Hi Giannis Ouzounis ,

First I think the TM project settings has not penalties. This would explain why you get 100 % matches when there are tag differences. Please check your settings here:

From my point of view, a small formatting penalty (such as 1) makes sense.

Now, to explain the locked segments, I think you have this configuration in the Pre-translate task:

Could you please check it as well?

Hi Giannis Ouzounis ,

First I think the TM project settings has not penalties. This would explain why you get 100 % matches when there are tag differences. Please check your settings here:

From my point of view, a small formatting penalty (such as 1) makes sense.

Now, to explain the locked segments, I think you have this configuration in the Pre-translate task:

Could you please check it as well?

Dear Jesús,

Thank you for looking into this.

The penalty settings in the project are exactly you show. The pre-translate options as well, with Lock context matches.

I agree that there would be a high fuzzy match for the TU without the tags, and the other match should have been chosen.

Next, I did two tests: First, I unlocked everything in my file, I ran a Pre-translate with Overwrite existing translation if better match found, but it didn't fix the issue. Still, the translation without the tags was there. Then, I tried to Clear Target Segment in the whole file, and pre-translate again. This time the correct TU was picked up.

I have no idea why it happened this way. A colleague of mine actually created the project, but I still tested everything on this specific project without changing any of the existing settings. Could the cause be a difference in Studio version?