When we create a new version from an older one, we need to do a Translation Merge from the older one to keep 75% similar texts which has been erased by the updated documents.
This process works well BUT for texts that are in fuzzy state before the TM.
Indeed, fuzzy texts are not recorded inside the translation memory, and thus not copied from it because of that.
As our reviewer use them to indicates to the translatator that something needs to be discussed, it breaks their review.
It's possible that Zanata works better if the trnaslation is refused instead of fuzzied.
I need to test this possibility on the debug-version (eg by refusing some text and creating a second debug-version from the first)
When we create a new version from an older one, we need to do a Translation Merge from the older one to keep 75% similar texts which has been erased by the updated documents.
This process works well BUT for texts that are in fuzzy state before the TM.
Indeed, fuzzy texts are not recorded inside the translation memory, and thus not copied from it because of that.
As our reviewer use them to indicates to the translatator that something needs to be discussed, it breaks their review.
It's possible that Zanata works better if the trnaslation is refused instead of fuzzied.
I need to test this possibility on the debug-version (eg by refusing some text and creating a second debug-version from the first)
I have added comments on the french translation in the debug-version, inside the file Abandoned House
Both have been rejected.
I have created a new version debug-version-comments which should allow to made some tests.
The version is still in "creating" state at the moment :)
I have added comments on the french translation in the `debug-version`, inside the file `Abandoned House`
Both have been rejected.
I have created a new version `debug-version-comments` which should allow to made some tests.
The version is still in "creating" state at the moment :)
Unfortunately after tests I am unable to find a way to keep these comments with fuzzy or rejected translation :'(
I appears than when a text is changed, Zanata assumes comments are non-consistent anymore. Which is a way to see things.
So when the translation is fuzzy or rejected, the copy also assumes it's not worth to keep.
I have no ideal solution at the moment, sorry fot that.
Unfortunately after tests I am unable to find a way to keep these comments with fuzzy or rejected translation :'(
I appears than when a text is changed, Zanata assumes comments are non-consistent anymore. Which is a way to see things.
So when the translation is fuzzy or rejected, the copy also assumes it's not worth to keep.
I have no ideal solution at the moment, sorry fot that.
When we create a new version from an older one, we need to do a Translation Merge from the older one to keep 75% similar texts which has been erased by the updated documents.
This process works well BUT for texts that are in fuzzy state before the TM.
Indeed, fuzzy texts are not recorded inside the translation memory, and thus not copied from it because of that.
As our reviewer use them to indicates to the translatator that something needs to be discussed, it breaks their review.
It's possible that Zanata works better if the trnaslation is refused instead of fuzzied.
I need to test this possibility on the debug-version (eg by refusing some text and creating a second debug-version from the first)
There are two points:
I have added comments on the french translation in the
debug-version
, inside the fileAbandoned House
Both have been rejected.
I have created a new version
debug-version-comments
which should allow to made some tests.The version is still in "creating" state at the moment :)
Unfortunately after tests I am unable to find a way to keep these comments with fuzzy or rejected translation :'(
I appears than when a text is changed, Zanata assumes comments are non-consistent anymore. Which is a way to see things.
So when the translation is fuzzy or rejected, the copy also assumes it's not worth to keep.
I have no ideal solution at the moment, sorry fot that.