Forum: Wordfast support
Topic: Can Wf Classic show TM match differences using text formatting (red/blue/strikethrough, etc.)?
Poster: Samuel Murray
Post title: @JL
[quote]JL01 wrote:
Red border means that the target text is different from the existing TM target unit when the source text is [the same]. [/quote]
I suppose you meant "when the source text is the same". Yes, that is true. Even though WFC is capable of detecting that the text is different, it does not highlight those differences, which means the translator has to compare the TM target text with the segment target text manually (and this is made more difficult by the fact that the TM target text is displayed in a smaller font).
[quote]In other words, a red border means that the target has been modified without the segment being entered into the TM, which is usually the case when target text is edited without opening the segment.[/quote]
It does not mean that the segment's target text has been modified at any stage -- it simply means that the segment's target text is different from that of the TM. There can be many reasons for this.
For example, if in the second half of the file you encounter a sentence that is identical to a previously translated sentence, but which must be translated differently, you would change the proposed match, and WFC would update the TM with that new text. Later, when you review your translation, segment by segment, and you encounter the first instance of that identical source text sentence, WFC warns you with a red border, but you still have to manually check both target texts to make a decision about whether in this particular case you should overwrite the existing target text with the TM's target text or retain your original translation.
There is no rule in translation that identical segments with identical meaning should always have identical target text. When I review a translation for which the client had said "make no changes purely for the sake of consistency, but correct only actual errors", and I find in two different files two identical segments with two different translations, then it would have been incredibly helpful to see how the two translations differ, to be able to decide whether to accept both, or not.
But yes, sometimes it is caused by the translation having been modified without opening the segment -- but it is not always feasible to open each individual segment. For example, when doing a spell-check, or doing a find/replace operation on the entire file.
Topic: Can Wf Classic show TM match differences using text formatting (red/blue/strikethrough, etc.)?
Poster: Samuel Murray
Post title: @JL
[quote]JL01 wrote:
Red border means that the target text is different from the existing TM target unit when the source text is [the same]. [/quote]
I suppose you meant "when the source text is the same". Yes, that is true. Even though WFC is capable of detecting that the text is different, it does not highlight those differences, which means the translator has to compare the TM target text with the segment target text manually (and this is made more difficult by the fact that the TM target text is displayed in a smaller font).
[quote]In other words, a red border means that the target has been modified without the segment being entered into the TM, which is usually the case when target text is edited without opening the segment.[/quote]
It does not mean that the segment's target text has been modified at any stage -- it simply means that the segment's target text is different from that of the TM. There can be many reasons for this.
For example, if in the second half of the file you encounter a sentence that is identical to a previously translated sentence, but which must be translated differently, you would change the proposed match, and WFC would update the TM with that new text. Later, when you review your translation, segment by segment, and you encounter the first instance of that identical source text sentence, WFC warns you with a red border, but you still have to manually check both target texts to make a decision about whether in this particular case you should overwrite the existing target text with the TM's target text or retain your original translation.
There is no rule in translation that identical segments with identical meaning should always have identical target text. When I review a translation for which the client had said "make no changes purely for the sake of consistency, but correct only actual errors", and I find in two different files two identical segments with two different translations, then it would have been incredibly helpful to see how the two translations differ, to be able to decide whether to accept both, or not.
But yes, sometimes it is caused by the translation having been modified without opening the segment -- but it is not always feasible to open each individual segment. For example, when doing a spell-check, or doing a find/replace operation on the entire file.