Jun 1, 2012

memoQ 6: more power under the hood

Yesterday, Kilgray held its virtual memoQ Day on ProZ. Given what a disaster the ProZ online conference interface was in the first virtual conference I attended long ago, I had no intention of looking in, but since this was the first occasion on which the new version 6 of memoQ was to be presented in public, I decided to "visit" after all. The format and management have improved since my last experience with these virtual conferences, and it was definitely worth the time. I encourage those with access to the recordings of the sessions to see what's on offer; there may be many things of value. Despite over 1000 participants during the day, the event worked well.

The memoQ 6 presentation was in the evening for the Central European time zone, so I did not catch that session live. The day was long enough before then. When I did look at it, I was pleased to see confirmation of may useful changes I had heard about and see a live demo of many more I had not expected. Given that the software is still pre-beta, I expect there will be a number of changes and additions before the release, so I will simply share my impressions using the list of running feature notes I kept and a few screenshots I made. And a few addenda from conversations at memoQfest 2012.

  • A smaller footprint – previous versions of memoQ suffered increasingly from “bloat”. (That statement reveals an attitude I like - SDL, which takes up about ten times as much space and uses more resources, does not seem to be similarly inclined to streamline.)
  • Vastly improved scalability and handling of large files and large numbers of files. Speed improvements of an order of magnitude in many cases.
  • memoQ 6 has added support for binary MS Office (DOC & XLS) formats – and can be used without MS Office being installed. This new functionality also includes the server APIs for sophisticated, automated workflows – no more manual saving to RTF or DOCX, etc.
  • TM corruption problems have been decreaseda and can no longer crash servers; servers generate diagnostic exports for better troubleshooting
  • Improved concordance functions – no more result count error, additional viewing format. LiveDocs can now be opened from the concordance and edits/corrections can be made to TMs from the concordance window.Meta data can now be used as sorting and filtering criteria in the concordance.
  • "Content connectors" now in the Translator Pro edition to watch folders and automate operations such as file imports and statistics in projects
  • Versioning features will be identical in the Translator Pro and Project Manager editions. Now anyone can create a comparison view of two source versions.
  • Preferred languages will be added to the top of selection lists, saving time. This is a major improvement over the previous workaround of hacking the XML configuration file.
  • Translation results can be ordered by preference
  • The horizontal editing option has been changed to "pin" the segment being translated in the middle of the text view
  • Spelling checks for locked rows
  • Confirm locked & proofread segments in TM
  • A new QA warning shows TM matches not used
  • 4 new MT engines integrated
  • Several new server workflows, including "cattle calls" and one with a subvendor group to maintain translator/reviewer anonymity from the higher level service provider
  • Enterprise IT integration: Active Directory integration, proxy support, security protocols
  • memoQ 5 & 6 can co-exist, and each needed for the respective servers; opening a mQ5 project will migrate it to mQ6 (create an upgraded *copy* of your project). This actually sucks. There should be middleware to ensure that any client version of memoQ can talk to any server version. The current incompatibilities cause too many headaches.
  • Inconsistent translations can be extracted directly into a view in the QA module! This feature saves a regular memoQ view combined from as many documents as were checked.
  • An AUTOSUGGEST feature which mines resources to suggest target content
  • TMX as a translation document format (this has interesting possibilities for TM maintenance).
This list of notes is by no means an exhaustive representation of what we can expect in about a month. And I do believe there are a number of other interesting things scheduled for minor releases in the months to follow. The progress Kilgray has made with the "refactored" version of memoQ will benefit everyone and surely inspire further rounds of "Follow the Leader" with the competition.


There are introductory webinars for memoQ 6 scheduled in June. See Kilgray's schedule for details and sign up.

8 comments:

  1. Hi Kevin,

    thanks for your nice summary, but where did you see the Autosuggest feature? I skipped through the presentation, but I must have missed it. Anyhow it doesn't seem to be of particular importance to them.

    Kind regards

    Noe

    ReplyDelete
  2. Noe, AutoSuggest wasn't in the demo on ProZ, because the feature is said not to be stable enough to show in the current alpha version. But I know it's on the schedule, because the head of development, Gábor Ugray, mentioned it as a priority several times in our conversations, and the feature was confirmed again just last week in a chat I had with Denis Hay, who described its handling of data in a little more detail. The situation isn't much different from the period before the release of version 5 last year - the early presentations only show the major features which are essentially finished, tested and stable, and the weeks before the release see major progress. On the whole, Kilgray has been bettera t meeting its release targets than almost any other software team I know, but given the enormous changes to the code base and features in version 6, I won't be disturbed if there is some delay to get things right. Not after I used to wait years for promised minor upgrades to DVX and other tools.

    ReplyDelete
  3. Thanks Kevin for sharing your insider knowledge. Hopefully there will be more unexpected new features.
    I suggested something like an automatic LiveDoc download and alignment feature, where you just have to enter URLs. But that's wishful thinking.

    Kind regards

    Noe

    ReplyDelete
  4. The great thing with Kilgray is that you may wish, and it might come true. Just let them know your whishes.

    ReplyDelete
  5. Hi Anna-Lena,

    I told them that that in the last update there was nothing in there for me. This time at least it looks better, more speed is good for every-one.
    In my opinion the link to online resources has to be strengthened. Aligning on the fly would be one step in this direction, because we live and work in webland.

    All the best

    Noe

    ReplyDelete
    Replies
    1. Nothing in version 5? I would dispute that strongly. The final feature set was a lit different than originally announced, but in the end, version 5 offered a lot of new value for freelancers. The split implementation of versioning (and its limited features, no accept/reject mechanism) was a real disaster, but the terminology extraction features alone made my year in translation IT. Despite the current limits, it's better suited for practical use than anything similar I have seen, and they don't try to soak you for a bucketful of cash for an "extra" like others do. There is no one in this profession for whom terminology extraction should not be highly relevant and worthwhile. Those who don't get that point should be more than a little afraid. The cascading filters and ability to configure your own text filters were also very useful on many occasions, though I will admit they are probably less relevant to the average freelancer who is challenged enough with routine MS Office documents.

      Delete
  6. Hi Kevin,

    I heard your speech in favour of term extraction last year at the Powwow in Berlin, but I never used it since then. Also the cascading filter was not of any use for me. In the beginning I thought it would work with embedded Office documents but that's not the case. This actually would be a reason to give Deja Vu a try which seems to handle such files.
    In Webinar they confirmed that there'll be a predictive text feature. Let's wait and see.

    ReplyDelete
  7. Noe, I can't overemphasize the value to be gained from that term extraction feature. Over the years I have done so well with tools like that, adding a very significant chunk to profits and making a lot of customers very happy. Just today another one came back with a great project - a continuation of a badly botched financial report that got sorted out a few months ago under an impossible deadline - only because of memoQ's term extraction and QA. Oh, and LiveDocs too. Term extraction is solid gold.

    I agree that it is regrettable that only Star Transit and DVX2 handle compound MS Office documents properly. In a pinch I can always prep the memoQ projects with those tools (like I do sometimes with various versions of Trados), but I would prefer that Kilgray develop its own filter for handling embedded Office objects.

    ReplyDelete

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