"Michael Werliin, PDI" <mwe(a)post.dk&gt; writes: That's where *you* can help. Documentation is a very important project but not many people have expressed interest in it. You can ask a lot of question where you find the document is unclear or needs examples. Then feedback the response to XEmacs manual. A lot of people will benefit from your work! -- Yoshiki Hayashi

CCs trimmed. Daniel> Unless I am blind, the documentation for mule is very It is. However, to the extent Mule works, it is supposed to be pretty transparent. As far as I know, we have few Hebrew-speaking users. (Maybe none.) Any Hebrew support we have is based mostly on contributions incorporated by people who don't speak the language. Be warned! I hope you'll persevere with XEmacs, but in this particular case FSF Emacs may be a better option. (See below.) Daniel> sketchy. For example, I have not been able to find any Daniel> information on how to enable Hebrew input. If I execute Daniel> 'select-input-method' and choose "hebrew", there is no Daniel> effect. (This is true for other languages as well). See? You didn't need any documentation to find it. :-/ I don't understand the "no effect." I cannot find a Hebrew input method; that doesn't mean it doesn't exist, but we apparently don't distribute one. For me, typing C-u C-m C-\ hebrew --> *beep* [no match]. Don't you get that error? If not, something is seriously awry. Input methods are not named by languages in Emacs, they are bound to character sets. In the case of Hebrew, the map is 1--1, so the name of the input method would be "hebrew" (possibly with modifiers, use TAB to check for completions), but for Western European languages like German and Italian, the input method would be "latin-1-prefix" or "latin-1-postfix" (depending on whether you prefer to type accents before or after the Latin character it modifies), not "german" and "italian". If the latin-1 methods don't work, then you probably do not have the `leim' package installed. (LEIM = Library of Emacs Input Methods). Use Options | Manage Packages | List and Install in the menubar. You need this for the hebrew.el library to work, even if you don't want all the input methods. Sorry, the package system has a bit of a granularity problem! You also must install the mule-base package. The FSF distributes a trivial Hebrew input method with Quail as hebrew.el, so we should be able to use it. This just maps the standard QWERTY characters to Hebrew characters. I'll set it up, test it a little, and send it to you with installation instructions later today. If you already have FSF Emacs available, you may be able to just do M-x load-file /path/to/emacs-20.x/lisp/hebrew.elc (if you have the XEmacs leim package installed, otherwise the input manager `quail' will not be found; FSF's quail probably won't work). Under X with a properly configured XEmacs (--with-xim=xlib or --with-xim=motif) you should be able to use LANG=hebrew in your environment (or however you are supposed to say that on your platform) to get XIM input support using Shift-SPC to enter the Hebrew input mode. You may need to add an XMODIFIERS="＠im=YourInputEngineHere" in the environment, but "YourInputEngineHere" is extremely platform- specific, and unfortunately I know very little about Hebrew. It would be very helpful to us if (1) you could provide more information about your current configuration (XEmacs's Installation file would be wonderful, "M-x describe-installation" should bring it up in a buffer), (2) tell us what you expect from Hebrew support in general, (3) tell us whether you are using a Hebrew keyboard, or need to map from a US version, and (4) if you know what the ASCII->Hebrew map should be, I'd like to know to compare it with the hebrew.el version. Please be aware that BIDI support in XEmacs is presently non-existent. I've been aiming at doing something about this. Unfortunately I am not an expert, don't speak Hebrew or Arabic, and almost all of the experts have chosen to work with the mainline GNU Emacs. Their code is mostly distributed directly to the active participants in the bidi ML at gnu.org. I am not aware of a generally available repository for it offhand, although direct requests to the authors will get you their packages. As of last March, though, even the packages available for FSF Emacs were not very good, according to presentations by their authors at the Multilingualization Conference! To be fair, most of them considered their packages very usable; the main defects were in nested BIDI handling, and relatively long learning curves for users. I'll have to look again at BIDI handling in FSF Emacs, I haven't tried that since March. Daniel> Presumably fonts need to be added and enabled, but I find Daniel> no directions for this. This is a platform-specific problem, not really related to XEmacs. If your platform can do Hebrew at all, and the Hebrew fonts are installed in "the usual places", XEmacs should be able to find them without further intervention by the admin or the user. On an X platform (I assume you are from the "Lucid" in your X-Mailer header), you can get intlfonts.tar.gz (there may be a version number in there) from a GNU mirror. I think the current version is about 1.3, and that this is in the intlfonts directory under the GNU directory. Warning: there are a lot of fonts there, and the FSF doesn't break them out. If you want to save space by getting rid of the Asian fonts, you'll need to fix fonts.dir and fonts.alias too. You may need to do a `mkfontdir' by hand, and `xset fp rehash' for the X server to pick them up on the fly. These are probably not beautiful fonts. There is also a copy of those fonts on ftp.xemacs.org, but I believe it's out of date. If you need more help than that, write me offline, it's definitely not appropriate on XEmacs-beta. Daniel> Please excuse me if this message is too newbie-ish for Daniel> this list. Even an expert can't use a function that's just plain _missing_. Thanks for bringing this to our attention! -- 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."

Yoshiki> Pehaps you can copy hebrew.el from FSF Emacs leim Yoshiki> package, but XEmacs doesn't understand the escape Yoshiki> sequence it uses. I don't know whether it is XEmacs bug Yoshiki> (I don't have the list of registered escape sequences) Of course you do. It's in file-coding.c, thank you very much, Ben! Yoshiki> but if it is, it needs to be fixed. It's a non-standard Mule sequence. According to ISO-2022 (4ed, 1994-12-01) it is "reserved for standardization." Mule uses it to invoke a TYPE96 charset into G0. (Gag, choke.) The problem isn't that, though. XEmacs handles that fine (see etc/HELLO). The real problem is XEmacs still has the notion of directionality in the _definition_ of character sets, so it can't find the Hebrew charset without the direction escape sequence CSI 2 ]. Works fine (as far as this non-Hebrew-speaker can tell). Will submit patch and commit soon. -- 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."