Jan 10, 2010

The experience of a new MemoQ user

I have been testing MemoQ for about two years and working actively with it since the breakthrough version 3.5 release. Shortly after I purchased my license, Kilgray offered a great group buy opportunity through ProZ (for far less than I had paid), so we bought a second license for my partner, who had developed a certain fascination with MemoQ or at least the effect that working with it had on me. For weeks she heard such strange squeals of delight from my adjacent office that she began to suspect that my mistress was hiding under the desk. Much to her disappointment, I was simply enjoying the clean design and innovative features of the software.

Although I installed the software on her laptop, it wasn't until many months (and quite a few updates) later that she got around to using it. There's simply too much going on in this office on many days with a heavy work schedule and three hunting dogs who demand their share of attention. Then about two weeks ago, we had a high volume rush project that pushed our capacity to a scary limit and demanded great efficiency in collaboration. Since we don't currently have a good TM server running, I decided to set up the master project on my machine and export bilingual exchange files for parts of the files we worked on. In addition to that, regular exports of TMs and terminologies were exchanged. We also reviewed each other's work by exchanging views in "MemoQ bilingual" format (as this also permitted the exchange of written comments). In the course of several very hard days of work I answered a number of beginner's questions and was occasionally greeting with a "Wow!" or a "This is soooooo much better than...", so I thought that perhaps a little "in-house interview" on Monique's new software experience might be in order.

Translation Tribulations: Ms. Simmer, could you give us some background on your previous experience with translation environment tools?

Monique Simmer: You mean CAT tools? I started using Trados in the 1990s and continue to process files with it [editor's note: Trados Classic] today. I've been using Déjà Vu since 2000. I tried SDLX at some point but didn't like it. And now MemoQ.

TT: What did you like about Trados [Classic]?

MS: Nothing.

TT: What's not to like?

MS: I'm not a sheep and don't fall for sleazy marketing tactics. It doesn't handle numbers and dates. It's full of bugs. Often the target files won't save from TagEditor or there are cleaning problems. Do you really want the whole list? Its creators ought to be tarred and feathered!

TT: How would you compare Déjà Vu with MemoQ?

MS: I don't have enough experience with MemoQ to make a comprehensive comparison. The filters seem to be better. I find that the labeling of menus on the MemoQ screens is not very intuitive sometimes. I got confused between importing and exporting the bilingual exchange files and "ordinary" files. 
On the positive side, it's an easy program to use overall. It's übersichtlich as we say in German. I like the fact that the keyboard shortcuts could be customized so the ergonomics are basically the same as when I use Déjà Vu. And the QA window was very, very useful for checking my work at the end. It caught a few places where I had mistyped numbers.
I also like the fact that I can use more than the two databases, which is the limit for my Déjà Vu Pro license. We've been talking about spending over 1000 euros to upgrade my DV license so I can do this, and now I can do it - and generate uncleaned Trados Word files for some of our customers - without spending all that money on DVX Workgroup.
Oh, yes... one thing that drives me up the wall about the MemoQ editor is that I can't reposition words by dragging and dropping, and the leftmost character is often dropped in selections.

TT: I'm told that version 4 of MemoQ, scheduled to be released next month, will feature a new editor without those problems.

MS: Let's hope so! So far I understand why you like the program. And the kind of support that Kilgray provides is really important.


  1. Though I am a loyal MemoQ user, I must say that I like the concordance and align features in Trados (classic) better. First, the concordance in Trados does not produce any of those annoying "expression x occurs in too many segments" results as MQ does. Second, it gives the user more search freedom, ignoring the endings (critical for Slavic languages). Third, I can set it to "show most recent translation units first", which is often very useful. Fourth, the concordance window is not fully resizable - so I have to close it so it doesn't clatter the view.
    Now for align: it is simply easier to understand and produces better results than the align module in MQ.
    But these are my only issues with MQ.

  2. @Arkadi: I haven't used the MQ align feature yet so I can't comment on it, though in all my time as a DVX user, I did tend to prefer doing aligment in Trados WinAlign. My main issue with Trados Classic is its inefficiency as a translation environment; some of the other features were not bad, really. Hard to beat the speed of import/export from the TM with other tools in that price class or below.

    As for the concordance, it's two-dimensional. I can resize it on the x-axis and the y-axis. What is your definition of fully resizable? I can't change the column sizes in the window. Is that it? MQ can indeed ignore endings by using wildcards; I do this all the time for German adjectives or cases where I want to avoid variations for plurals, etc. So I don't understand the advantage you claim for Trados in that respect. As for cluttering the view, I just pull the concordance over onto my second wide-screen monitor and leave it. Don't know what I'd do without that extra screen space.

  3. Kevin
    Thank you for your comment.
    I think I should have made myself clearer. Yes, you can use wildcards but that means the extra moves that are avoided if you use Trados (sorry to sound so stubborn on this one). Then, if I key in unquoted phrases, the MQ concordance often comes up with that error message I mentioned ("expression x occurs in too many segments") and doesn't show anything - that's annoying. Fuzzy searching (if I can put it this way) in the Trados concordance simple works better for me:). Maybe I just have to learn how to use this feature in MQ more efficiently.
    Yes, with an additional monitor, the concordance window is not a problem. Perhaps, I should try a two-monitor configuration.
    But overall I am very happy with MQ for reasons you have so convincingly exposed in your several MQ posts.

  4. @Arkadi: OK, now I understand. I haven't compared the concordance matching in TWB & MQ much. My comparisons have been more with respect to my main tool, DVX, which is currently not as good as MemoQ for concordancing.

    I can definitely recommend that second monitor. My partner used that arrangement for years; I even bought a second monitor for someone that worked with us in the summer a few years ago. After he was gone it sat unused for nearly a year, and when I finally connected it and saw the huge benefits (for the LG Flatiron Slim also the ability to rotate 90° and show me full A4 pages larger than a paper printout), I was very angry with myself for not using it sooner. Now I want a third monitor....

  5. It's gratifying to see that you can't even get your partner to use that ridiculous "TeNT" expression. Time to feed that to the hunting dogs perhaps?

    My experiences match those of Ms. Simmer quite closely. I'm looking forward to mouse editing.

    I love the way that termbases work in MQ, and the 'dump all the terms in the edit box sometimes provides something surprisingly usable.

    I'm hoping MQ doesn't improve beyond it's usefulness and develop functions that I don't need.

  6. @Kevin
    That's somewhat off the topic, perhaps.
    Do you possibly know how to convert a Multiterm base to a MemoQ termbase other than using the Multiterm convert?

  7. @Rod: Those spoiled mutts are too stuffed with roast duck and other delicacies to take an interest in chewing up TEnTs. They would probably get indigestion anyway.

    Watch this space for comments on "improvements". So far what I have seen is simply encouraging, not overblown.

  8. @Arkadi: regarding Multiterm migration, sure. But I need more info. What version of Multiterm are you using, do you have synonym structures, etc.? Most of the converters you'll find in online Yahoogroups forums and elsewhere are not good, because they assume 1:1 source/target data structures.

    Back before the new Multiterm data structure appeared with version 6.x I wrote a drag & drop WSH macro to convert v5.x MultiTerm data into complete sets of pairs for import into any program that takes tab-delimited data. Thus if four source term synonyms were paired with two target term synonyms, you would get eight pairs for your data import. The script I wrote is ugly, because I had been out of the programming game for a few years when I wrote it, but if you need a good laugh and can improve it, I'll be happy to look for it and send it to you. I think the Microsoft objects I was instantiating also had some issues with different text file formats, but I never looked into the matter much.

    Some tools also read various MultiTerm formats and can theoretically export them. Déjà Vu, for example. However, this function always used to crash for me, and I haven't re-tested it with Build 316 to see if it works now. That also means I don't know if it resolves pairs correctly. Across is also theoretically useful for data migrations; it's the only program I know that will read the exports from Star Transit (MARTIF?). But when I tried to export from my older version the results were not pretty. You might look and see if the program claims some sort of MultiTerm compatibility and test the latest version. As you know, MemoQ can export MultiTerm XML, but I'm not aware of any plans to permit an import.

    For MT version 6+, if you are competent with XSL or other tools for transforming XML, you might just do that and convert the XML files to whatever you want.

    I know that's not much of an answer. I haven't had the need to do this for a while, so I'm not currently up to date with the latest options. Maybe someone who is aware of the latest, greatest things can share a tip. But it will, in any case, be important to know what version you are concerned with.

  9. @Kevin
    Thanks a lot.
    Never used Multiterm in fact.
    I just thought there was a (relatively) easy solution. But it looks like there is not. I think I'll ask the client if he can send me a source file so I can convert it to a MemoQ termbase.

  10. It might indeed be easy, Arkadi. But version is still important. There are still companies maintaining old MultiTerm 5.5 databases. (BTW, certain older versions of MultiTerm do not need a license to operate. So if you must use a v5.5 database for some reason, this is technically possible. Of course the legalities should be sorted out with the customer and SDL.) In any case, find out the version and maybe I can give you a clear answer. Contact me privately if you need some conversion help.

  11. Arkadi

    Have you tried exporting Multiterm termbases using the Default setting (XML) and simply importing that into MemoQ? That works for me.

    Generally if something doesn't work first time, you can open XML and CSV files in Excel and hack them about until they do work. I haven't found the need to use any external other tools.

    The flexibility of MemoQ in handling legacy Trados resources is yet another of its attractions. I was overjoyed at how simple it is.

  12. @Kevin
    Thank you,Kevin. If I don't solve it myself, I'll contact you privately as you kindly suggested.

    The problem is I don't even have Multiterm installed on my machine, so I don't think I can use the default option and convert the M-base to XML (if I understood you correctly).

  13. MemoQ currently specifies CSV and TMX (an XML variant) as possible import input. I'm surprised that you say it swallows the MultiTerm XML export as "TMX". I'll have to try that; it would be a fine thing if it works.

  14. @Arkadi: The previous comment's "you" was Rod, of course. If your client doesn't know how to give you the export as you need it or you have an ancient version of MultiTerm here, we can make appropriate suggestions or provide help. Find out first what version you are faced with. That is the most important information.

  15. Sorry, I've obviously got my TMX and XML mixed up. In fact, I don't think I have attempted importing XML to MemoQ, although I do frequently simply change the extension of XML files to CSV or TXT with no problem. (I've been a bit distracted trying to retranslate the documentation for a medical product that itself hasn't been localized properly - the client was bright enough to spot that the documentation stinks, but not enough to see that the product localization is a dangerous nightmare.)

    I don't know of any way to get termbases out of MultiTerm without directly exporting them, so I'll shut up now! I'm curious to know if there is a way.

  16. There's no way around that export step. What you can export depends on the version, however. It may be hard to believe, but there are still plenty of customers with old version 5 formatted termbases.

  17. I am fairly new to MemoQ and just jumped in with the Server version after spending 1.5 hours downloading, installing, configuring, loading termbase and TM, then analysing a file. Very impressive and kudos for ease-of-use!

    Is there a way to edit text in the Alignment window? Is that something coming in 4.0?


  18. @Doug: I missed this comment when you made it. The current versions of memoQ allow editing of the source and target text of an alignment.


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