Alexey> Currently XEmacs beeps and says "'iso-next-group' not Alexey> defined" when switching keyboard layout. Do you know of a good on-line (or off-line, it might be in my library) reference for these ISO layout management facilities (in English)? Is this ISO 9995 stuff? (An alternative reference, eg, ECMA, would be appreciated, last I checked my library did not have ISO 9995.) There is another patch pending to fix a small subset of the "millions of useful keysyms not recognized" problems we still have. If possible, I'd like to do something sane with all these keysyms (both the literal input ones and the keyboard control ones) at once. Alexey> Please, apply it to current beta or tell me why it is not Alexey> needed. Of course it's needed (or at least useful enough that something with equivalent effect should be applied, if it isn't). The reason things like this don't get applied is because they fix symptoms for a subset of users, but past history makes it plain that they frequently cause problems when we try to comply with standards or fix bugs for other users later. Help us be standards-conformant, please. I think that is the appropriate attitude to take in the development branch. -- Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Ask not how you can "do" free software business; ask what your business can "do for" free software.

Alexey> - you think that according to some standard of Alexey> specification, XEmacs should do something special about Alexey> iso-next-group (and -first-group, and -last-group, btw); Alexey> or we will see that it explicitly lets application to Alexey> ignore those keysyms; Yes. XEmacs should know what to do with those, and it should do whatever it does in conformance to the spec. Alexey> - this is a part of a large problem (e.g., the problem Alexey> with unability to do Alexey> M-x describe-key RET Cyrillic_A Ooh, that's nasty. I don't know what to do about it offhand, but I bet you're right: we need to handle that down in C. You might be able to get some sense by defining that iso-next-group as a prefix key, even though it really isn't. I was thinking of something much simpler, though: the fact that we only implement about 20% of the keysyms in <X11/keysymdef.h>, although for the graphical ones it would be trivial to fix that by using define-key a few hundred times. Alexey> Stephen, how do you in your English/Japanese enviroment do Alexey> the switches between layouts? Or you do not use xkb, and Alexey> never see iso-next-group? I do use xkb, but I don't know anyone in Japan who uses a Japanese layout. Instead, even native speakers (unless they are professionally trained typists) normally use an input method which translates from roman letters to Japanese phonetics (there are only about 70 of them and they are extremely regular in construction), plus a few control keys to switch between modes. Alexey> ISO 9995 does not seem to be online, and as far as I can Alexey> tell from its description (Keyboard Layouts for Text and Alexey> Office Systems), it does not govern specifics like X Alexey> client/server communication :) My guess is that what XKB does for ISO 9995 is to define keysyms to deal with ISO keyboard control functions. So really from XEmacs's point of view this should all be transparent. It's quite possible that XEmacs should just throw them away, since XEmacs does not normally want to change the keyboard layout at the X level. Alexey> So, what's the plan? Alexey> Should we read the Xkb spec to find out if there is Alexey> something said about the need for an application to Alexey> respond to group changes keysym? That's the place to start, since it's easily available. Alexey> Should we find ISO 9995 and see if it governs something of Alexey> interest? Yes. I'll take another look around my library system. Also I can talk to the people at the Mule lab, which is only about 2 km from my apartment. Alexey> Where is that "pending patch" you're talking about? Somewhere on XEmacs Beta ... it's really very simple, though. It just amounts to a request for a bunch of (define-key global-map keysym ...) definitions that would make a Frenchman happy. Much simpler than the C-h k iso-next-group Cyrillic_A problem. -- Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Ask not how you can "do" free software business; ask what your business can "do for" free software.