Pavel Tsvetkov

Pavel Tsvetkov said:

This is a known problem but according to RWS, it is not a bug but a feature. The idea is to discourage people from sending bilingual files for review and then importing them back. RWS wants everyone to buy a Trados license and use that.

Sometimes I'm lost for words when I read the comments in here.  Where did you read this, even "albeit not directly"?

yyukimura said:

I am trying to import the reviewed Bilingual Review File

The reason we have this feature in the software is exactly because we want people to be able to use it.  I can only think Pavel is referring to something else.

There can be any number of reasons why the bilingual review file fails to update the Studio project correctly but it's hard to identify why you are having a problem without being involved in the process you have gone through.  Perhaps this article will be of help?

https://multifarious.filkin.com/2014/07/26/efer-ad/

Pavel Tsvetkov

Pavel Tsvetkov said:

This is a known problem but according to RWS, it is not a bug but a feature. The idea is to discourage people from sending bilingual files for review and then importing them back. RWS wants everyone to buy a Trados license and use that.

Sometimes I'm lost for words when I read the comments in here.  Where did you read this, even "albeit not directly"?

yyukimura said:

I am trying to import the reviewed Bilingual Review File

The reason we have this feature in the software is exactly because we want people to be able to use it.  I can only think Pavel is referring to something else.

There can be any number of reasons why the bilingual review file fails to update the Studio project correctly but it's hard to identify why you are having a problem without being involved in the process you have gone through.  Perhaps this article will be of help?

https://multifarious.filkin.com/2014/07/26/efer-ad/

Hi Paul,

Thank you for your comment.

Here is a post of mine that is now 2 years old: https://community.sdl.com/product-groups/translationproductivity/f/studio/26216/sld-trados-2019-is-still-incapable-of-properly-importing-reviewed-bilingual-files-but-is-this-a-feature

I have been a professional translator for 14 years now, and for a considerable part of this period I have used Trados. Importing back a reviewed bilingual file fails in about 80% of the time I have tried it. It either fails completely (no changes are reflected whatsoever), or it fails in part (some changes are reflected), which is even worse as I have to manually check change by change to make sure that they are all there. It is simply unworkable.

Now, I have come to believe that this is how SDL/RWS wants it to be. It is either that, or these two companies have been unable to properly implement this "feature" over a very long period of time. How is that possible?

Also, if you take a look at that previous discussion, you actually directed me to make a suggestion on http://ideas.sdl.com for the empty segments in an exproted bilingual file to then be recognized and imported as new translations. This was flat-out rejected by SDL:

Hi,

We are not considering this change. By design, this feature does not support "reviewing" empty segments - i.e. translating from scratch is not supported in the bilingual export. A CAT tool should be used to translate any segments from scratch to ensure best productivity. So, as a consequence, it is important to make sure target segments are all available in the exported file. (Again, it's meant for review, not from scratch translation).

Thanks

Daniel

So, my take on all of this is that SDL/RWS have included the "feature" but have handicapped it by design, and it was also made very unreliable (it simply does not work most of the time), because the idea is for people to buy licenses for Trados not to use Word to review or translate and then import in Trados.

It is always the case with companies who enjoy a leading position on the market. Canon does the same with their cameras: they can make the best hybrid camera right now, but instead they make photo cameras with limited video capabilities, and they have a separate line of video cameras that cannot be used to take photos. Would customers want to buy the best hybrid camera in the world made by Canon? Yes. Can Canon make such a camera? Yes. Will Canon make it? No.

So, this explanation "a CAT tool should be used to translate any segments from scratch to ensure best productivity" just gave me a glimple at SDL in-house thinking at the time. It is not only that the tool has been handicapped against the wishes of paying customers, but it was actually done in their best interest. Orwellian language.

Best regards,

Pavel
 

Pavel Tsvetkov

ok.  I see where you are coming from now.

Pavel Tsvetkov said:

my take on all of this is that SDL/RWS have included the "feature" but have handicapped it by design, and it was also made very unreliable

This is of course not true at all.  I can recall having a discussion with the memoQ team many years ago when we first introduced our version of this feature.  Our development team actually tried to make this more useful by supporting features not supported by memoQ at the time.  It was an impressive piece of work, but also risky because it's so easy to break it.  Hence the reason it was intended to be used only for review, which means that you have already translated the file and have some light edits on it.

Prior to this we didn't have this feature at all and we only introduced it because it was thought to be a missing feature some users wanted.  Most of our high volume users would not use this at all as they need a more controlled environment.

I think your logic is unfortunately coming from a place of negativity about us as a company.  The sort of reasoning you have gone through simply doesn't happen.  If we introduce a feature it's because we want it to be available and for the most suitable purpose... which we have always said was review.

If your workflow fails 80% of the time I'd be interested to know, and I mean really know, how you are managing this.  It sounds to me as though you are abusing the feature and expecting more from it than it is designed to be used for.  Whilst I can sympathise with you if you need a workflow that works mainly outside of a controlled environment, the best advice I would give you is to find a tool that supports it better.  We don't make any secret what this feature is intended for in Trados Studio, never have.  There is no Orwellian theory at play here.

It's more a case of use the tools for the jobs they are designed for, and if they don't do what you need then find one that does.

Hi Paul,

Thank you for sharing your thoughts.

Paul said:

If your workflow fails 80% of the time I'd be interested to know, and I mean really know, how you are managing this.

In some situations, it is the only available option. I sometimes work with doctors and subcontract medical projects to them. They know medicine well as they have the education, but they are not linguists and they refuse to use a CAT tool. They cannot be persuaded to buy Trados and even if they could find it for free, they would not use it.

So, because the review feature has been implemented so poorly:

 – I cannot reliably send a file for them for review as I am not sure that when I get it back it will be imported correctly. It fails almost every time.

 – I cannot use the workflow where I use Trados to create the project on my end, and then send the actual file for translation to them, and then import it in Trados to edit it and deliver. Even if I manually add some text to all target segments (not to send them empty), I still cannot send the file for review, and then import it back: most of the time Trados will still fail, or will implement only part of the edits.

So, back to square one: the maker of Trados is not interested in making this work. What decision makers at SDL/RWL fail to understand is that some people will never buy Trados, and would not even use it under any circumstances. Trados or any other CAT tool. These are not potential customers that you could win over by making the purchase of Trados the only viable option. You will never have them as customers.

But by refusing to make this feature a reliable function of Trados, first SDL and now RWL affect a large segment of the professional translation industry, and it has surely affected my bottom line over the years. And I have been loyally paying you for Trados since version 2007. That is 14 years now.

So, again, I apologise if my comments hurt someone's feelings at SLD/RWL but I just state the facts as they exist. And as a paying customer for 14 years now, I hope at least I am entitled to stating the obvious.

Best regards,

Pavel  

Pavel Tsvetkov

Pavel Tsvetkov said:

I apologise if my comments hurt someone's feelings at SLD/RWL but I just state the facts as they exist. And as a paying customer for 14 years now, I hope at least I am entitled to stating the obvious.

You're not hurting anyone's feelings.  You do seem to be determined to use the solution for something it's not intended to be used for.  Hardly our fault!  It's not a poorly implemented solution, just a misused one.

Pavel Tsvetkov said:

I sometimes work with doctors and subcontract medical projects to them. They know medicine well as they have the education, but they are not linguists and they refuse to use a CAT tool.

So in this case I would suggest you take responsibility for the process properly and own it in a way to avoid where the problem are.  If they do the translation work for you why don't you send them an Excel file instead where they are very unlikely to break the file.  Then you have two options at least:

1. copy the excel target column into the bilingual file for review that you create yourself, immediately after creating it, or maybe
2. convert the translated Excel to a TM and pre-translate the project.

If the people you work with won't use a CAT tool then I imagine they won't be paying much attention to any tags either, so probably better leave this sort of thing to someone like yourself who knows how to deal with them if they are there.

Thank you, Paul!

I will try working around the problematic export for/import from review function as you suggest, but in the "tight deadlines and high expectations" environment I function in, this will most probably not be a workable option if it takes more time to use, and may still fail at the import stage.

Now, back to my post from 2 years ago. When you directed me to make a suggestion on http://ideas.sdl.com for the implementation of a new feature: the new translations in an exproted bilingual file to be recognized and accepted by Trados at the import stage, the idea was gaining a lot of traction – professional translators and paying customers were voting it up. You voted for it as well.

It was moving up the ranks so quickly that SDL had to step in and explain they would never implement it, no matter what paying customers wanted.

So, outside the realm of conspiracy theories, why would SDL refuse to implement a feature paying customers support and want?

Best regards,

Pavel

Pavel Tsvetkov

Pavel Tsvetkov said:

So, outside the realm of conspiracy theories, why would SDL refuse to implement a feature paying customers support and want?

I actually think Daniel has been very clear there.  Your idea was about using this workflow to translate from scratch, so you were sending empty target columns and the tool has never been designed for this.

You are basically after a way to allow non-translators to translate in MS Word with no controls whatsoever and Daniel is right not to consider this. I don't believe the ROI is really there even if we did consider it.  We would have to completely change the way this feature works and its simply not worth the effort.  We are getting closer and closer to a very smooth online/offline working environment and giving users like this access to a browser environment to translate in is a far better solution, and so simple to use even a Doctor could use it ;-)

I'm even surprised that you don't want a better solution than exchanging Word files.

Notwithstanding this, perhaps you should consider just asking them to translate in a new word file and then you can align them.  There are so many workarounds to this minor problem (including the mentioned copy source to target before you create the file) I would not be very happy to see developer effort going into this feature as a completely new thing when we have so many other things to do that should take priority.

We have many customers, and frankly I'd be surprised to see a significant number approving your suggestion as a priority.

Pavel Tsvetkov said:

But by refusing to make this feature a reliable function of Trados, first SDL and now RWL affect a large segment of the professional translation industry

I doubt this.  I think the professional translation industry would prefer a more robust solution for translation with proper control, and this is exactly what we are striving for.  Not a hack using an exchange of word files.

Thanks for the link to this article, it helped me solve my problem with import in a very simple way. You've saved my day!