Heinrich Pesch wrote:
I use translate to fuzzy icon or the alt+*-shortcut. I wonder if Jerzy's Ctrl+alt+plus works differently?
When the document is pretranslated, the "translate to fuzzy" command (ALT+*) opens every segment and stops when it finds a fuzzy match.
The "save close open next no 100%" command skips pretranstalted segments from TM (100% match) and opens only the next fuzzy match. With long documents and a lot of pretranslated text this saves you much time.
When the relation between known segments and unknown (less than 100% match) is high (much old text but little new) you might export the unknowns to rtf and count, if you get the same result as in the analysis. I had in at least two cases (with two different customers) a lot more exported words than the analysis had indicated.
Never tried that, so I can´t tell you why.
The problem with those 100% matches really is, that it does not mean that the match from the TM reallys fits into that context you are translating at the moment. At least when translating from German you very often have to decide, if the verb is infinitive or imperitive. "Kabel anschließen" in a capture has to be translated differently than "Kabel anschließen" in an order list. At least in the languages I know.
I always have a bad fealing when using translate to fuzzy, but what to do when the customer does not pay for 100% matches. Close your eyes and I'll kiss you, tomorrow I'll miss you!
Well, it is not different, when translating from German into Polish. For example some expressions are written differently, when masculinum or femininum is meant.
Therefore you need to allow more 100% matches in your TM. In such cases - with the penalty for more 100% matches set - the automatic translation will always stop there and let you decide, which 100% match you chose.