We often lock segments before publishing a project in GS. This is done to protect previously translated content or to make sure that content that must not be translated stays untouched.

Our analysis is set to exclude locked segments and it yields correct results within Studio.

As soon as we publish a file to GS and change its phase from "Preparation" to "Translation" our locked segments re-appear in the word count although they are still locked.

This can cause "credibility" problems with our translators when we do not include locked segments in their work order but their analysis shows the higher, all-inclusive, number.

We can reproduce the issue easily with IDML files, but we have also noticed that it happens in other file formats.

Studio 2017, SR1
GroupShare 2017, SR1