Last modified: 2014-04-15 06:22:15 UTC
The suggestions from translation memories already tell you how much the two (source text) strings match, but then it's left to the translator to look for the differences, to start with. As a translator, I want to see what are the changed/removed/added words so that I can quickly change the proposed translation to match my original. Drawbacks: 1) visual clutter, 2) [IMHO] wrongly suggesting that those are the only parts requiring check, while the whole suggestion might be wrong in the new context.
*** Bug 38581 has been marked as a duplicate of this bug. ***
Pau, can you assess if there is any benefit or this feature request from a usability perspective? I'm skeptical at first, aside from if there are algorithms that can identify which parts cause the Levenshtein distance.