For the last few months I have been keeping tabs on MemoQ's latest releases, which have come out with refreshing regularity. Every so often there was enough interesting material to write about, but I always held out. Until now.

This last week MemoQ 3.5 was released, and here are the new features according to Kilgray's (the Hungarian company behind MemoQ):

·         Longest substring concordance

·         Wildcard concordance and wildcard in terms

·         STAR Transit filter

·         Bi-directional language enhancements

·         Horizontal edit view

·         XML preview feature

·         PowerPoint 2007 filter

·         Drastic server speed improvements

I downloaded the new version and specifically looked at the first three features, which I found truly ground-breaking.

Let's start with the first feature first, the oddly named "longest substring concordance" (or, just as odd: LSC). This is an attempt to automate concordance searches according to user-defined parameters (under Tools> Options> Subsegment leverage you can set the minimum number of concordance hits and/or the minimum number in words/letters/characters) and without interrupting the workflow. In the Translation results pane there are now not only translation memory matches, terminology matches, and assembled rows (available since version 3.2) but also ominous-looking matches of subsets of the string that needs to be translated with empty targets. These are the LSC matches. Clicking on them will produce the Concordance dialog, which will list all (or a predefined number of) the appearances of that particular substring in the translation memory within the context of the complete strings in the TM.

Confused?

Imagine you have this (real-life) sentence:

The starter motor rotates the engine during the start sequence, by driving it through the reduction gear unit assembly.

MemoQ might find that "the reduction gear unit" is a worthwhile subsegment and will display this in the Translation results pane with an empty target. Double-clicking on it would bring up the Concordance dialog with these options:

The starter motor rotates the engine during the start sequence, by driving it through the reduction gear unit assembly.

A coupling assembly mechanically connects the main output drive shaft of the reduction gear unit to the driven unit.

Individual accessory drive pads for the main lube oil pump are also incorporated on the reduction gear unit.

with their respective translations. If you wanted to use any of those, you would just need to highlight the part of the translation you want to use and select Insert selected.

I really like this feature because it offers a new granularity to TM materials without being obtrusive (the program does not slow you down by displaying an additional dialog like a comparable feature in Trados, plus you are free to use the displayed option or not), it enables easy paste access to the desired translation, and MemoQ's superior search-and-lookup speed allows it to operate without even seeming to slow down the search process too much.

Speaking of the Concordance dialog, that same dialog is also used for the enhanced concordance features. (A "concordance search" is the process of manually highlighting one term or phrase in the source segment, pressing a shortcut key -- in the case of MemoQ, it's Ctrl+K -- and accessing all occurrences of that in the TM.) What makes MemoQ's concordance feature attractive is the possible automatic addition of a wildcard character. In the Concordance dialog you can select the option Add wildcard to selected text, and for the current and next search(es) MemoQ automatically adds an asterisk (*) to each of the terms. This will make MemoQ look for 0 or more additional characters to the term in question, something that is particularly helpful for languages with heavy flexion. (Plus, I really like this because I always forget where the asterisk is on the German keyboard and I get tired of switching back to the English keyboard to enter it manually.)

The last feature that really caught my interest was the Transit compatibility feature. Now, Transit is a great program, but it's really different, and many folks just don't want to spend the time to learn to use the free Satellite edition (even if they might miss out on something). For those, this feature will be very useful. It allows you to import a PXF file -- this is a Transit-specific package file with the translation files, reference material (i.e., TM), and glossary data -- extract the translation files as well as the TM content, translate it, and then send it back to your client as a TXF file -- Transit's return package format.

In general it works very well. There are a few glitches -- a couple of strings (out of a few thousand in a test run) were unduly protected as tags, and I had to switch my import mode to get all my data, but it's impressive that the MemoQ developers were able to use Transit logic to "harvest" reference material that is readily available in the TM you selected when you started translating. What this feature is not able to do is automatically import the TermStar glossaries. This is somewhat unfortunate because Transit projects tend to be very terminology-heavy because of its excellent terminology tool.

Here are some other caveats with the new version: Word 2007 is still not supported -- though PowerPoint and Excel 2007 are -- and TBX, the termbase exchange standard, is also not supported yet.

This morning I had a chance to talk with the owner of a translation agency who has been using the server edition of MemoQ for a while, just to get an idea of what his take on performance and user acceptance was. He was very positive overall. Compared to other server-based products (Idiom WorldServer, Logoport, and Across) he reported equal or better server response times. He also liked the option for translators to check out resources to work offline or the online document storage. This last feature allows translators not only to share translation memories and terminology databases, but also the actual documents, which -- optionally -- can be server-based as well. Thus, multiple users can have access to large documents that can be translated and edited at the same time for faster turnaround. His assessment of that process was kind of interesting: He ran into problems with translators not working successively from top to bottom through documents, creating havoc for the poor editors; however, that seems to be less a technical limitation than an organizational one.

As far as user acceptance, he acknowledged that not all his translators were super-eager to adopt a new tool at the drop of a hat, but it helped that they did not have to pay for the program. With MemoQ's mobile licensing concept, he is able to assign temporary full licenses to his users. Interestingly -- and he was not the first to mention it -- Déjà Vu users in particular are struggling with the idea of using a different tool.

Speaking of licensing, I am interested to see how MemoQ's licensing scheme will be adopted once it's time for it. Last fall Kilgray adopted a new system in which all upgrades are free for a year after purchase, no matter how major or minor they might be; after that year, a 20% annual fee is applicable for further upgrades. This is certainly not an uncommon practice in the software industry in general, but as far as I know it is unusual for the individual user sector in our industry. We'll see what happens come fall of 2009.