouraged to using an automated and yet old-fashioned FTP system to download and upload their data -- the folks at XTRF assumed that it's too cumbersome for translators to use the existing vendor portal.

Here is the good news regarding all of this, though: XTRF's development focus during the next months will be to iron out these issues and make the user interface easier. In fact, some of those easier-to-use elements are already emerging. I mentioned the client portal as an example, but I really liked the graphical workflow editor as well.

The way that XTRF helps companies to implement the system in a way reflects the complexity: "The idea of implementation is that you do it bit by bit," said Tomasz. So rather than throw out all existing systems at once and replace it with the new system, you want to start with the less complicated projects so that you have a smooth transition.

To find out more, I talked with representatives of Venga, a relatively young Californian translation and localization company (they've been around for approximately five years). They started to use XTRF about two years ago and did not really have a prior system in place that they needed to replace. I asked them what convinced them to go with XTRF rather than a system like Plunet, for instance. They said they chose what seemed at the time to be the greater flexibility and customizability of XTRF. Venga heavily utilizes XTRF's deep integration with their preferred TEnT, memoQ, so they essentially command most memoQ processes from the XTRF interface (exceptions to that include processes like DTP, QA, or -- of course -- troubleshooting). The same is true for the integration of their accounting system QuickBooks with XTRF; in fact, here they can use either QuickBooks or XTRF to access the other system.

When I asked about the training necessary for the average project manager to familiarize herself with XTRF, they estimated one to two weeks for the basic tasks and an additional one to two weeks for the more complex tasks.

The aforementioned integration with translation environment systems is a big deal for the usability of a translation management system. Just a couple of years ago, management systems prided themselves on having an "integration" when they could process the analysis results of a TEnT -- well, hello, Shallow Hal! -- but these times are clearly gone. The integrations that XTRF offers are much deeper and driven on an API level, meaning that the different software systems can actually communicate with each other. The existing integration with TEnTs includes memoQ, XTM, MemSource, and to some degree Trados Studio (for instance, it's possible here to create packages and send those to translators through XTRF's interface).

When I asked the Venga representative about the level of customizability of XTRF, she said -- very wisely and diplomatically -- well, there are some things you can do and other things you can't.

Isn't that always the right answer, though?

 

At the risk of inflicting some whiplash on you, let's go from the really big system for LSPs and corporate translation departments to a rather bare-bones system for freelance translators and very small agencies. Kobalt Languages in Spain recently released the cloud-based system 4Visions (hosted by Amazon's AWS service).

I like 4Visions' concept and potential, but I can also see that there a lot of things that need to improve before it can really have an impact.

The idea is that the system takes care of most business aspects of a freelancer's life, including tracking projects and creating and tracking invoices. 4Vision does all that, but it's a little too disconnected to make it truly helpful. For instance, when you create a quote or a project, the data is not automatically transferred to the invoice section of the tool, requiring you to enter all the data again when creating the invoice. That seems unnecessary and needs to be changed.

The good news is that Ricard Sierra, the brains behind 4Vision, is aware of having to fix some shortcomings (here's a good opportunity for you to help mold a tool to your specifications -- I can promise that he will listen). Some of the things that he is planning to implement include a report feature (a crucial feature in a tool like this) and an automated integration with Google Calendar so that alerts about missed payments or deadlines are sent to you that way. It would be great if the handsome dashboard with a great view of your activities that you now see when opening 4Visions's homepage were also integrated into Google Calendar, and via that route could then be integrated into tools like Outlook. (And you're worried about using Google Calendar because of data security? The NSA does not share that concern!)

4Vision's current price for a professionally working freelance translator is 29 euro per month (which actually was raised from 19 euro since the last time I talked to Ricard.) A little steep if you ask me -- but, again, I'm sure that Ricard would listen to concerns about that, too.

ADVERTISEMENT

Kilgray has released memoQ 2013 R2 - come and explore!

Based on your feedback, Kilgray has added a wealth of productivity boosters to memoQ that will make your work easier and more effective.

Monolingual review, PDF import with full formatting, enhanced compatibility with other translation tools, online projects with packages, roles in translation memories and memoQ TM search tool are just a few of the many features that memoQ 2013 R2 brings.

Let the features do the talking: click here, download, install and enjoy using memoQ 2013 R2!

If you are not familiar with memoQ, you can download a full-featured 45-days trial

4. Windows 8.1

After all the hoopla about Windows 8, Microsoft quickly released a free update: Windows 8.1. In previous years this would have been a full new (and paid) version, but in an era of poor reviews it's just a .1 version. (You can find my review of Windows 8 and how it relates to translators in the archives in the 216th issue of the newsletter.)

I did have a quick look at it on my secondary computer -- which actually did not turn out to be such a quick look since I ran into some upgrade hiccups -- and . . . I was kind of disappointed. I did not think that Windows 8 was exactly terrible to start with, but I did not like the silly tile view on the Start screen and the absence of the Windows (Start) button on the desktop view. At first glance the Windows button has actually been restored in this new version -- but that really is only the first glance: Clicking it just opens the -- yes, you guessed it -- tile view with the many unnecessary apps. So I once again installed the free third-party application, Classic Shell, which gives me a pretty close version of the old Start menu.

One thing that I liked was that you can actually start your computer on your desktop, but you have to tell it to do so: On the desktop, right-click on the taskbar, select Properties, and open the Navigation tab. Check When I sign in or close all apps on a screen, go to the desktop instead of Start.

Bingo.

Oh, there is one really cool feature that Windows 8.1 introduced which simultaneously was also introduced for Windows 7 through the normal Windows Update channel (KB2852386):the ability to delete copies of old Windows Update files from your hard disk. Doesn't sound too exciting, but I was able to reclaim 4GB of space on my older Windows 7 computer with this method.

OK, "exciting" is probably not the right word for it, but it's helpful nevertheless. Here is the description for how to do it in Windows 7, and the process is just about the same in Windows 8.1.

5. On Good Terms with Terminologists

I recently had a talk with Barbara Inge Karsch, a well-known terminologist whom I have been cooperating with on some projects. I put myself into a strange position with that talk because she clearly knew what she was talking about -- terminology -- and I really didn't, at least not on a theoretical level. Still, I was really happy with the result and learned a lot in the process. It was published in the ATA Chronicle and it's available online on the ATA's website.

And speaking of the ATA -- I can't wait to see many of you later this week in San Antonio! 

6. New Password for the Tool Kit Archive

As a subscriber to the Premium version of this newsletter you have access to an archive of Premium newsletters going back to 2007.

You can access the archive right here. This month the user name is toolbox and the password is sanantonio.

New user names and passwords will be announced in future newsletters.

The Last Word on the Tool Box Newsletter

If you would like to promote this newsletter by placing a link on your website, I will in turn mention your website in a future edition of the Tool Box newsletter. Just paste the code you find here into the HTML code of your webpage, and the little icon that is displayed on that page with a link to my website will be displayed.

Here is a webpage that proudly display the Tool Box logo:

www.guiadotradutor.com (By the author of Guia do tradutor: melhores pr�ticas)

If you are subscribed to this newsletter with more than one email address, it would be great if you could unsubscribe redundant addresses through the links Constant Contact offers below.

Should you be  interested in reprinting one of the articles in this newsletter for promotional purposes, please contact me for information about pricing.

� 2013 International Writers' Group