Description of problem: ibus-anthy doesn't learn Kanji candidates when converting plural segments. However, it learns candidates when converting single segment. Version-Release number of selected component (if applicable): ibus-anthy-1.1.0.20090603-1.fc11.i586 How reproducible: See, below. Steps to Reproduce: 1. input single segment こうそくで -> 高速で 2. convert single segment to another Kanji for learning こうそくで -> 光速で 3. input single segment again for checking こうそくで -> 光速で 4. input plural segments (use learned Kanji) こうそくでいどうする -> 光速で移動する 5. convert plural segments to another Kanji for leaning こうそくでいどうする -> 高速で移動する 6. input plural segments again (don't use learned Kanji) こうそくでいどうする -> 光速で移動する Expected results: In step 6. it should use learned Kanji. (use same Kanji in step 5.) こうそくでいどうする -> 高速で移動する Additional info: Originally, this problem was reported on a Japanese blog. http://fun.poosan.net/sawa/index.php?UID=1248624046 On the blog entry, he said that SCIM-anthy can learn the first candidate even if converting plural segments. I checked that was true by using SCIM-anthy.
Could you try the nightly RPM if you see your problem? https://admin.fedoraproject.org/updates/F11/FEDORA-2009-8610 I cannot reproduce your problem with the latest.
Thanks! The latest package(ibus-anthy-1.1.0.20090814-1.fc11.i586) works perfect!
Closing this bug with the submitter's comment. *** This bug has been marked as a duplicate of bug 508358 ***