Hi there,

The negative word count is caused by merging/splitting segments and can be resolved by running Batch Tasks/Translation Count.

Here is a photo illustration of the process:

Untouched segments, 43 untranslated chars:

Merged segments, 86 untranslated chars (the amount doubled):

Merged segments translated, 43 untranslated, 43 translated chars (still doubled):

Segments after runing Translation Count (back to normal):

Hope this helps,

Levente

I see that this is over 8 years old, and in the mean time (especially with the last few releases of Studio, including 2021), the problem, instead of being corrected, has got worse. I’m not convinced that negative word counts and similar absurd results are only due to merged or split segments: I’m occasionally getting them also when I do not split or merge. So, as I work through a document, I see the word count decrease towards zero... then go below zero, with the indication that I have already translated 104% of the document, and no indication of how much more there is to translate. At that point I have to run again a batch task, just to have an idea of the correct word count, only for the same behavior happening again later in the day.

SDL (now RWS) programmers should take a look and fix this problem, in a way that does not occur even when splitting or merging segments. Take a look at memoQ: their word count is rock solid.