Apr 26, 2010

memoQ projects now translatable and editable in an RTF editor!

It's been nearly four weeks since my last blog post here; since completing the requirements for my German hunting license at the end of March I have been very busy getting to know the 385 hectare area in which I'll be hunting (in part by doing a GPS survey of the area on foot and marking important places on maps using Google Earth and other orientation tools). Business and personal matters as well as preparations for the court battle in Zurich with the infamous non-payer of translation work Dominic de Neuville (formerly of the now insolvent Language Promotions GmbH, currently DBA "Transit") have been a real handful. And with all of that I had the preparations for the Spring Natural Abilities Test (Verbandsjugendprüfung) for two young hunting dogs, Ajax vom Bernsteinsee and Daniavizsla's Barack.

In the whole whirl of activity I had rather lost track of what was going on in Kilgray's world, though I knew I had to get my tail in gear and do some preparation for my talk on tool interoperability with memoQ at the upcoming memoQfest 2010 in Budapest. I remembered the promise of the company's director of development, Gábor Ugray, to complete a feature I've been pestering Kilgray about for nearly two years by the time of the conference, but I figured that it wouldn't happen since I hadn't heard about it for a while.

O ye of little faith. In my two years of contact with the company, Kilgray has seldom failed to meet a promised target, and when plans are changed, the users are informed well in advance and plausible, acceptable reasons for the changes are given along with the new plan. This is nearly unheard of in the translation industry. But once again, the boys in Budapest have been true to their word.

I returned from dog trials in Osnabrück near midnight yesterday, tired and sick from breathing the pollution produced by my chain-smoking friend on a 450 km return trip. (I thought about walking to spare my lungs, but the dogs wanted to get home faster.) This morning I downloaded my e-mail (a horror of hundreds of messages for a few days' absence) and discovered a fat gold nugget in all the electronic mud I had to pan: the public beta of memoQ version 4.2! It's available for download at For a "minor" update, it has some pretty major features, one of them being the long-awaited exportable and importable RTF tables, similar to the "external views" or bilingual exports from Déja Vu.

This feature can be used in a number of ways that are important to translators, translation agencies and translation consumers. For many years I have used this feature in Déjà Vu to export commented segments to indicate terminology uncertainties, source text errors, points to consider and more. I also export entire translations for review, comment and correction by persons who do not use CAT/TEnT tools, i.e. a typical translator consumer in industry. The modified versions can be re-imported with a few menu selections and mouse clicks, saving me a massive amount of time updating projects and translation memories. This is the feature I have missed the most over the past year as a heavy user of Kilgray's memoQ environment, and I have been hoping that something similarly useful would be added to the toolkit there.

Although there are apparently a few quirks in the first beta version with this feature (such as an indication of more changes than I actually make to the RTF file), it seems to work quite well, and it exceeds my initial expectations for the export in most respects. The comments column was too narrow for my taste in the test export I made, but perhaps that's because there were no comments. I cheated and re-sized the columns the way I wanted them in the screenshot at the end of this post.

The new feature is accessed in the Translations view of the Project home tab in memoQ. The Export bilingual option offers the following possibilities:

Selecting Two-column RTF causes the Next button to appear:

Clicking the Next button shows the following:

As far as I can tell at this point there isn't a "commented segments only" feature yet, but I'll do my best to get the development team drunk next week and talk them into it. But I'll bet they're already ahead of me on that one. The exported RTF file (with the cheat of resizing the empty comments column) looks like this:

There seems to be a little code page issue with the German "ä" in the title, but hey, it's beta and that doesn't affect the functionality. This RTF file can be edited by someone in MS Word, WordPad, Open Office swriter or other tools and re-imported to the memoQ project. By "editing" I also mean translation. The target column can be exported empty and filled in by a translator using only word processing software. In this way, for example, memoQ can be used as a tool to facilitate the translation or editing/review of all the file formats it works with by persons who use only a word processor!

The re-import into memoQ worked well in my test. In one case I chopped out most of the RTF file and left only a few lines of interest - the ID column ensured that the modified segments were re-imported accurately to the right place, even when I changed the order of the lines and introduced gaps in the text sequence.

Modified segments can be identified and checked quickly using the usual memoQ selection features for edited text:

What I would like to see though is a mark-up view or some sort of change history like one has for the re-import in Déjà Vu. But this is still a great start and removes the last obstacle to full implementation of memoQ in all my translation processes.


  1. Thanks, Kevin.

    BTW, do you happen to know if there is a way to prepare projects in memoQ and export them for translation in SDLX?

  2. @barbudo: Of course you ask me this question a few weeks after I de-installed SDLX for testing purposes :-) Nonetheless there are likely a number of options. Doesn't SDLX handle Trados bilingual RTF/DOc files? If it does, you can produce those exports from memoQ. Or you can use these new RTF table exports in version 4.2 and copy the source column into a separate document, translate it, then paste the results into the target column and re-import. (Or make things easier by populating the target column with the source and hiding all the text you don't want translated if SDLX has an "ignore hidden text" option. The guidelines I wrote on prepping RTF and Word files with content to exclude (available here as a PDF) might help you in this regard.

  3. Thanks for the great news, Kevin!

    I also asked Kilgray staff to add feature enabling the user to export the segments with comments.

  4. @Mykhailo: You can do exactly that with this feature. If you want only commented segments in your RTF export, then create a corresponding filtered view and export that as a bilingual RTF table. Denis Hay at Kilgray sent me that tip yesterday after reading this post.

  5. Ah, how I wish I were inside that TeNT pissing out, rather than the other way round at present!

    Until Kilgray sorts out the way it displays Japanese, all these beneficial new developments will be useless to me. And Kilgray have not given any indication of when they intend to fix the problem.

    Actually the problem with Japanese in MemoQ is like the expression TeNT - it's godawful to look at. So I shall keep pissing into it until the occupants get the message!

  6. Rod, while you are relieving yourself, you might consider that the particular feature I mention here - the importable/exportable RTF tables for translation and review - will be utterly unaffected by the character display issue you mention. An exported table can be translated using whatever tool you please, including a word processor.

    This, by the way, is a feature for which I have waited for a total of about two years now, because it is a critical part of my workflow with Déjà Vu (another tool about which I think you have anything nothing good to say). During those two years Kigray continued to develop other aspects of the software as a matter of priority. In fact, until last year's memoQfest 2009, I'm not really sure they understood what I wanted or whether it was important to anyone but me. Then an interesting thing happened: in an open discussion after one of the presentations, a half dozen members of the audience discovered that they had been asking for the same thing, just calling it by different names. Suddenly it became a priority (among other priorities for version 4), and now we have a better implementation of the feature than what I know and love from Déjà Vu.

    Was I frustrated with the wait? You bet. Was I tempted to unzip and let it rip on the only TEnT tool provider I know that seems to have a viable strategy for satisfying both the corporate and freelance sides of this business, that provides excellent, friendly support and that has staff who are allowed to act like real people? No way. There were simply too many positive things going on for me to reach for the long knife.

    I have some - not a lot - background in your working language, Japanese, though when I learned it I don't think there were any word processors usable by humans that could deal with it. Come to think of it, there weren't any word processors. That came later. In any case, although you have articulated this issue with font display a number of times on Watercooler and elsewhere and Wenjer has confirmed the same or something like it for Chinese, I haven't really understood what you're talking about. It's obvious that there is a problem, one which needs to be corrected before the product is viable for the Far East market, but a combination of foggyheadedness (lack of sleep on my part) and tone (on yours) makes it hard to grasp the point.

    I don't know what's on the plan for this in Budapest, nor do I lose sleep at night worrying about it, because it's not my market. I would, of course, look at things very differently in your position. However, as I recall you had a good, working system with MQ 3.6, didn't you? And the TMs and termbases are compatible at the desktop level. So patience with concurrent productivity is in fact an option with the software. Try waiting as long as I did for RTF tables before you play with your zipper.

    When I see that people like Angelika Zerfass (a German trainer who is competent in Japanese and/or Chinese) are heavily involved in Kilgray's effort, I have some confidence that these things will be worked out. But one thing I have learned to respect highly in the two years I have watched the company closely: this team has an excellent understanding of how to set priorities, and when a good case is made or presents itself, priorities are adjusted accordingly. They are simply brilliant in this regard, and my hat is off to them even if I often don't get what I want when I want it.

  7. Hi Kevin,

    I see you're in the TeNT even as we speak.

    The Kilgray people have no excuse not to understand the matter I raised since I sent them screenshots of the problem and a careful, atonal analysis. You can see the shots here;

    For people who aren't used to looking at Japanese, the way the text is rendered may look OK. But let me assure you, it isn't. By way of analogy, it's a bit like trying to work all day with a Fraktur font. It's a big step backwards from earlier versions.

    Since you're there, could I ask you to give one of the Kilgray people a poke from me about this? I've come to rely on memoQ and I was looking forward to some improvements. Well, they came, but hobbled with something that looks like Fraktur. I'm sure you can appreciate how frustrating that would be. If you could just convey that, I'd be delighted to readjust my position with regard to the TeNT.

    I'd be much obliged to ye.

  8. @Rod: The matter has been under discussion for two days already. It will be dealt with, but there the developer who rewrote the editing engine needs to spend some time to understand the issue more clearly. The people in the Kilgray camp who actually read the languages affected haven't seen the relevant information yet. I will forward it to them personally to make sure that everyone understands the matter. If Wenjer has good examples for Chinese I would like to see those too.

    Actually, as far as I could tell from our chat, the developer understand at a theoretical level what's wrong. He explained the technical challenge of the underlying technology in more detail than I could follow, but the upshot is that he's very aware of it and simply needs to get some specific feedback from the right people. It will get fixed, I'll bet my bicycle on it.

  9. Thanks Kevin!

    That's good news.

    I will of course be very happy to cooperate with any testing and feedback that might need to be done.

    (If you're betting something, I already have a decent bike, so why not put up your new shooting iron instead?)

    Enjoy the rest of your conference, and I look forward to your updates.


  10. @Rod: Not betting the drilling, but you can have my first-born son instead. If I lose the bet, I'll just have to find a volunteer to bear him first.

    But even if I were to bet the rifle I'm sure to keep it. Yesterday at lunch the developer responsible for the new memoQ editor sat on my left and a charming outside consultant to Kilgray who teaches students of Japanese and Chinese translation, uses those languages herself and is aware of the problem sat on my right. Before the main course arrived they were talking. So I believe that matters will get worked out. Maybe not tomorrow, but within an objectively reasonable time, and with her support, appropriate testing by a good number of persons with the relevant linguistic competence is assured.


Notice to spammers: your locations are being traced and fed to the recreational target list for my new line of chemical weapon drones :-)