JF Delannoy 

JF Delannoy said:

How can I apply the enriched MT retroactively - to a former batch?

Can you explain what you mean by this?  What is "enriched MT"?

  Also here...

JF Delannoy said:

Not I am reviewing the AT segments, and want to apply the enriched MT to them.

JF Delannoy said:

How can I apply the MT

I am quite confused by what you are tying to do here.  Are you post-editing or applying MT?

Perhaps you can simply describe the steps you are taking in simple language we're probably familiar with.

I proceed by chunks of about 50 files.
For every segment in chunk n I am sure of, I validate, and it updates ("enriches") the MT.

Then I load the next 50 files - chunk n+1.
I'd like to apply the better MT there and supersede what can be safely superseded as being a better rendition.
I want to batch it; at present, I am Ctrl-T-ing several hundred segments one by one.

I proceed by chunks of about 50 files.
For every segment in chunk n I am sure of, I validate, and it updates ("enriches") the MT.

Then I load the next 50 files - chunk n+1.
I'd like to apply the better MT there and supersede what can be safely superseded as being a better rendition.
I want to batch it; at present, I am Ctrl-T-ing several hundred segments one by one.

JF Delannoy 

Allow me to rephrase this to make sure I understand...

1. You create projects in batches of 50 sdlxliff files at a time.  These files have been machine translated by others.
2. You post-edit the first 50 files, updating your translation memory with the reviewed translations.
3. You then create a project with the next batch of 50 sdlxliff files and want to run them through your translation memory before starting your post-editing work

If this is correct then all you need to do is run the pre-translation batch task when you create each project making sure that you use this option (perhaps!):

Or maybe you do this:

1. You create a project with all of the sdlxliff files.  These files have been machine translated by others.
2. You open up the files in batches of 50 files at a time.
3. You post-edit the first 50 files, updating your translation memory with the reviewed translations.
4. You then close these files and open the next 50 and want to run them through your translation memory before starting your post-editing work

If this is correct then all you need to do is run the pre-translation batch task on the next set of 50 files before you open them making sure that you use the option above (perhaps!).

Either way it all hinges on you using the pre-translation batch task against the translation memory you are continuously updating.

In another project, not so large:

Applying anew Pre-Translate on a translated text (after populating the TM) should give the status CM to what was with status NMT, right?

It does not work. NOK.

Yet I have the settings above

So I am presently doing Ctrl-T 400 times, segment by segments.

#

I have tried with "Translation overwrite: always overwrite" - it ruins it. I find dozens of NMTs - Draft that I had already moved to CM'ed - Translated status.

#

So back to the Ctrl-T chore.