Hi Gwen Gondry Gwen Gondry

I think this should be fairly easy. First you use the initial, complete document (not the revised version). Enter the translations you got from your 3 translators. This could probably be done with PerfectMatch or else you create a TM from the 3 sdlxliffs and have Studio "pretranslate from TM". Make sure only the newly created TM is active for that. Now you should have a complete initial version of the document with translations.

Then add the revised version to the project and select the just created complete initial version sdlxliff for your PerfectMatch step.

I think this should work, and it should not take long.

Daniel

Hi Daniel,

This indeed looks like the most efficient and safest option. I was originally reluctant to go through a pre-translation process to avoid any risk but I suppose there is no risk if the TM we use only contains the original document and nothing else.

I'm wondering now if there's a real value-added in performing a Perfect Match task.

We could probably just use the pre-translation and lock context matches. I'm having a hard time figuring out what more we'd be getting from the Perfect Match, unless I'm missing something...

Thanks a lot for your help!

Gwen

Gwen Gondry

You could try PerfectMatch first, and if that fails try pretranslate. Context Matches are pretty safe in my experience - unless you have lots of little files with just one Segment per file, but this does not seem to be your scenario.

Would be good to get Paul Filkins opinion on what advantages a PerfectMatch has over a Context Match.

Daniel

Gwen Gondry

I just tested this a little and I did this:

1. created a project with a large... ish file
2. split the file into three using split and merge
3. translated the three split files as a separate project
4. ran perfect match three times selecting each of the three translated split files one at a time

This works perfectly and has the following advantages over using your TM:

1. the result is a Perfect Match.  This means that in situations where you might have had inconsistent translations but it was important to keep them, the updated file will use the correct one.
2. the result can retain the original fuzzy values, comments that were in the files and segment statuses.  If you use your TM everything becomes a 100% match and you lose any other context that might have been important.

In a nutshell these are probably the most important advantages.

Gwen Gondry

I just tested this a little and I did this:

1. created a project with a large... ish file
2. split the file into three using split and merge
3. translated the three split files as a separate project
4. ran perfect match three times selecting each of the three translated split files one at a time

This works perfectly and has the following advantages over using your TM:

1. the result is a Perfect Match.  This means that in situations where you might have had inconsistent translations but it was important to keep them, the updated file will use the correct one.
2. the result can retain the original fuzzy values, comments that were in the files and segment statuses.  If you use your TM everything becomes a 100% match and you lose any other context that might have been important.

In a nutshell these are probably the most important advantages.