We are trying to evaluate the impact of issues with Japanese Full Text Search on the "Find potential duplicates" search feature in Work Items view. (See jazz.net defect 206451 and enhancement 230165).

For RTC 3.0.x, we know that the Full Text Search field in the lower left hand corner of Eclipse client [with "Search for ID or Text"]  will search for work items and match by score according to how many characters found in a variety of work item fields [such as Description and Discussion] match the search string.

1. Does the "Find potential duplicates" feature [available from Work Items view when right-clicking on a work item], use the same search engine as the Full Text Search engine? Or does it work independently? We ask this because we need to know if the "Find potential duplicates" feature is affected by the same issues/limitations of the Full Text Search engine.

2. Does the "Find potential duplicates" feature rely on the "Summary" string in work items to find hits?
--> From our testing, the Full Text Search is similar but different. For "Find potential duplicates" it will not find another work item because of text found in other work items in fields other than the Summary field. It seems to rely only on the Summary field. Is this understanding correct?

3. How does the "Find potential duplicates" feature find hits? What does it compare?

4. If the behaviour of RTC is different for our questions 1, 2 or 3 above, according to RTC version, please explain the change.
--> We are aware that there have been changes to the Full Text Search engine from RTC 4.0, but not aware of fundamental changes in implementation that would affect the answers to our questions 1-3.

Many thanks in advance.