DGT-OmegaT 3.2 (TEST) update 18 published

During the two last weeks, lot of tests have been done in order to make release 3.2 of DGT-OmegaT, actually in test, ready for production in a near future. Some intermediate versions have been available internally for a small amount of time, that's the reason why we switch directly from update 15 to update 18.

Major problem with search screens

Branch 3.2 of OmegaT introduced the possibility to view progress of long searches, such as pre-translation with matches or glossaries. Unfortunately, we later realized that this made small searches considerably slower, because the refresh was done too frequently. This new version does it only once per second, so that queries which take less than one second (almost everything except pre-translation) are not impacted.

Branch 3.1 did not implement progression screen, so it is not affected. Branch 3.4 will receive the same correction in the next release.

Added features

Normally this version should not receive new features because it is a release candidate. But our users wanted some features in advance. So, we implement directly in 3.2 features which have low impact on the application, to avoid introduction of new bugs. Features with more impact, or more powerful variants of features mentioned here, are delayed to release 3.4

Added features:

• Auto-adjust pre-translate prefix: in previous versions, when you selected one of the pre-translate methods (source, match, machine translation, text) the prefix was immediately changed. Now it is also true when you select a score type in matches case, so that you display the same score you used for search

• Can restrict where to search match: when you use Pre-translate by Match, now you can restrict in which file(s) you want to find matches. Note that this does not change where changes are applied: they are always applied on ongoing translations!

• Synchronize editor score display with matches pane: when you edit matches pane you can select which score(s) you want to display; now the editor's delimiter will try to display exactly the same scores, meaning that if you modify the matches pane you also modify the editor
• Make adjusted score case sensitive: in standard OmegaT all scores are intentionally case-insensitive; this is logical for word-based scores, but for adjusted score, since all elements of the segment are taken in account, users expect that score is 100% if and only if the two segments are exactly identical, which was not the case. We note that OmegaT did it intentionally, but contrarily to what is written here, we did experiment that in older versions, this score was case-sensitive. So this is nothing more than rolling back to behaviour of older versions.