• State Verified Answer
  • Replies 17 replies
  • Answers 1 answer
  • Subscribers 435 subscribers
  • Views 6849 views
  • Users 0 members are here
Options
Related

Studio 2017 SR1 wordcount goes nuts

Ann De Visscher

Hello all,

I downloaded SR1 the day before yesterday. It doesn't change much to the issues I'm having since purchasing 2017, multiple crashes and very (very very) slow context searches for segments with a single all caps word...

But a new issue is the constant faulty translated words wordcount (you know, below and on the right ot the screen, when your translation is finished but trados says there are still words to translate and then you have to perform the "number of translated words" (nombre de mots traduits) task to have the right number or percentage of translated words.

Besides, I often import tmx files in an existing sdltm file when creating a project and it goes slower than with the first version.

Furthermore the preparation of the project is also slower than before.

So, not very impressed with the corrective at the moment...

 

Cheers to you all

 

Ann
Rate translation
×
Suggest better translation
×
Moderator UI
×
  • Get AI Suggestion
Parents Reply Children
  • 0 in reply to Paul
    Hi Paul,

    Thank you for your answer. I did try this solution, but to no avail. Besides, for the moment, slow search is the least of my problems. I just wish Studio wouldn't shut down every five minutes. Sometimes with an error message, but sometimes, it just vanishes promptly. Now with SR1, I can add that wordcount problem for wich I systematically must run the translated wordcount task after translating a file. And, more than before, I have the validation error that makes it impossible to clean the file. (I have to use Shift+F12 in editor mode to have a clean file). All those problems are only encountered with mqxlz files or xlf files prepared with a memoq engine (my biggest client :-( so I don't think this is hardware related...

    Best regards.

    Ann
  • 0 in reply to Paul
    And I just encountered a new problem with SR1 : when editing a translation unit (F2) in translation results, the translation unit stays the same even after pressing save... so editing is not possible anymore...
  • 0 in reply to Ann De Visscher
    Hi Ann De Visscher,

    >when editing a translation unit (F2) in translation results, the translation unit stays the same even after pressing save.
    We had this problem too. You can fix it by disabling Lookahead. It's under File > Options > Editor > Automation

    The word count problem I have heard from several people too. Currently your workaround to run the batch task is the only way I know...
  • 0 in reply to Jesse Good
    Thank you for your answer, Jesse! Editing translation units works again. All the other issues stay the same, even after a complete reinstall last weekend (with antivirus deactivated). I guess I will wait for SR2, but I wonder if there is a compatibility problem with memoq... I lose a lot of time (and patience) with all this.
  • 0 in reply to Ann De Visscher
    What makes you think there is a compatibility problem with memoQ? If you are referring to the wordcount issue... this is a problem generally that we don't seem able to fix. I think a better indication of progress is to use the WIP report in your batch tasks. The indication in the bottom of the taskbar has been a refresh problem since Studio 2009 and we have never been able to resolve it. Frankly I'd remove it altogether until until we can make it work reliably.

    Paul Filkin | RWS Group

    ________________________
    Design your own training!
    You've done the courses and still need to go a little further, or still not clear? 
    Tell us what you need in our Community Solutions Hub

  • 0 in reply to Paul

    Dear Paul and Jerzy,

    Thank you for taking the time to answer my post. English is not my main language and so I hope I will be able to be as clear and complete as I can.  

    Paul says : “What makes you think there is a compatibility problem with memoQ? “

    The issues I’m having are only occurring with memoQ files, that’s why I thought maybe that was the reason, but of course I can’t be sure of that. The main issue is the constant crashing and about that, I should maybe give some more context. This only happens with memoq files, of which a big part is pretranslated, with many segments consisting of only one or two words in all caps (always takes time to find segment matches) and since a big part is pretranslated, I rapidly browse from one segment to the other (I still have to check the pretranslated segments for errors), it then feels like the segment match search is saturated because of the fast segment change. Then, I stop in one segment, wait for a segment match and the program crashes. Sometimes with a message and sometimes it just suddenly disappears on me. So I have to restart the program which takes 2 or 3 minutes and makes me lose precious time… I also have to say that these are files in which I do a lot of segment merging.

    Jerzy says : “The project preparation (here the analysis and populate project translation memories or pretranslate task) ARE slower, because they use fragment matching and match repair. When preparing a project, switch this option off and check then. I do that very often - I switch this option off for preparation and finalization of projects. However, as preparation is usually done in parallel to other work, it does not hurt that much, so I can leave the setting switched on for this. »

    I will try this, but I can’t say I’m glad to have to do this since I would expect improvements from a new version of a program instead of slower performances… Sometimes the deadline is short and you want to begin your work as soon as possible. Sometimes it takes more than 30 minutes to prepare a project. I also wonder : if fragment matching and match repair are off when preparing the project, will the segment search while translating not be even slower than it already is? Sometimes I wait for 3 to 5 minutes before I get a segment search result.

    Jerzy says : “As for the wordcount in the lower right corner - this is just an orientation, very useful and much better than just a percentage. But to have exact numbers you can switch to Files view, where you can have the exact number of words still left to translate. If this is not shown correctly, just run "Translation wordcount" batch and the data will be exact again. »

    About the wordcount errors : it’s not that I really absolutely need to know the wordcount at all times. I just wanted to state that this problem occurred when starting to work with 2017 SR1. Never had that before or very rarely, and the real problem is that I always have to do the “translation wordcount” task to have a 100% translated file before cleaning it, which, again, takes time.

    My hardware is fine, I think, I work with windows 10, 1.80 Ghz processor, 8 ram, 64bit processor.

    All these issues occurred since upgrading from Studio 2015 to Studio 2017. I already had a contact with someone from SDL not so long after the upgrade, but no solution was found.

    I’m glad to work with SDL, I’m used to it, but I really wish I wouldn’t lose so much time on technicalities especially since I’m going to be on a tight schedule soon going back to school at 43

    Again, thank you very much for your time.
  • 0 in reply to Ann De Visscher
    In this very specific case, where your file is more pretranslated than not and you have more short segments, do not use fuzzy match repair and switch off the look ahead feature. This two are responsible for the delays. In your case mainly the Look ahead feature is of no use, as you stop in pretranslated segments, while the TM has already prefetched the next hit for the next NOT translated segments. This is why it then takes so long.

    In the end of the day it is like with all new products in life. A new TV has many improvements, but we mainly use only few (like the Internet access to watch movies via my Amazo account). What we never use is Internet browsing with the TV or media streaming from our network. So is the TV really better than the old one? Same with improvements in Word, where an average user has never utilized more than 10% (my guess) of program features.
    Studio has many improvements, which do not necessarily fit in every job. So for proofreading jobs for example all fragment matching and fuzzy repairs are of really no use, so it is better to switch them off.
    This is like driving a car - when you are in the city, you drive slow and can switch it in Eco mode (at least it is what I can do in my car). The engine response is like would you have removed at least 50% of its power. The accelerator pedal feels like rubber - you keep pressing it, but nothing really happens. But you are driving with speed limit to 50 km/h, so it is not necessary to use the whole engine power, which would bring this very same car in less than 5 seconds up to 100 km/h. Your benefit is reduced fuel consumption. But then you go with this car on the motorway ans switch it to Sport mode. It now feels like would you have a completely different model. Not only it takes less than 5 seconds to go for 100 km/h, but even if you are fast or very fast and then still push the accelerator pedal further down, the car jumps forward like a jet during take-off. So is Studio too - exactly adaptable to your needs. And when you do that, your benefit is less effort for better quality in shorter time.

    _________________________________________________________

    When asking for help here, please be as accurate as possible. Please always remember to give the exact version of product used and all possible error messages received. The better you describe your problem, the better help you will get.

    Want to learn more about Trados Studio? Visit the Community Hub. Have a good idea to make Trados Studio better? Publish it here.

  • 0 in reply to Jerzy Czopik
    Dear Jerzy,

    I'm very sorry for this late reply. I tried deactivating fuzzy match repair and I must say, while I didn't yet have to translate files with which I generally have the biggest issues, SDL Studio doesn't crash anymore (one single crash in nearly two weeks)!
    So I think your analysis and solution were perfect.
    I still think that it's weird that a feature could (ever) crash the program and cause so much problems, but hey, I'm glad that I will lose much less time now.

    Thank you very very much!
  • 0 in reply to Paul
    Regarding "Studio 2017 SR1 wordcount goes nuts", I personally found out that the problem, among other circumstances, appears as soon as you start splitting long translation segments into shorter sub-segments.
    Another bug brought by SR1: after some time the project TMs stop suggesting even obvious analogies (I personally adjusted the lowest analogy limit at 70%) and I must restart Studio to get the TMs working again!
    Regards.
    Gérard LAMBERT
  • 0 in reply to Gérard LAMBERT
    Hi ,

    Have you tried turning off Lookahead? This may account for the apparent non working TMs that work when you restart Studio. Look in File -> Editor -> Automation

    Paul Filkin | RWS Group

    ________________________
    Design your own training!
    You've done the courses and still need to go a little further, or still not clear? 
    Tell us what you need in our Community Solutions Hub