Michael> The following has been an annoyance for a while, but now I decided to speak Michael> up :-) Michael> I have a habit of making the useless caps lock key to act as control key, Michael> which I do in .Xmodmap. I also do things with Alt_R and such. Michael> When XEmacs starts up, it gives me a mouthful. Why does it Michael> care? What you are doing is not ICCCM-compliant, I believe. The ICCCM says: Convention Clients should determine the meaning of a modifier bit from the KeySyms being used to control it. which seems to mean that xemacs should interpret your xmodmapped capslock key with its _original_ meaning. You have to have one of the Control keys generate control. You want to do something like this: (editing required) xmodmap \ -e 'clear lock' \ -e 'clear control' \ -e 'keycode 126 = Control_L' \ -e 'add control = Control_L' \ -e 'add control = Control_R' # if you have a Control_R key But users typically don't figure this out, and what you're doing typically works even though it's illegal, so maybe we should turn the warning off. Has this been FAQized? If not, it should be.

OG> On Thu, Jan 18, 2001 at 04:20:29PM +0900, Martin Buchholz OG> wrote: OG> What should be FAQized, the fact that XEmacs spews you a OG> warning, it may be useful to actually read it? Give Jamie the last word, otherwise you'll regret it. http://www.jwz.org/gruntle/autobogotification.html -- University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Institute of Policy and Planning Sciences Tel/fax: +81 (298) 53-5091 _________________ _________________ _________________ _________________ What are those straight lines for? "XEmacs rules."

Michael> When XEmacs starts up, it gives me a mouthful. Why Michael> does it care? Because XEmacs does very complex things with keymapping given the chance. MB> What you are doing is not ICCCM-compliant, I believe. MB> You have to have one of the Control keys generate control. Michael> Why? I have already assigned control to caps lock. I No, you haven't (I'm deliberately interpreting control -> Control keysym, as opposed to control -> control modifier bit) . That's not your fault; X11 key mapping is brain dead. (Well, actually, to be fair it was designed before our ancestors had evolved forebrains.) Michael> don't have keys to waste. I actually bind the ctl key Michael> (on the PC eyboard) to save-some-buffers ! Martin isn't speaking clearly. What he meant is "you have to have a key whose keysym is Control_[LR] generate the control modifier." If you have hands like Akeem Olajuwan, you could put the control modifier bit on KeypadMinus---but you also need to put a Control keysym there. Michael> I do something similar: Not really. Michael> remove Control = Control_L Michael> remove Lock = Caps_Lock Michael> keysym Control_R = Caps_Lock So far, so good. You now have no key that can do lock, and you have a key in the lower right corner that does control. That key generates the keysym Caps_Lock. The key in the lower left corner generates the keysym Control_L, and the key to the left of the "A" generates the keysym Caps_Lock. Michael> add Control = Caps_Lock You have added the control modifier to the key to the left of "A" and (redundantly) to the key in the lower right corner. Michael> keycode 37 = F22 I assume this keycode corresponds to the key in the lower left corner? That key now generates the keysym F22. You have no Control keys anymore, although you can generate control keystrokes. Summarizing: Physical Keysym Modifier ------------------ --------- -------- Left of "A" Caps_Lock control Lower left corner F22 Lower right corner Caps_Lock control Michael> It is probably [keycode 37 = F22] that causes the warnings Nope. The last line is completely unexceptionable. Look at the errors: 37 != 0x6d or 0x42. MB> what you're doing typically works even though it's illegal, It depends on whether it matters if you have a Control key or not. It probably doesn't. Unless you're Ben Wing, and can't hold down Control and another key simultaneously on very bad days. Most of us aren't Ben, and can chord, so we typically think "it works." Howevva ... some implementations of menu accelerators "do something" (like highlight the menu bar) as soon as you hit the Alt key. This will break, for everybody, if there is no Alt key. Doesn't matter whether or not there is an Alt modifier that allows you to produce keystrokes like Alt-F. Michael> I think it should be illegal to sell keyboards that have Michael> Control in the wrong place :-) Caveat emptor. :-) MB> so maybe we should turn the warning off. Michael> Please do. Please don't. I'll have to revert it: Jamie knows my email, and I don't look good as he does in "black and crispy." If you have to do something, distribute xkeycaps(1X). http://www.jwz.org/xkeycaps/ Michael, it's not your fault that X11's keyboard model is braindead. You also (probably; 200 keyboards is a _lot_ of keyboards[1]) don't have to do anything about it personally, thanks to the fact that the halflife of Jamie's brain is measured in months (or maybe years). <RANT> We, on the other hand, are going to undoubtedly be dealing with braindeath of the sort that caused the (now obsolete and removed, praise be) Mandrake-bashing entries in PROBLEMS, because Linux (and probably the BSDs) are spawning efforts to standardize and debase^W make user-friendly all user interfaces. For years. It is _not_ true that XEmacs is no more complex than any other app (for God's sake Martin, you're the Bullshit/Defeat guru for this project! you should know better). And it's supposed to get more complex (have you forgotten that Ben plans to add menu accelerators for keyboards that have more than 1 extra modifier key?) </RANT> Footnotes: [1] And none of them are JIS, and only about 2 know anything about Japanese. But that's OK, XFree86's (and Linux's) Japanese keyboard data are all broken anyway. Another day, another <RANT>. -- University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Institute of Policy and Planning Sciences Tel/fax: +81 (298) 53-5091 _________________ _________________ _________________ _________________ What are those straight lines for? "XEmacs rules."

Matt Tucker <tuck(a)whistlingfish.net&gt; writes: It is also my experience that, when I do things right, XEmacs no longer complains. I know because I set up caps lock to be control, left windows key to be Alt, left Alt key to be meta, and the two right-hand-side windows keys to be Super and Hyper. As long as you resist the temptation to just add the appropriate modifier key to the keysym, everything should work.

Matt> -- "Stephen J. Turnbull" <turnbull(a)sk.tsukuba.ac.jp&gt; spake thusly: [...] Matt> It might also help if someone were to point out the commands Matt> that need to be added to the current configuration to get Matt> the appropiate behavior. The omission was deliberate. Getting it right is a job for Superman, or a computer. Pay no attention to the man behind the curtain. Or in front of it, for that matter. Just get XKeyCaps.[1] FAQ patch coming up. Footnotes: [1] Jamie did not pay me to write this. -- University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Institute of Policy and Planning Sciences Tel/fax: +81 (298) 53-5091 _________________ _________________ _________________ _________________ What are those straight lines for? "XEmacs rules."

Stephen J. Turnbull writes: and he could...

On Thu, Jan 18, 2001 at 01:52:34AM -0500, Michael Kifer wrote: The semantics of the modifier bits ModShift, ModLock, and ModControl are predefined. It does not make sense to assign ModControl to any keysym other than Control_L or Control_R, or to assign any modifier bits to the "control" keysyms other than ModControl. You can't turn a "control" key into a "meta" key (or vice versa) by simply assigning the key a different modifier bit. You must also make that key generate an appropriate keysym (Control_L, Meta_L, etc). OG.