kato%cs.titech.junet@utokyo-relay.CSNET (Akira Kato) (01/19/89)
I am not sure about the Sigma's activity, however, that is not the only group which discusses the Japanese input scheme on X Window System. There is a SIG of JUS (Japan Unix Society) in which it is also discussed. In the viewpoint of compatibility, one standard have to be established. My guess now is that the Japanese input scheme is rather complicated and we have to develop several candidates first. Then we can select one of them considering that it can also be applied to Korean, Chinese, and other languages easily. Sigma's proposal has not been posted for the public review, yet. In the JUS group one idea is proposed by Mr. M. Ishisone (ishisone%sra.junet@uunet.uu.net). It is under development and premature now. I introduces its outline below: The system consists of two processes: an X client who would like to get Japanese texts (A) and an X client which controls the key strokes and kana-kanji conversion(B). In his prototype implementation, actual kana-kanji conversion is performed in a separate process, WNN's jserver daemon. That daemon acts as a back-end process of B. The key issue here is the protocol between A and B. Process A first finds out which client (process) can control the kana-kanji conversion by the selection mechanism. Then A requests B through a client-message to grab the key strokes and to perform the kana-kanji conversion. B opens an input- only window overlapping to the A's image and perform the conversion interactively. (In this stage, B may talk jserver daemon over the network) Finally, B returns the result Kanji string back to A in a property. ------------ | (implementation dependent) --->| X server |--- | | ------------ | |--> invisible to A | -- selection -- | | | V | ------------- clnt msg -------------- ----------- | |---------->| | IPC | | | A, client | | B, control |<-------->| jserver | | |<----------| | | | ------------- property -------------- ----------- A member of our group, Mr. M. Nakamura, is going to participate the X conference which will be held soon. He will bring a tape so that he can demonstrate the above scheme if he can use SUN3/SUNOS3.x at the conference. Thanks, Akira Kato
rws@EXPO.LCS.MIT.EDU (Bob Scheifler) (01/19/89)
He will bring a tape so that he can demonstrate the above scheme if he can use SUN3/SUNOS3.x at the conference. I should have made this clear at the outset, but vendor demos are NOT permitted at the X Conference, nor is vendor literature. Any examples of such found will be summarily unplugged or removed. Sorry, we want to follow MIT guidelines for this event.