Neil Allen

The screenshot on the left shows that you have created your project using a Project Translation Memory.  The one on the right is what happens if you simply add a Translation Memory directly.

When you create your project you have several options... the most important being these wrt your problem:

Choose "Prepare" and you'll get a project TM.

Neil Allen said:

In this configuration, WordBird, or at least its "copy non translatables"  function, does not seem to work.

Perhaps this is something Angela Sigee needs to look at?  I don't know if Project TMs are supported with her plugin.

Neil Allen

The screenshot on the left shows that you have created your project using a Project Translation Memory.  The one on the right is what happens if you simply add a Translation Memory directly.

When you create your project you have several options... the most important being these wrt your problem:

Choose "Prepare" and you'll get a project TM.

Neil Allen said:

In this configuration, WordBird, or at least its "copy non translatables"  function, does not seem to work.

Perhaps this is something Angela Sigee needs to look at?  I don't know if Project TMs are supported with her plugin.

Ok, Paul, my mistake. This arose because for ages I have been using "prepare without project TM" and then going back and loading tmx files into the created memory AFTER the project is created, because there are no circumstances when I ever want a blind pretranslate carried out on an unknown project before I have had a chance to look at the files and possibly lock some of the segments, particularly in the case of xliff files received from other tools which may have already been pretranslated.

I recently changed to creating my projects using Analyse Only, as this allows me to load the translation memory during the project creation stage without having the unwanted pretranslation. However, this creates a "project memory" and stops WordBee from working.

Probably the least inconvenient is to carry on creating projects using Analyse Only and then remove the memory and add it back in. Ideally, in a user-friendly world, it would be possible to remove the unwanted pretranslation stage

Thanks for pointing this out, anyway.