Ian Barwick wrote:
I'm based temporarily in Japan and have acquired a 109 key USB Japanese keyboard for my laptop (running SuSE 9.0). It works fine, apart from the five "extra" keys (backslash, yen-sign and the three "henkan" keys around the spacebar),
The backslash and yen-sign keys are not "extra". Normally the backslash key inputs a yen-sign when not shifted or an underscore when shifted. Normally the yen-sign key inputs a yen-sign when not shifted or a vertical bar (pipe or or operator) when shifted. The unshifted forms are redundant but the shifted forms are not. Also if the font is not Japanese then the yen-sign displays as a backslash since it is the same code point, but again it doesn't matter which key it came from. Kernel developers have repeatedly broken the input of the backslash and yen-sign keys, especially on USB keyboards, but sometimes on PS/2 keyboards too. Sometimes I could fix it myself, sometimes I needed help. My fixes have always been ignored. But one time I finally convinced someone who isn't a kernel developer, and he persuaded a kernel developer to fix it, which he did, though not using my fix. If you can install a kernel from the 2.4 series (probably anything after 2.4.10 or thereabouts) then it should work for you. Rumour has it that the latest 2.6 series (probably 2.6.5) works too, but I haven't tested it yet. Other actual "extra" keys (hankaku/zenkaku/Kanji, Microsoft sperm key, muhenkan, henkan, hiragana/katakana/Ro-maji, and Microsoft other key) usually don't need to produce any input in Linux. I'm not sure if ATOK uses the Japanese henkan keys, but the usual (free) Linux IMEs don't.