1
0
mirror of https://git.savannah.gnu.org/git/emacs.git synced 2024-11-24 07:20:37 +00:00
emacs/etc/FAQ
1990-12-10 05:14:16 +00:00

2401 lines
87 KiB
Plaintext

This is a list of frequently asked questions (FAQ) about GNU Emacs with
answers. This list is designed to cut down on the noise in the
gnu.emacs.help newsgroup (which is also the help-gnu-emacs mailing list)
that results from people repeatedly asking many of the same questions.
This list is maintained by me, Joe Wells. My e-mail address is
jbw@bu.edu. If jbw@bu.edu doesn't work try jbw@bucsf.bu.edu.
Feel free to contribute text for the answers that I haven't written yet,
to suggest new questions, to suggest changes in the wording, to suggest
deletions, etc. Please include either "FAQ" or "frequently asked
questions" in the subject of messages you send to me to help keep me from
accidentally filing your message in the wrong mail folder.
Here is a short list of new questions and questions that are now answered:
New questions:
Why doesn't "C-c C-c" interrupt subprocesses under HP/UX?
How do I make Emacs send 8-bit characters to my terminal?
Why does ispell sometimes ignore the local dictionary?
Where can I get Emacs with NeWS support?
Questions with improved answers:
Has someone written an GNU Emacs Lisp package that does XXX?
How do I get Emacs running on VMS under DECwindows?
Why does Emacs crash from a bus error in XLookupString under OpenLook?
Why does Emacs hang with message "Unknown XMenu" with X11R4?
How do I use regexps (regular expressions) in Emacs?
How do I make Emacs "typeover" or "overwrite" when I type instead of
always inserting?
How do I get Emacs for my PC?
How can I highlight a region?
Where can I get GNU Emacs Lisp packages that don't come with Emacs?
Where can I get the latest VM, Supercite, GNUS, Calc, Calendar, or Patch?
Minor changes:
Where does the name "Emacs" come from?
In C mode, can I show just the lines that will be left after #ifdef
commands are handled by the compiler?
What does "M-x command" mean?
How do I get a printed copy of the GNU Emacs manual?
What do things like this mean: C-h, M-C-a, RET, etc.?
Why doesn't GNUS work anymore via NNTP?
What does Emacs do every time it starts up?
Enjoy,
--
Joe Wells <jbw@bu.edu>
-------------------------------------------------
GNU Emacs Frequently Asked Questions with Answers (part 1 of 2)
Sat Dec 8 01:30:53 1990
If you are viewing this file in a GNU Emacs Buffer, you can type "M-2 C-x
$" to get an overview of just the questions. Then, when you want to look
at the text of the answers, just type "C-x $".
To search for a question numbered XXX, type "M-C-s ^XXX:", followed by a
C-r if that doesn't work, then type ESC to end the search.
Notation Used in the Answers (READ THIS SECTION FIRST!)
You may skip this section if you are reasonably familiar with GNU Emacs.
Some of these are not actually frequently asked questions, but knowing
them is important for understanding the answers to the rest of the
questions.
1: What do things like this mean: C-h, M-C-a, RET, etc.?
C-a means while holding down the "Control" key press the "a" key. The
ASCII code this sends will generally be the value that would be sent by
pressing just "a" minus 96 or 64. Either way it will be a number from 0
to 31.
M-a means while holding down the "Meta" key press the "a" key. The
ASCII code this sends is the sum of the ASCII code that would be sent by
pressing just "a" and 128.
M-C-a means while holding down both the "Control" key and the "Meta" key
press the "a" key.
C-M-a is a synonym for M-C-a.
RET means press the "Return" key. RET is the same as C-m. This sends
ASCII code 13.
LFD means press the "Linefeed" key. LFD is also the same as C-j. This
sends ASCII code 10. Under Unix, ASCII code 10 is more often called
"Newline".
DEL means press the "Delete" key. DEL is the same as C-?. This sends
ASCII code 127.
ESC means press the "Escape" key. ESC is the same as C-[. This sends
ASCII code 27.
SPC means press the "Space" key. This send ASCII code 32.
TAB means press the "Tab" key. TAB is the same as C-i. This send ASCII
code 9.
To read more about this online, type "C-h i m emacs RET m characters
RET", and also "C-h i m emacs RET m keys RET".
2: What do you mean when you write things like this: type "ESC a"?
I will enclose key sequences that are longer than one key inside double
quotes. These notations refer to single keys (some with modifiers):
C-x, M-x, M-C-x
RET, LFD, DEL, ESC, SPC, TAB
I separate these from other keys within double quotes by spaces. Any
real spaces that I write inside double quotes can be ignored, only SPC
means press the space key. All other characters within double quotes
represent single keys (some shifted).
3: What if I don't have a Meta key?
Instead of typing M-a, you can type "ESC a" instead. In fact, Emacs
converts M-a internally into "ESC a" anyway.
4: What if I don't have an Escape key?
Type C-[ instead. This should send ASCII code 27 just like an Escape
key would.
5: What does "M-x command" mean?
"M-x command" means type M-x, then type the name of the command, then
type RET.
M-x is simply the default key sequence that invokes the command
"execute-extended-command". This command allows you to run any Emacs
command if you can remember the command's name. If you can't remember
the command's name, you can type TAB and SPC for completion, and ? for a
list of possibilities. An Emacs "command" is any "interactive" Emacs
function.
NOTE: Your system administrator may have bound other key sequences to
invoke execute-extended-command. A function key labeled "Do" is a good
candidate for this.
To run non-interactive Emacs functions, use M-ESC instead and type a
Lisp form.
6: What do things like this mean: etc/SERVICE, src/config.h,
lisp/default.el?
These are the names of files that are part of the GNU Emacs
distribution. The GNU Emacs distribution is so big that it is divided
into several subdirectories; the important subdirectories are named
"etc", "lisp", and "src".
If you use GNU Emacs, but don't know where it is kept on your system,
start Emacs, then type "C-h v exec-directory RET". The directory name
that is displayed by this will be the full pathname of the "etc"
directory of your installed GNU Emacs distribution.
Some of these files are available individually via FTP or e-mail, see
question 16.
7: What are FSF, LPF, OSF, GNU, RMS, FTP, BTW, and GPL?
FSF == Free Software Foundation
LPF == League for Programming Freedom
OSF == Open Software Foundation
GNU == GNU's Not Unix
RMS == Richard M. Stallman
FTP == File Transfer Protocol
BTW == By the way
GPL == GNU General Public Licence
NOTE: Avoid confusing the FSF, the LPF, and the OSF. The LPF only
opposes look-and-feel copyrights and software patents. The FSF aims to
make high quality free software available for everyone. The OSF is a
commercial organization which wants to provide an alternative,
standardized version of Unix not controlled by AT&T.
Sources of Information and Help
8: I'm just starting GNU Emacs; how do I do basic editing?
Type "C-h t" to invoke the self-paced tutorial. Typing just C-h is
how to enter the help system.
WARNING: your system administrator may have changed C-h to act like DEL.
You can use M-x help-for-help instead to invoke help. To discover what
key (if any) invokes help on your system, type "M-x where-is RET
help-for-help RET". This will print a comma-separated list of key
sequences in the echo area. Ignore the last character in each key
sequence listed. Each of the resulting key sequences invokes help.
NOTE: Emacs's help facility works best if help is invoked by a single
key. The variable help-char should hold the value of this character.
9: How do I find out how to do something in GNU Emacs?
There are several methods for finding out how to do things in Emacs.
You should become familiar with the online documentation for Emacs. The
complete text of the Emacs manual is available online in a hypertext
format via the "Info" manual reader. Type "C-h i" to invoke Info.
You can order a hardcopy of the manual from the FSF. See question 12.
You can list all of the commands whose names contain a certain word
(actually which match a regular expression) using the "command-apropos"
command. Type "C-h a" to invoke this command.
You can list all of the functions and variables whose names contain a
certain word using the "apropos" command. M-x apropos invokes this
command.
You may find that command-apropos and apropos are extremely slow on your
system. This will be fixed in Emacs 19. If you can't wait that long,
I can send you the "fast-apropos.el" file that contains the fix. This
file also contains a "super-apropos" command that will list all the
functions and variables whose documentation strings contain a certain
word.
There are many other commands in Emacs for getting help and information.
To get a list of these commands, type "C-h C-h C-h".
10: Where can I get GNU Emacs on the net (or by snail mail)?
Look in the files etc/DISTRIB and etc/FTP for information. If you don't
already have GNU Emacs, see question 16.
11: Where can I get help in installing GNU Emacs?
Look in the file etc/SERVICE for names of companies and individuals who
will sell you this type of service.
This file in the version 18.55 distribution is not completely up to
date. The imminent release of 18.56 will help that. An up-to-date
version of the SERVICE file is available on prep.ai.mit.edu. See
question 16 for how to retrieve this file.
12: How do I get a printed copy of the GNU Emacs manual?
You can order a printed copy of the GNU Emacs manual from the FSF for
$15. A set of 6 manuals is $60.
The full TeX source for the manual also comes with Emacs, if you're
daring enough to try to print out this 300 page manual yourself. (We're
talking about killing an entire tree here. This kind of printer usage
can make you very unpopular. Besides, if you order it from the FSF, the
price may be tax-deductible as a business expense.)
If you absolutely have to print your own copy, and you don't have TeX,
you can get a PostScript version from:
Anonymous FTP:
site: abcfd20.larc.nasa.gov
file: /pub/docs/emacs-18.55.PS.Z (size: 472K)
Carl Witty <cwitty@cs.stanford.edu> writes:
The Emacs manual is also available online in the Info system, which is
available by typing "C-h i". In this form, it has hypertext links and
is very easy to browse or search; many people prefer it to the printed
manual.
13: Has someone written an GNU Emacs Lisp package that does XXX?
Probably. A listing of Emacs Lisp packages, called the Lisp Code
Directory, is being maintained by Dave Brennan <brennan@dg-rtp.dg.com>.
You can search through this list to find if someone has written
something that fits your needs.
This list is file "LCD-datafile.Z" in the Emacs Lisp Archive. (See
question 14.) The files "lispdir.el.Z" and "lispdir.doc.Z"
in the archive contain information to help you use the list.
14: Where can I get GNU Emacs Lisp packages that don't come with Emacs?
First, check the Lisp Code Directory to find the name of the package you
are looking for. (See question 13). Then check local archives and
the Emacs Lisp Archive to find a copy of the relevant files. Then, if
you still haven't found it, you can send e-mail to the author asking for
a copy.
You can access the Emacs Lisp Archive via:
Anonymous FTP:
site: tut.cis.ohio-state.edu
directory: /pub/gnu/emacs/elisp-archive
file: README.Z
NOTE: Any files with names ending in ".Z" are compressed, and you should
use "binary" mode in FTP to retrive them. You should also use binary
mode whenever you retrieve any files with names ending in ".elc".
Submissions for the archive should be sent to Dave Sill <de5@ornl.gov>.
If you have ftp access, the submission can be deposited in the directory
/pub/gnu/emacs/elisp-archive/incoming on tut.cis.ohio-state.edu.
Whenever possible submissions should contain an LCD-datafile entry since
this helps reduce administrative overhead for the maintainers. Before
submitting anything, first read the file "guidelines.Z", which is
available in the archive.
15: Where can I get documentation on GNU Emacs Lisp?
Obtain the GNU Emacs Lisp Reference Manual for Emacs 18 under Unix. It
is available from the FSF for $50. The latest revision is edition 1.02
dated July 1990.
Within Emacs, you can type "C-h f" to get the documentation for a
function, "C-h v" for a variable.
Also, as a popular USENET saying goes, "Use the Force, Read the Source".
You can also get the Texinfo source for the manual, if you are daring
enough to try to print this 550 page manual out yourself. This is
available in the Emacs Lisp Archive and also via:
Anonymous FTP:
site: prep.ai.mit.edu
file: /pub/gnu/elisp.tar.Z
A set of pregenerated Info files comes with the Texinfo source for the
Emacs Lisp manual, so you don't have to format it yourself for online
use. (You can create the Info files from the Texinfo source.) See
question 21 for details on how to install these files online.
16: What informational files are available?
This isn't a frequently asked question, but it should be!
The following files are available in the "etc" directory of the GNU
Emacs distribution, and also the latest versions are available
individually via:
Anonymous FTP:
site: prep.ai.mit.edu
directory: /pub/gnu/etc
APPLE -- Special Report: Apple's New Look and Feel
DISTRIB -- GNU Emacs Availability Information,
including the popular "Free Software Foundation Order Form"
FTP -- How to get GNU Software by Internet FTP or by UUCP
GNU -- The GNU Manifesto
INTERVIEW -- Richard Stallman discusses his public-domain
UNIX-compatible software system
with BYTE editors
MACHINES -- Status of GNU Emacs on Various Machines and Systems
MAILINGLISTS -- GNU Project Electronic Mailing Lists
SERVICE -- GNU Service Directory
SUN-SUPPORT -- includes the popular "Using Emacstool with GNU Emacs"
These files are available in the "etc" directory of the GNU Emacs
distribution:
DIFF -- Differences between GNU Emacs and Twenex Emacs
CCADIFF -- Differences between GNU Emacs and CCA Emacs
GOSDIFF -- Differences between GNU Emacs and Gosling (Unipress??) Emacs
COPYING -- GNU Emacs General Public License
NEWS -- GNU Emacs News, a history of user-visible changes
In addition, all of the above files are available directly from the FSF
via e-mail. Of course, please try to get them from a local source
first.
These files are available from the FSF via e-mail:
* GNU's Bulletin, June, 1990 -- this file includes:
GNU'S Who
What Is the Free Software Foundation?
What Is Copyleft?
GNUs Flashes
Possible New Terms for GNU Libraries by Richard Stallman
"League for Programming Freedom" by Michael Bloom and Richard Stallman
Free Software Support
GNU Project Status Report
GNU Wish List
GNU Documentation
GNU Software Available Now
Contents of the Emacs Release Tape
Contents of the Pre-Release Tape
Contents of the X11 Tapes
VMS Emacs and Compiler Tapes
How to Get GNU Software
Freemacs, an Extensible Editor for MS-DOS
Thank GNUs
FSF Order Form
* GNU Task List
* GNU Coding Standards
* Legal issues about contributing code to GNU
* GNU Project Status Report
17: What is the current address of the FSF?
Snail mail address:
Free Software Foundation
675 Massachusetts Avenue
Cambridge, MA 02139, USA
Phone number:
(617) 876-3296
E-mail addresses:
gnu@prep.ai.mit.edu {probably read by Len Tower}
gnulists@prep
18: What is the current address of the LPF?
Snail mail address:
League for Programming Freedom
1 Kendall Square, Number 143
Post Office Box 9171
Cambridge, MA 02139, USA
Phone number:
(617) 243-4061 { or 243-4091, I'm not sure ... }
E-mail address:
league@prep.ai.mit.edu
19: Where can I get the up-to-date GNU stuff?
The most up-to-date official GNU stuff is normally kept on
prep.ai.mit.edu and is available for anonymous FTP. See the files
etc/DISTRIB and etc/FTP for more information. (To get copies of these
files, see question 16.)
20: Where can I get the latest VM, Supercite, GNUS, Calc, Calendar, or Patch?
1. VM:
Author: Kyle Jones <kyle@xanth.cs.odu.edu>
Mailing lists: info-vm-request@uunet.uu.net
info-vm@uunet.uu.net
bug-vm-request@uunet.uu.net
bug-vm@uunet.uu.net
Anonymous FTP:
Last released version: 4.41
site: tut.cis.ohio-state.edu
file: /pub/gnu/vm/vm-4.41.tar.Z
Beta test version: 5.18
site: abcfd20.larc.nasa.gov
directory: /pub/vm
2. Supercite:
Author: Barry Warsaw <warsaw@warsaw.nlm.nih.gov>
Latest version: 2.1
Mailing lists: supercite@warsaw.nlm.nih.gov
supercite-request@warsaw.nlm.nih.gov
Anonymous FTP:
site: durer.cme.nist.gov [129.6.32.4]
files:/pub/gnu/elisp/supercite-2.1.shar.0# (where # is replaced by 1 or 2)
NOTE: Superyank is an old version of Supercite.
3. GNUS:
Author: Masanobu Umeda
Latest version: 3.13
Anonymous FTP:
site: tut.cis.ohio-state.edu
files: /pub/gnu/gnus/3.13.##-of-11.Z (where ## is replaced by 00 -- 11)
4. Calc:
Author: Dave Gillespie <daveg@csvax.cs.caltech.edu>
Latest version: 1.06
Anonymous FTP:
site: csvax.cs.caltech.edu [131.215.131.131]
or any comp.sources.misc archive
5. Calendar/Diary
Author: Ed Reingold <reingold@cs.uiuc.edu>
Latest version: 4
E-mail:
Subject: send-emacs-cal
body: your best internet e-mail address
6. Patch
Author: Larry Wall <lwall@jpl-devvax.jpl.nasa.gov>
Latest version: 2.0 patchlevel 12
Anonymous FTP:
site: any comp.sources.unix archive
or, for Europeans:
site: archive.cs.ruu.nl [131.211.80.5]
file: /pub/UNIX/patch-2.0.tar.Z
21: How do I install a piece of Texinfo documentation?
First create Info files from the Texinfo files with M-x
texinfo-format-buffer. There is also a program called "makeinfo" that
provides better error checking and runs faster. Currently, it comes
with the Emacs Lisp manual (which comes on the GNU Emacs tape).
For information about the Texinfo format, read the Texinfo manual which
comes with Emacs. This manual also comes installed in Info format, so
you can read it online.
Neither texinfo-format-buffer nor the makeinfo program install the
resulting Info files in Emacs's Info tree. To install Info files:
1. Move the files to the "info" directory in the installed Emacs
distribution. See question 6 if you don't know where that
is.
2. Edit the file info/dir in the installed Emacs distribution, and add a
line for the top level node in the Info package that you are
installing.
22: What is the real legal meaning of the GNU copyleft?
RMS writes:
The legal meaning of the GNU copyleft is less important than the spirit,
which is that Emacs is a free software project and that work pertaining
to Emacs should also be free software. "Free" means that all users have
the freedom to study, share, change and improve Emacs. To make sure
everyone has this freedom, pass along source code when you distribute
any version of Emacs or a related program, and give the recipients the
same freedom that you enjoyed.
If you still want to find out about the legal meaning of the copyleft,
please ask yourself if this means you are not paying attention to the
spirit.
23: What are appropriate messages for gnu.emacs.help/gnu.emacs.bug/comp.emacs
etc.?
The file etc/MAILINGLISTS discusses the purpose of each GNU
mailing-list. (See question 16 on how to get a copy.) For
those which are gatewayed with newsgroups, it lists both the newsgroup
name and the mailing list address.
comp.emacs is for discussion of Emacs programs in general. This
includes GNU Emacs along with various other implementations like JOVE,
MicroEmacs, Freemacs, MG, Unipress, CCA, etc.
Many people post GNU Emacs questions to comp.emacs because they don't
receive any of the gnu.* newsgroups. Arguments have been made both for
and against posting GNU-Emacs-specific material to comp.emacs. You have
to decide for yourself.
24: Should I send questions directly to people who post a lot on the net,
instead of posting myself?
I don't know about the others, but please don't send such messages to me
just because you don't want to disturb a lot of people. Send messages
to individuals that concern things that individual has written.
GNU Emacs, all its Variants, and other Editors
25: Where does the name "Emacs" come from?
EMACS originally was an acronym for Editor MACroS. The first Emacs was
a set of macros written by Richard Stallman for the editor TECO (Tape
Editor and COrrector {Text Editor and COrrector?}) on a PDP-10.
(Amusing fact: many people have told me that TECO code looks a lot like
line noise.)
26: What is the latest version of GNU Emacs?
GNU Emacs 18.55. GNU Emacs 18.56 is coming out with some bug fixes, but
no additional functionality.
27: When will GNU Emacs 19 be available?
Good question, I don't know. For that matter, neither do the
developers. It will undoubtedly be available sometime in the 1990s.
28: What will be different about GNU Emacs 19?
From the latest "GNU's Bulletin":
Version 19 approaches release with a host of new features: before and
after change hooks; X selection processing (including CLIPBOARD
selections); scrollbars; support for European character sets; floating
point numbers; per-buffer mouse commands; interfacing with the X
resource manager; mouse-tracking; Lisp-level binding of function keys;
and multiple X windows (`screens' to Emacs).
Thanks go to Alan Carroll and the people who worked on Epoch for
generating initial feedback to a multi-windowed Emacs. Emacs 19
supports two styles of multiple windows, one with a separate screen for
the minibuffer, and another with a minibuffer attached to each screen.
More features of Version 19 are buffer allocation, which uses a new
mechanism capable of returning storage to the system when a buffer is
killed, and a new input system---all input now arrives in the form of
Lisp objects.
Other features being considered for Version 19 include:
* Associating property lists with regions of text in a buffer.
* Multiple font, color, and pixmaps defined by those properties.
* Different visibility conditions for the regions, and for the various
windows showing one buffer.
* Incremental syntax analysis for various programming languages.
* Hooks to be run if point or mouse moves outside a certain range.
* Source-level debugging for Emacs Lisp.
* Incrementally saving undo history in a file, so that recover-file also
reinstalls buffer's undo history.
* Static menu bars, and better pop-up menus.
* A more sophisticated emacsclient/server model, which would provide
network transparent Emacs widget functionality.
Also, Emacs 19 will have a Emacs Lisp source code debugger.
29: Is there an Emacs that has better mouse and X window support?
Emacs 18 has some limited X Window System support, but there are
problems. Emacs 19 will have amazing mouse and window support. Right
now, there is a modified version of Emacs 18.55 called "Epoch" which has
greatly improved mouse and window support. The latest version of Epoch
is available via:
Anonymous FTP:
site: cs.uiuc.edu
directory: pub/epoch-files/epoch
file: epoch-3.2b (b stands for beta)
NOTE: Epoch only works with the X Window System; it does not work on
ordinary terminals.
30: How do I get Emacs for my PC?
GNU Emacs won't run on a PC directly under MS-DOS. There have been
reports of people running GNU Emacs under a special program under MS-DOS
on 286 or 386 machines, but I don't know the details.
You can try a variety of similar programs such as MG, MicroEmacs,
Freemacs, etc.
Russ Nelson <nelson@sun.soe.clarkson.edu>, the author, describes
Freemacs:
* Freemacs is free, and it was designed from the start to be
programmable.
* Freemacs is the only IBM-PC editor that tries to be like GNU Emacs.
* Freemacs can only edit files less than 64K in length.
* Freemacs doesn't have undo.
Carl Witty <cwitty@cs.stanford.edu> writes about Freemacs:
Better is Freemacs, which follows the tradition of ITS and GNU Emacs
by having an full, turing-complete extension language which is
incompatible with everything else. In fact, it's even closer to ITS
Emacs than GNU Emacs is, because Mint (Freemacs' extension language)
is absolutely illegible without weeks of study, much like TECO.
You can get Freemacs several ways:
1. Anonymous FTP:
site: simtel20.army.mil, directory: PD:<MSDOS.FREEMACS>
site: grape.ecs.clarkson.edu [128.153.13.196], directory: /e/freemacs
2. E-mail:
address: archive-server@sun.soe.clarkson.edu, or
archive-server%sun.soe@omnigate (BITNET), or
{smart-host}!sun.soe.clarkson.edu!archive-server (UUCP)
body: help
3. Snail mail:
address: Russell Nelson, 11 Grant St., Potsdam, NY 13676
Send $15 copying fee, and specify preferred floppy disk format:
5.25", 1.2 Meg, or 5.25", 360K, or 3.50", 720K
I don't know much about MicroEmacs. All I know is that the latest
version is 3.10 and it is available via:
Anonymous FTP:
site: durer.cme.nist.gov
Another Emacs for small machines is Jove (Jonathan's Own Version of
Emacs). The latest version is 4.9. It is available via:
Anonymous FTP:
sites: hipl.psych.nyu.edu, cs.toronto.edu, cayuga.cs.rochester.edu
Yet another Emacs is "mg", which used to stand for MicroGNUEmacs, but
now just stands for "mg". It is available via:
Anonymous FTP:
site: snow.white.toronto.edu
31: Where can I get Emacs for my Atari ST?
The latest version for the ST is available via:
Anonymous FTP:
site: cs.uni-sb.de [134.96.7.254]
directory: /pub/atari/emacs
32: Where can I get Emacs with NeWS support?
Chris Maio's NeWS support package for GNU Emacs is available via:
Anonymous FTP:
site: columbia.edu
file: /pub/ps-emacs.tar.Z
site: tut.cis.ohio-state.edu
file: /pub/gnu/emacs/ps-emacs.tar.Z
Binding Keys to Commands
33: Why does my key binding fail?
Most likely, it failed because "ESC [" was already defined. Evaluate
this form first:
(define-key esc-map "[" nil)
34: Why doesn't this [terminal or window-system setup] code work in my
.emacs file, but it works just fine after Emacs starts up?
This is because you're trying to do something in your .emacs file that
needs to be postponed until after the terminal/window-system setup code
is loaded. This is a result of the order in which things are done
during the startup of Emacs.
In order to postpone the execution of Emacs Lisp code until after the
terminal/window-system setup, set the value of the variable
term-setup-hook or window-setup-hook to be a function which does what
you want.
See question 72 for a complete explanation of what Emacs
does every time it is started.
Here is a simple example of how to set term-setup-hook:
(setq term-setup-hook
(function
(lambda ()
(if (string-match "^vt220" (or (getenv "TERM") ""))
;; Make vt220's "Do" key behave like M-x:
(define-key CSI-map "29~" 'execute-extended-command)))))
35: How do I use function keys under X Windows?
If compiled on a Sun, Emacs recognizes the function keys that are
normally on a Sun keyboard, and you can bind them the same way you
normally bind Sun keys. Each function key will generate a key sequence
that looks like "ESC [ ### z", where ### is replaced by a number.
If not compiled on a Sun, the function keys will appear to Emacs in a
way remarkably similar to the keys of a VT220 terminal. Each function
key will generate a key sequence that looks like "ESC [ ### ~".
For the complete list of the numbers which are generated by the function
keys, look in the file src/x11term.c.
However, this may be inadequate for you if you have function keys Emacs
doesn't know about. Johan Vromans <jv@mh.nl> explains:
There are a number of tricks that can be helpful. The most elegant
solution, however, is to use the function "x-rebind-key". This function
is commented out in the source for good reasons --- it's buggy.
It is rather easy to replace this function with the function
epoch:rebind-key from the Epoch distribution.
After implementing this, all keyboard keys can be configured to send
user definable sequences, e.g.
(x-rebind-key "KP_F1" 0 "\033OP")
This will have the keypad key PF1 send the sequence \eOP, just like an
ordinary VTxxx terminal.
Another method is to handle the keys in the C source. Although this may
be slightly more efficient, it is much less flexible.
Finally, some people have established a convention to have the special
keys send a sequence of characters that indicate the key and its
modifiers instead of a plain escape sequence. This is also a good
method. The advantage of sending the plain escape sequences is that you
require only one setup for Emacs running under X and on normal
terminals, e.g.
(x-rebind-key "KP_F1" 0 "\033OP")
will have key PF1 send \eOP, and, probably in another setup file:
(global-set-key "\eOP" 'specific-function)
to bind PF1 to a function. The last statement applies for normal VTxxx
terminal also.
36: How do I tell what characters my function keys emit?
Use this function by Randal L. Schwartz <merlyn@iwarp.intel.com>:
(defun see-chars ()
"Displays characters typed, terminated by a 3-second timeout."
(interactive)
(let ((chars "")
(inhibit-quit t))
(message "Enter characters, terminated by 3-second timeout.")
(while (not (sit-for 3))
(setq chars (concat chars (list (read-char)))
quit-flag nil)) ; quit-flag maybe set by C-g
(message "Characters entered: %s" (key-description chars))))
Problems with Key Bindings and Input
37: Why does Emacs spontaneously go into "I-search:" mode?
Your terminal (or something between your terminal and the computer) is
sending C-s and C-q for flow control, and Emacs is receiving these
characters and interpreting them as commands. (The C-s character
normally invokes the isearch-forward command.)
For a more detailed discussion, read the file PROBLEMS in the Emacs
distribution.
38: What do I do if my terminal is sending C-s and C-q for flow control and
I can't disable it?
Use this piece of Emacs Lisp:
(set-input-mode nil t)
39: How do I make Emacs use C-s and C-q for flow control instead of for
commands?
Same answer as previous question.
40: How do I use commands bound to C-s and C-q (or any key) if these keys
are filtered out?
I suggest swapping C-s with C-\ and C-q with C-^:
(swap-keys ?\C-s ?\C-\\)
(swap-keys ?\C-q ?\C-^)
See question 41 for the implementation of swap-keys.
41: How do I "swap" two keys?
When Emacs receives a character, you can make Emacs behave as though it
received another character by setting the value of
keyboard-translate-table. The following Emacs Lisp will do this for
you, allowing you to "swap" keys.
WARNING: the value of C-g (7) is hard coded in several places both
before and after the translation specified by keyboard-translate-table
is carried out. Thus, swapping C-g with another key may cause problems;
rebinding C-g may also cause problems. (Well, I've never done it, so I
don't know for sure.)
(defun swap-keys (key1 key2)
"Swap keys KEY1 and KEY2 using map-key."
(map-key key1 key2 t)
(map-key key2 key1))
(defvar map-keys-alist nil
"Association list of key mappings currently in effect.
If (FROM . TO) is an element, that means key FROM is currently mapped to TO.")
(defun map-key (from to &optional no-update)
"Make key FROM behave as though key TO was typed instead.
If optional argument NO-UPDATE is non-nil, the key-mapping does not take
effect until a subsequent map-key or unmap-key."
(let ((alist-entry (assq from map-keys-alist)))
(if alist-entry
(setcdr alist-entry to)
(setq map-keys-alist (cons (cons from to) map-keys-alist))))
(or no-update (map-keys-update)))
(defun unmap-key (key)
"Undo any mapping of key KEY."
(setq map-keys-alist (delq (assq key map-keys-alist) map-keys-alist))
(map-keys-update))
;; Makes keyboard-translate-table reflect the key mappings in
;; map-keys-alist.
(defun map-keys-update ()
(if (null map-keys-alist)
;; Emacs runs fasted if keyboard-translate-table is nil
(setq keyboard-translate-table nil)
(let ((max-key-mapped
;; Find the mapped key with largest value
(apply 'max
(mapcar
(function
(lambda (x)
(car x)))
map-keys-alist)))
(i 0))
;; keyboard-translate-table doesn't have to be any longer than
;; necessary. This speeds up Emacs.
(setq keyboard-translate-table (make-string (1+ max-key-mapped) 0))
(while (<= i max-key-mapped)
(aset keyboard-translate-table i (or (cdr (assq i map-keys-alist)) i))
(setq i (1+ i))))))
42: Why does the "Backspace" key invoke help?
The Backspace key (on every keyboard I've used) sends ASCII code 8. C-h
sends the same code. In Emacs by default C-h invokes "help-command".
The easiest solution to this problem is to use C-h (and Backspace) for
help and DEL (the Delete key) for deleting the previous character.
For some people this solution may be problematic:
1. They normally use Backspace outside of Emacs for deleting the
previous character typed. This can be solved by making DEL be the
command for deleting the previous character outside of Emacs. This
command will do this on many Unix systems:
stty erase ^?
2. The person may prefer using the Backspace key for deleting the
previous character because it is more conveniently located on their
keyboard or because they don't even have a separate Delete key. In
this case, the best solution is to swap C-h and DEL:
(swap-keys ?\C-h ?\C-?)
See question 41 for the implementation of swap-keys.
43: How do I type DEL on PC terminal emulators?
Someone whose name I forgot wrote:
Most PCs have deficient keyboards that don't have both Backspace and
Delete keys. Whether C-h (backspace) or DEL is generated by the
"Backspace" key varies from one terminal emulator to another. If you're
lucky, you can reconfigure the keyboard so that it generates DEL. If
not, you will have to hunt to figure out what keystroke will do it ---
possibilities include various shifted and controlled versions of
"Backspace", the "Del" key on the numeric keypad (which might depend on
"Shift" or "NumLock"), or perhaps C-? (Control-?).
If this is too hard, you may want to swap the delete key with some other
key. See question 42.
Building/Installing/Porting Emacs and Machine/OS-Specific Bugs:
44: Why do I get an "f68881_used undefined" error, when I build Emacs on my
Sun 3?
Barry A. Warsaw <warsaw@cme.nist.gov> writes:
Some of the code that is being linked on the "ld" line of emacs'
build command has been compiled with the -f68881 option. Most
common reason is that you're linking with X libraries which were
built with -f68881 option set. You need to either remove all
dependencies to the 68881 (may mean a recompile of the X libraries
with -fswitch or -fsoft option), or you need to link emacs with
the 68881 startup file /usr/lib/Mcrt1.o. Make this change to
src/ymakefile:
change: #define START_FILES crt0.o
to: #define START_FILES crt0.o /usr/lib/Mcrt1.o
The order of these start files is critical.
45: Why does Emacs crash under SunOS 4.1?
Under SunOS 4.1, Emacs dies with this message:
Fatal error (6).Abort
There is a bug in the "localtime" routine supplied with SunOS 4.1.
A private function called by tzsetwall zeroes the byte just past an
eight byte region it mallocs. This corrupts GNU malloc's memory pool.
Put "#define SYSTEM_MALLOC" in src/config.h.
Or, for the purists:
Put "#define SUNOS_LOCALTIME_BUG" in src/config.h and apply the following
patch to src/malloc.c.
(Patch by Eirik Fuller <eirik@elf.tn.cornell.edu>.)
>*** malloc.c~ Fri Apr 7 22:12:10 1989
>--- malloc.c Tue May 8 07:07:22 1990
>***************
>*** 476,482 ****
>--- 476,487 ----
> multiple of 8, then figure out which nestf[] area to use.
> Both the beginning of the header and the beginning of the
> block should be on an eight byte boundary. */
>+ #ifdef SUNOS_LOCALTIME_BUG
>+ /* SunOS 4.1 localtime scribbles on the ninth byte. */
>+ nbytes = (n + ((sizeof *p + 15) & ~15) + EXTRA + 15) & ~15;
>+ #else
> nbytes = (n + ((sizeof *p + 7) & ~7) + EXTRA + 7) & ~7;
>+ #endif
> {
> register unsigned int shiftr = (nbytes - 1) >> 2;
>
46: Why does Emacs occasionally cause segmentation faults and machine
panics on my DECstation 3100?
Bruce Cole <cole@dip.cs.wisc.edu> writes:
> We're running Emacs 18.54 on an Ultrix Risc DECsystem 5400. Three
> times we've had the machine hang with the following message:
>
> panic: tblmod on invalid pte
>
> Ultrix support tells us this is caused by Emacs. Has anyone
> experienced this? DEC says it only happens on RISC boxes.
This is due to a MIPS specific Ultrix kernel bug. I sent DEC a
description of the bug with a bug fix. The Kernel bug manifests itself
with Emacs since Emacs uses a non-standard data start address on Ultrix
MIPS machines.
I haven't often seen Emacs cause MIPS machines to panic. Usually you
just see one of the following errors when you try to start up Emacs:
segmentation fault (core dumped)
emacs: Bad address
Out of memory
data size rlimit exceeded, pid 6523, process tcsh (for example)
Until DEC fixes their kernel, you can avoid the bug by changing the data
start address used by Emacs.
> I have built two newer versions of GNU Emacs and it still happens. I
> should also mention that these faults happen quite rarely to some
> users (I may see it once a month), while other users see it a few
> times a week or even daily!
The problem only occurs when a MIPS machine is doing a lot of paging.
Users who don't cause their workstation to page will not see this
problem.
Patch to GNU Emacs 18.55:
>*** m-pmax.h Thu Jun 8 11:53:55 1989
>--- m-pmax.h.new Mon Jul 9 10:21:21 1990
>***************
>*** 1,3 ****
>--- 1,7 ----
> #include "m-mips.h"
> #undef LIBS_MACHINE
> #undef BIG_ENDIAN
>+ #undef LD_SWITCH_MACHINE
>+ #undef DATA_START
>+ #define DATA_START 0x10000000
>+ #define DATA_SEG_BITS 0x10000000
47: How do I get Emacs running on VMS under DECwindows?
Hal R. Brand <BRAND@addvax.llnl.gov> is said to have a VMS save set with
a ready-to-run VMS version of Emacs for X Windows. It is available via:
Anonymous FTP:
site: addvax.llnl.gov
Johan Vromans <jv@mh.nl> writes:
Getting Emacs to run on VMS with DECwindows requires a number of changes
to the sources. Fortunately this has been done already. Joshua Marantz
<josh@viewlogic.com> did most of the work for Emacs 18.52, and the mods
were ported to 18.55 by Johan Vromans <jv@mh.nl>. Also included is the
handling of DEC's LK201 keyboard. You need to apply the changes to a
fresh Emacs 18.55 distribution on a Unix system, and then you can copy
the sources to VMS to perform the compile/link/build.
The set of changes have been posted a number of times three times the
last 12 months, so they should be widely available.
48: Why won't src/alloc.c compile on a SPARCstation?
{Could someone send me a better description of the symptoms ... Does it
really not compile, or does it refuse to link, or does it die when it's
run?}
An object in src/alloc.c is declared as an array of char. However, on a
SPARCstation, this object must be aligned on a longword boundary.
gcc will align this object not on a longword boundary because it thinks
this is ok because the object is an array of char.
Here is a solution:
(Patch from the Epoch source code.)
>*** dist-18.55.dist/src/alloc.c Mon Feb 13 05:23:35 1989
>--- dist-18.55/src/alloc.c Mon Nov 27 23:30:13 1989
>***************
>*** 665,673 ****
> #define NSTATICS 200
>
>- char staticvec1[NSTATICS * sizeof (Lisp_Object *)] = {0};
>-
> int staticidx = 0;
>
> #define staticvec ((Lisp_Object **) staticvec1)
>
> /* Put an entry in staticvec, pointing at the variable whose address is given */
>--- 665,676 ----
> #define NSTATICS 200
>
> int staticidx = 0;
>
>+ #ifdef __GNUC__
>+ Lisp_Object *staticvec[NSTATICS] = {0};
>+ #else
>+ char staticvec1[NSTATICS * sizeof (Lisp_Object *)] = {0};
> #define staticvec ((Lisp_Object **) staticvec1)
>+ #endif
>
> /* Put an entry in staticvec, pointing at the variable whose address is given */
{Does anyone understand why one might want to declare staticvec as an
array of char instead of an array of LispObject?}
49: Why do shells started from Emacs suspend themselves under SunOS 4.1?
In SunOS 4.1, when subprocesses are started by an Emacs process started
by "rsh", the process stops with a message like this:
Process shell stopped (tty input)
SunOS 4.1 uses the POSIX standard's specification for job control.
Basically, a process must be a session leader to acquire a new
controlling terminal.
A quick fix is to run Emacs as "/usr/etc/setsid emacs" instead.
To fix it in Emacs, use this patch:
(Patch by Eirik Fuller <eirik@elf.tn.cornell.edu>.)
>*** /tmp/,RCSt1a12635 Fri Jul 6 11:29:10 1990
>--- process.c Sat Jun 16 14:04:55 1990
>***************
>*** 1031,1036 ****
>--- 1031,1039 ----
> int j = open ("/dev/tty", O_RDWR, 0);
> ioctl (j, TIOCNOTTY, 0);
> close (j);
>+ #ifndef USG
>+ setpgrp (0, 0);
>+ #endif
> }
> #endif /* TIOCNOTTY */
>
50: Why does Emacs crash from a bus error in XLookupString under OpenLook?
The Sun OpenLook libraries use the XComposeStatus parameter on
XLookupString for composing key sequences. I'm told this is actually
supposed to be an optional parameter.
(Patch by David J. MacKenzie <djm@eng.umd.edu>.)
>*** dist-18.55.dist/src/x11term.c Wed Aug 2 21:56:42 1989
>--- dist-18.55/src/x11term.c Fri Jul 13 19:22:32 1990
>***************
>*** 1213,1216 ****
>--- 1213,1218 ----
> #endif /* not sun */
>
>+ static XComposeStatus status;
>+
> internal_socket_read(bufp, numchars)
> register unsigned char *bufp;
>***************
>*** 1223,1227 ****
> BLOCK_INPUT_DECLARE ();
> XEvent event;
>- XComposeStatus status;
> KeySym keysym;
>
>--- 1225,1228 ----
>***************
>*** 1330,1334 ****
> nbytes = XLookupString (&event,
> mapping_buf, 20, &keysym,
>! 0);
> /* Someday this will be unnecessary as we will
> be able to use XRebindKeysym so XLookupString
>--- 1331,1335 ----
> nbytes = XLookupString (&event,
> mapping_buf, 20, &keysym,
>! &status);
> /* Someday this will be unnecessary as we will
> be able to use XRebindKeysym so XLookupString
51: Why doesn't "C-c C-c" interrupt subprocesses under HP/UX?
Paul Stravers <stravers@dutentb.tudelft.nl> writes:
In HP/UX we can (must) combine the two distinct calls
1. ioctl (fildes, TIOCGPGRP, &gid);
2. kill (-gid, signo);
into the single call
1+2. ioctl (fildes, TIOCSIGSEND, signo);
This is apparently the case under at least HP/UX versions 6.5 and 7.0.
Here is a set of patches courtesy of Piet van Oostrum <piet@cs.ruu.nl>
that fixes these problems:
1. Signals to child processes work now (like the C-c C-c mentioned above)
2. When running under X11, C-g now correctly interrupts long-running commands
3. When running under X11, the window is redislayed even if emacs is doing
some work (e.g. if your window becomes visible after being behind another
one.
4. A nasty bug in alloca is solved. This bug is very infrequent unless you
apply the other patches.
{I don't know for sure whether all of these patches are applicable only
under HP/UX. Perhaps someone can tell me.}
>*** alloca.s.~1~ Fri Sep 15 16:35:39 1989
>--- alloca.s Tue May 1 17:18:08 1990
>***************
>*** 92,103 ****
> sub.l &MAXREG*4,%d1 # space for saving registers
> mov.l %d1,%sp # save new value of sp
> tst.b PROBE(%sp) # create pages (sigh)
> move.w &MAXREG-1,%d0
> copy_regs_loop: /* save caller's saved registers */
>! mov.l (%a1)+,(%sp)+
> dbra %d0,copy_regs_loop
>! mov.l %sp,%d0 # return value
>! mov.l %d1,%sp
> add.l &-4,%sp # adjust tos
> jmp (%a0) # rts
> #endif /* new hp assembler */
>--- 92,105 ----
> sub.l &MAXREG*4,%d1 # space for saving registers
> mov.l %d1,%sp # save new value of sp
> tst.b PROBE(%sp) # create pages (sigh)
>+ mov.l %a2,%d1 # save reg a2
>+ mov.l %sp,%a2
> move.w &MAXREG-1,%d0
> copy_regs_loop: /* save caller's saved registers */
>! mov.l (%a1)+,(%a2)+
> dbra %d0,copy_regs_loop
>! mov.l %a2,%d0 # return value
>! mov.l %d1,%a2 # restore a2
> add.l &-4,%sp # adjust tos
> jmp (%a0) # rts
> #endif /* new hp assembler */
>*** keyboard.c.~1~ Wed Apr 4 14:53:49 1990
>--- keyboard.c Thu Apr 26 14:46:27 1990
>***************
>*** 719,724 ****
>--- 719,773 ----
> errno = old_errno;
> }
>
>+ #ifdef HPUX
>+ /* tells if an alarm is outstanding for emptying the socket */
>+ int socket_alarm;
>+
>+ static void
>+ empty_socket_interrupt()
>+ {
>+ int nread;
>+ char buf[256 * BUFFER_SIZE_FACTOR];
>+ register int i;
>+ int old_errno = errno;
>+
>+ if (waiting_for_input) return;
>+
>+ fcntl (fileno (stdin), F_SETFL, O_NDELAY);
>+
>+ /* Read what is waiting. */
>+
>+ nread = (*read_socket_hook) (0, buf, sizeof buf);
>+
>+ #ifdef EBADSLT
>+ if (nread == -1 && (errno == EAGAIN || errno == EBADSLT))
>+ #else
>+ if (nread == -1 && errno == EAGAIN)
>+ #endif
>+ nread = 0;
>+
>+ fcntl (fileno (stdin), F_SETFL, 0);
>+
>+ /* Scan the chars for C-g and store them in kbd_buffer. */
>+ kbd_ptr = kbd_buffer;
>+ for (i = 0; i < nread; i++)
>+ {
>+ kbd_buffer_store_char (buf[i]);
>+ /* Don't look at input that follows a C-g too closely.
>+ This reduces lossage due to autorepeat on C-g. */
>+ if (buf[i] == Ctl ('G'))
>+ break;
>+ }
>+
>+ /* reestablish interrupt */
>+ signal (SIGALRM, empty_socket_interrupt);
>+ alarm (1);
>+
>+ errno = old_errno;
>+ }
>+
>+ #endif /* HPUX */
>+
> /* read a character from the keyboard; call the redisplay if needed */
> /* commandflag 0 means do not do auto-saving, but do do redisplay.
> -1 means do not do redisplay, but do do autosaving.
>***************
>*** 733,738 ****
>--- 782,794 ----
> Lisp_Object tem;
> extern request_echo ();
>
>+ #ifdef HPUX
>+ if (socket_alarm) {
>+ alarm (0);
>+ socket_alarm = 0;
>+ }
>+ #endif
>+
> if ((c = unread_command_char) >= 0)
> {
> unread_command_char = -1;
>***************
>*** 869,874 ****
>--- 925,938 ----
> }
> }
>
>+ #ifdef HPUX
>+ if (read_socket_hook) {
>+ socket_alarm = 1;
>+ signal (SIGALRM, empty_socket_interrupt);
>+ alarm (1);
>+ }
>+ #endif
>+
> return c;
> }
>
>***************
>*** 1094,1119 ****
>
> #else /* no FIONREAD */
> #ifdef USG
> if (kbd_count)
> abort ();
>
> fcntl (fileno (stdin), F_SETFL, O_NDELAY);
>! kbd_ptr = kbd_buffer;
> if (read_socket_hook)
>! {
>! kbd_count = (*read_socket_hook) (0, kbd_buffer, sizeof kbd_buffer);
>! }
> else
>! {
>! kbd_count = read (fileno (stdin), kbd_buffer, sizeof kbd_buffer);
>! }
> #ifdef EBADSLT
>! if (kbd_count == -1 && (errno == EAGAIN || errno == EBADSLT))
> #else
>! if (kbd_count == -1 && errno == EAGAIN)
> #endif
>! kbd_count = 0;
> fcntl (fileno (stdin), F_SETFL, 0);
> #else /* not USG */
> you lose
> #endif /* not USG */
>--- 1158,1209 ----
>
> #else /* no FIONREAD */
> #ifdef USG
>+ char buf[256 * BUFFER_SIZE_FACTOR];
>+ register int i;
>+ #ifdef HPUX
>+ int mask;
>+ #endif
>+
> if (kbd_count)
> abort ();
>
>+ #ifdef HPUX
>+ if (socket_alarm)
>+ mask = sigblock (sigmask (SIGALRM));
>+ #endif
>+
> fcntl (fileno (stdin), F_SETFL, O_NDELAY);
>!
>! /* Read what is waiting. */
> if (read_socket_hook)
>! nread = (*read_socket_hook) (0, buf, sizeof buf);
> else
>! nread = read (fileno(stdin), buf, sizeof buf);
> #ifdef EBADSLT
>! if (nread == -1 && (errno == EAGAIN || errno == EBADSLT))
> #else
>! if (nread == -1 && errno == EAGAIN)
> #endif
>! nread = 0;
>!
> fcntl (fileno (stdin), F_SETFL, 0);
>+
>+ /* Scan the chars for C-g and store them in kbd_buffer. */
>+ kbd_ptr = kbd_buffer;
>+ for (i = 0; i < nread; i++)
>+ {
>+ kbd_buffer_store_char (buf[i]);
>+ /* Don't look at input that follows a C-g too closely.
>+ This reduces lossage due to autorepeat on C-g. */
>+ if (buf[i] == Ctl ('G'))
>+ break;
>+ }
>+
>+ #ifdef HPUX
>+ if (socket_alarm)
>+ sigsetmask(mask);
>+ #endif
>+
> #else /* not USG */
> you lose
> #endif /* not USG */
>*** x11term.h~ Fri Sep 15 15:20:26 1989
>--- x11term.h Thu Apr 26 14:03:25 1990
>***************
>*** 11,17 ****
>--- 11,22 ----
> #define sigmask(no) (1L << ((no) - 1))
> #endif
>
>+ #ifdef HPUX
> #define BLOCK_INPUT_DECLARE() int BLOCK_INPUT_mask
>+ #define BLOCK_INPUT() BLOCK_INPUT_mask = sigblock (sigmask (SIGALRM))
>+ #define UNBLOCK_INPUT() sigsetmask (BLOCK_INPUT_mask)
>+ #else /* not HPUX */
>+ #define BLOCK_INPUT_DECLARE() int BLOCK_INPUT_mask
> #ifdef SIGIO
> #define BLOCK_INPUT() BLOCK_INPUT_mask = sigblock (sigmask (SIGIO))
> #define UNBLOCK_INPUT() sigsetmask (BLOCK_INPUT_mask)
>***************
>*** 19,23 ****
>--- 24,29 ----
> #define BLOCK_INPUT()
> #define UNBLOCK_INPUT()
> #endif /* SIGIO */
>+ #endif /* HPUX */
>
> #define CLASS "emacs" /* class id for GNU Emacs, used in .Xdefaults, etc. */
>*** process.c.orig Fri Sep 15 16:28:36 1989
>--- process.c Fri Apr 20 13:17:06 1990
>***************
>*** 1057,1062 ****
>--- 1057,1079 ----
> signal (SIGHUP, SIG_DFL);
> #endif
> #endif /* HAVE_PTYS */
>+
>+ /* Give child clean slate for signals */
>+ #ifdef SIGCHLD
>+ #ifdef BSD4_1
>+ sigrelse (SIGCHLD);
>+ #else /* not BSD4_1 */
>+ #ifdef HPUX
>+ sigsetmask (0);
>+ #endif /* HPUX */
>+ #if defined (BSD) || defined (UNIPLUS)
>+ sigsetmask (0);
>+ #else /* ordinary USG */
>+ signal (SIGCHLD, sigchld);
>+ #endif /* ordinary USG */
>+ #endif /* not BSD4_1 */
>+ #endif /* SIGCHLD */
>+
> child_setup_tty (xforkout);
> child_setup (xforkin, xforkout, xforkout, new_argv, env);
> }
>***************
>*** 1413,1427 ****
> if (read_kbd && kbd_count)
> nfds = 0;
> else
>! #ifdef IBMRTAIX
> nfds = select (MAXDESC, &Available, 0, 0, &timeout);
> #else
>- #ifdef HPUX
>- nfds = select (MAXDESC, &Available, 0, 0, &timeout);
>- #else
> nfds = select (MAXDESC, &Available, 0, &Exception, &timeout);
> #endif
>- #endif
> xerrno = errno;
>
> if (fix_screen_hook)
>--- 1430,1440 ----
> if (read_kbd && kbd_count)
> nfds = 0;
> else
>! #if defined (IBMRTAIX) || defined (HPUX)
> nfds = select (MAXDESC, &Available, 0, 0, &timeout);
> #else
> nfds = select (MAXDESC, &Available, 0, &Exception, &timeout);
> #endif
> xerrno = errno;
>
> if (fix_screen_hook)
>***************
>*** 1771,1776 ****
>--- 1784,1790 ----
> error ("Process %s is not active",
> XSTRING (p->name)->data);
>
>+ #ifndef HPUX /* Can't get gid of tty's controlling session */
> #ifdef TIOCGPGRP /* Not sure about this! (fnf) */
> /* If we are using pgrps, get a pgrp number and make it negative. */
> if (!NULL (current_group))
>***************
>*** 1785,1790 ****
>--- 1799,1805 ----
> the child itself heads the pgrp. */
> gid = - XFASTINT (p->pid);
> #endif /* not using pgrps */
>+ #endif /* not HPUX */
>
> switch (signo)
> {
>***************
>*** 1800,1805 ****
>--- 1815,1827 ----
> flush_pending_output (XFASTINT (p->infd));
> break;
> }
>+
>+ #ifdef HPUX
>+ if (!NULL (current_group))
>+ ioctl (XFASTINT (p->infd), TIOCSIGSEND, signo);
>+ else
>+ kill (XFASTINT (p->pid), signo);
>+ #else /* not HPUX */
> /* gid may be a pid, or minus a pgrp's number */
> #ifdef BSD
> /* On bsd, [man says] kill does not accept a negative number to kill a pgrp.
>***************
>*** 1808,1813 ****
>--- 1830,1836 ----
> #else /* Not BSD. */
> kill (gid, signo);
> #endif /* Not BSD. */
>+ #endif /* Not HPUX */
>
> /* Put notices in buffers now, since it is safe now.
> Because of this, we know that a process we have just killed
>%
This is a separate patch to process.c, distinct from the immediately
preceding one. Both should be applied.
>*** process.c.~1~ Fri Apr 20 13:17:06 1990
>--- process.c Wed Oct 24 16:38:15 1990
>***************
>*** 1085,1095 ****
> /* If the subfork execv fails, and it exits,
> this close hangs. I don't know why.
> So have an interrupt jar it loose. */
>! signal (SIGALRM, create_process_1);
>! alarm (1);
>! if (forkin >= 0)
>! close (forkin);
>! alarm (0);
> if (forkin != forkout && forkout >= 0)
> close (forkout);
>
>--- 1085,1100 ----
> /* If the subfork execv fails, and it exits,
> this close hangs. I don't know why.
> So have an interrupt jar it loose. */
>!
>! { int old_trap = signal (SIGALRM, create_process_1);
>! int old_alarm = alarm (1);
>! if (forkin >= 0)
>! close (forkin);
>! alarm (0);
>! signal (SIGALRM, old_trap);
>! /* don't bother with possible 1 sec loss */
>! if (old_alarm) alarm (old_alarm);
>! }
> if (forkin != forkout && forkout >= 0)
> close (forkout);
52: What should I do if I have trouble building Emacs?
RMS writes:
If you try to build Emacs and it does not run, the first thing to do is
look in the file called PROBLEMS to see if a solution is given there.
If none is given, then please send a report by mail to
bug-gnu-emacs@prep.ai.mit.edu. Please do not send it to
help-gnu-emacs@prep.ai.mit.edu.
Sending to help-gnu-emacs (which has the effect of posting on
gnu.emacs.help) is undesirable because it takes the time of an
unnecessarily large group of people, most of whom are just users and
have no idea how to fix these problem. bug-gnu-emacs reaches a much
smaller group of people who are more likely to know what to do and have
expressed a wish to receive more messages about Emacs than the others.
Weird/Confusing Problems:
53: Does Emacs have problems with files larger than 8 Megs?
Most installed versions of GNU Emacs will use 24 bit signed integers
(and 24 bit pointers!) internally. This limits the file size that Emacs
can handle to 8388608 bytes.
Leonard N. Zubkoff <lnz@lucid.com> writes:
Putting the following two lines in src/config.h before compiling Emacs
allows for 26 bit integers and pointers:
#define VALBITS 26
#define GCTYPEBITS 5
See question 54 for an explanation.
54: Why does Emacs use 24 bit integers and pointers?
David Gillespie <daveg@csvax.cs.caltech.edu> writes:
Emacs is largely written in a dialect of Lisp; Lisp is a freely-typed
language in the sense that you can put any value of any type into any
variable, or return it from a function, and so on. So each value must
carry a "tag" along with it identifying what kind of thing it is, e.g.,
integer, pointer to a list, pointer to an editing buffer, and so on.
Emacs uses standard 32-bit integers for data objects, taking the top 8
bits for the tag and the bottom 24 bits for the value. So integers (and
pointers) are somewhat restricted compared to true C integers and
pointers.
Emacs uses 8-bit tags because that's a little faster on byte-oriented
machines, but there are only really enough tags to require 6 bits. See
question 53 to find how to recompile Emacs with 6-bit tags and
26-bit integers and pointers if space is at a premium for you.
55: Why does Emacs start up using the wrong directory?
Most likely, you have an environment variable named PWD that is set to a
value other than the name of your current directory. This is most
likely caused by using two different shell programs. "ksh" and (some
versions of) "csh" set and maintain the value of the PWD environment
variable, but "sh" doesn't. If you start sh from ksh, change your
current directory inside sh, and then start Emacs from inside sh, PWD
will have the wrong value but Emacs will use this value. See question
72.
56: How do I edit a file with a "$" in its name?
When entering a filename in the minibuffer, Emacs will attempt to expand
a "$" followed by a word as an environment variable. To suppress this
behavior, type "$$" instead.
57: Why does Shell Mode lose track of the shell's current directory?
Emacs has no way of knowing when the shell actually changes its
directory. So it tries to guess by recognizing cd commands. A number
of fixes and enhancements to Shell Mode have been written, check the
Emacs Lisp Archive (question 14).
58: Why doesn't Emacs expand my aliases when sending mail?
First, you must separate multiple addresses with commas.
Emacs normally only reads the ".mailrc" file once per session, when you
start to compose your first mail message. If you edit .mailrc, you can
type "M-ESC (build-mail-aliases) RET" to make Emacs reread .mailrc.
59: Why doesn't setting default-directory always work?
There is a separate value of default-directory for each Emacs buffer.
The value in the current buffer is the one that is used.
60: Why does my change to load-path not work?
If you added file names with tildes (~) in them to your load-path,
you'll need to do something like this:
(setq-default load-path (mapcar 'expand-file-name load-path))
61: Why does the cursor always go to the wrong column when I move up or
down one line?
You have inadvertently typed "C-x C-n" (set-goal-column) which sets the
"goal-column" to the column where the cursor was. To undo this type
"C-u C-x C-n".
If you make this mistake frequently, you might want to unbind this
command by doing (define-key ctl-x-map "\C-n" nil).
62: Why don't the command line switches documented in the man page work?
The "man page", ie. the file etc/emacs.1, hasn't been updated in a
while. The command line switches that it says work with X Windows are
for X10. See question 72 for details on what switches are
available using X11. {Would someone like to rewrite the man page? RMS
asked me and I haven't had time yet.}
63: Why does gnus-Subject-mail-reply chop off the last character of the
article headers?
There is a bug in some rnews functions that GNUS uses. Here is a patch:
(Patch supplied by Karl Kleinpaste <karl_kleinpaste@cis.ohio-state.edu>.)
>*** rnewspost.el.orig Thu Oct 5 10:32:47 1989
>--- rnewspost.el Thu Oct 5 10:32:47 1989
>***************
>*** 252,258 ****
> (save-restriction
> (narrow-to-region (point-min) (progn (goto-line (point-min))
> (search-forward "\n\n")
>! (- (point) 2)))
> (setq from (mail-fetch-field "from")
> subject (mail-fetch-field "subject")
> reply-to (mail-fetch-field "reply-to")
>--- 252,258 ----
> (save-restriction
> (narrow-to-region (point-min) (progn (goto-line (point-min))
> (search-forward "\n\n")
>! (- (point) 1)))
> (setq from (mail-fetch-field "from")
> subject (mail-fetch-field "subject")
> reply-to (mail-fetch-field "reply-to")
>***************
>*** 289,295 ****
> ;@@ but rnews.el lacks support
> (narrow-to-region (point-min) (progn (goto-char (point-min))
> (search-forward "\n\n")
>! (- (point) 2)))))
> (setq from (mail-fetch-field "from")
> news-reply-yank-from from
> ;; @@ not handling old Title: field
>--- 289,295 ----
> ;@@ but rnews.el lacks support
> (narrow-to-region (point-min) (progn (goto-char (point-min))
> (search-forward "\n\n")
>! (- (point) 1)))))
> (setq from (mail-fetch-field "from")
> news-reply-yank-from from
> ;; @@ not handling old Title: field
>***************
>*** 365,371 ****
> ;@@ but rnews.el lacks support
> (narrow-to-region (point-min) (progn (goto-char (point-min))
> (search-forward "\n\n")
>! (- (point) 2)))))
> (setq news-reply-yank-from (mail-fetch-field "from")
> ;; @@ not handling old Article-I.D.: field
> news-reply-yank-message-id (mail-fetch-field "message-id")))
>--- 365,371 ----
> ;@@ but rnews.el lacks support
> (narrow-to-region (point-min) (progn (goto-char (point-min))
> (search-forward "\n\n")
>! (- (point) 1)))))
> (setq news-reply-yank-from (mail-fetch-field "from")
> ;; @@ not handling old Article-I.D.: field
> news-reply-yank-message-id (mail-fetch-field "message-id")))
64: Why does Emacs hang with message "Unknown XMenu" with X11R4?
Many different X errors can produce this message. Here is the solution
to one problem:
X11 Release 4 now enforces some conditions in the X protocol that were
previously allowed to pass unnoticed. You need to put the X11R4 server
into X11R3 bug compatibility mode for Emacs's Xmenu code to work. You
can do this with the command "xset bc".
65: Why isn't the load average displayed by display-time?
The most likely cause of the problem is that "loadst" can't read the
special file /dev/kmem. To properly install loadst, it should be either
setuid to the owner of /dev/kmem, or is should be setgid to the group to
which /dev/kmem belongs. In either case, /dev/kmem should be readable by
its owner or its group, respectively.
Another possibility is that your version of Unix doesn't have the load
average data available in /dev/kmem. Your version of Unix might have a
special system call to retrieve this information (eg., inq_stats under
UMAX).
66: Why doesn't GNUS work anymore via NNTP?
There is a bug in NNTP version 1.5.10, such that when multiple requests
are sent to the NNTP server, the server only handles the first one before
blocking waiting for more input which never comes.
You can work around the bug inside Emacs like this:
(setq nntp-maximum-request 1)
(setq nntp-buggy-select t)
Or you could fix the NNTP server in one of these ways:
1. Test whether there is data in stdin's stdio buffer before calling
"select" using something like "((fileptr)->_cnt > 0)". Best choice.
2. Remove call to select by not defining "TIMEOUT" in "common/conf.h".
3. Put "setbuf (stdin, NULL)" in the NNTP server in the right spot.
This would be very inefficient.
4. Rewrite the server to use an alarm instead of "select".
67: Why can't I paste a selection I made in Emacs into my xterm window?
A person whose name I lost writes:
The strange interaction between Emacs, xterm, and the cut buffers is
because most xterm clients use selections instead of (well, superceding)
cut buffers. If there is a selection around, xterms will use that
instead of the cut buffer. Emacs doesn't know about selections, but you
can convince xterm to always use the cut buffer instead of selections by
adding this to your .Xdefaults:
! Solution by Thomas Narten, should work under X11R3 and later
! GNU Emacs only copies to CUT_BUFFER0. xterm by default wants to paste
! from the PRIMARY selection.
XTerm*VT100.Translations: #override \
~Meta <Btn2Up>: insert-selection(CUT_BUFFER0)
68: Why does ispell sometimes ignore the local dictionary?
You need to update the version of ispell to 2.0.02. A patch is
available via:
Anonymous FTP:
site: tut.cis.ohio-state.edu
file: /pub/gnu/ispell/patch2.Z
You also need to change a line in ispell.el from:
(defconst ispell-version "2.0.01") ;; Check against output of "ispell -v".
to:
(defconst ispell-version "2.0.02") ;; Check against output of "ispell -v".
Configuring Emacs for yourself:
69: How do I set up a .emacs file properly?
I'm going to be sending out a sample .emacs file in a separate message
later this month. It will be based largely on AT&T's sample .emacs
file, which was sent to me by Dan Jacobson. In fact it might be
identical.
70: How do you debug a .emacs file?
First start Emacs with the "-q" command line option. Then, in the
*scratch* buffer, type the following:
(setq debug-on-error t) C-j
(load-file "~/.emacs") C-j
(C-j stands for Control-J, ie., hold the control key and press J.)
If you have an error in your .emacs file, this will invoke the debugger
when the error occurs. If you don't know how to use the debugger do
(setq stack-trace-on-error t) instead.
WARNING: this will not discover errors caused by trying to do something
that requires the terminal/window-system initialization code to have
been loaded. See question 34.
71: How do I turn on abbrevs by default just in mode XXX?
Put this in your .emacs file:
(condition-case ()
(read-abbrev-file nil t)
(file-error nil))
(setq XXX-mode-hook
(function
(lambda ()
(setq abbrev-mode t))))
72: What does Emacs do every time it starts up?
This question is not itself frequently asked, but helps to answer other
questions.
Here is a complete explanation of the order in which things are done
during startup.
1. These command line switches are handled, only if they appear before
all other command line arguments, and only if they appear in this
order:
"-map". Only under VMS. Instead of "undumping", under VMS the
contents of a dump file are mapped over a section of memory (very
quickly initializing lots of variables).
"-t". The next argument is treated as the name of a terminal to use
instead of whatever is connected to the stdin, stdout, and stderr
FILE pointers.
"-display" (synonym: "-d"). The next argument is treated as the
name of an X Window System server with which to connect. This
overrides the DISPLAY environment variable.
"-nw". This means don't connect to an X Window server even if the
DISPLAY environment variable was set or the "-display" switch was
specified. Instead, Emacs talks to a regular dumb terminal.
"-batch". Emacs is noninteractive. The variable "noninteractive"
is t. This disables any full screen interaction either with a
dumb terminal or with a window system. In addition, Emacs will
die after processing all command line arguments.
All of Emacs's internal data structures are initialized between steps 2
and 5.
2. The "*scratch*" buffer is created and its value of default-directory
is set from the result of calling the "getwd" library routine.
(getwd does what the "pwd" program does.)
3. If X11 is being used, Emacs determines the name under which it looks
for X resource values. (I haven't checked how it does this under
X10.)
1. If there is a "-rn" command line switch, it is handled now, no
matter where it is on the command line. (Later it will be
processed again and ignored.) The next argument is used as
Emacs's resource name.
2. If the "-rn" switch was missing, Emacs looks for the WM_RES_NAME
environment variable.
3. If neither of the previous two steps succeed, Emacs uses the
filename with which it was invoked as its resource name.
4. If X11 is being used, these command line switches are handled now,
no matter where they appear in the command line. (Under X10, they
are not handled until step 13.) WARNING: this can cause strange
interactions with the processing of the remaining command line
arguments in step 13.
In addition, under both X11 and X10, X resources (ie. the stuff in
an .Xdefaults file) are handled now. Command line switches take
precedence over resources.
After each command line switch are given the equivalent resource
names in this format: (X10 name / X11 instance name / X11 class name).
"-r". Reverse video. (ReverseVideo/reverseVideo/ReverseVideo)
"-font" (synonym: "-fn"). Font. (BodyFont/font/Font)
"-wn". Window name (X11 only). (/title/Title)
"-in". Icon name (X11 only). (/iconName/Title)
"-i". Use built-in icon bitmap. (BitMapIcon//)
"-b". Border width (in pixels).
(BorderWidth/borderWidth/BorderWidth)
"-ib". Internal border width (in pixels) (default: 1).
(/internalBorder/BorderWidth)
"-geometry" (synonym: "-w"). Window shape and size. "-geometry" is
X11 only, but its synonym "-w" was available under X10.
(/geometry/Geometry)
Color options only:
"-fg". Foreground color. (Foreground/foreground/Foreground)
"-bg". Background color. (Background/background/Background)
"-bd". Border color. (Border/borderColor/BorderColor)
"-cr". Cursor color. (Cursor/cursorColor/Foreground)
"-ms". Mouse pointer color. (Mouse/pointerColor/Foreground)
In addition, under X10, the resource "Keymap" was handled now.
5. If your Emacs isn't "dumped", the default Lisp libraries are
"loaded" now.
6. Emacs enters a "recursive edit". The first time Emacs does a
recursive edit, steps 7 through 15 are followed, and then Emacs
enters its normal editing mode.
7. If the environment variable PWD exists, the value of
default-directory in the *scratch* buffer is set to the value of
PWD. This often causes problems and confusion when the value of PWD
doesn't actually contain the pathname of the current directory.
8. Now the "-no-init-file" (synonym: "-q") and "-user" (synonym: "-u")
command line switches are handled. They must follow the switches
mentioned in step 1, and must precede any other command line
arguments. Actually, both of these options can be specified;
whichever is last on the command line wins.
9. The user's ".emacs" file is loaded now, if the -no-init-file option
was not specified. If the -user option was specified, that user's
.emacs file is loaded instead. A user's .emacs file is a file named
".emacs" in the user's "home" directory. If no such file exists,
this step has no effect.
10. The system wide "default" file is loaded now, if the variable
inhibit-default-init is nil. The user can set this variable in the
.emacs file to skip this step.
11. If the *scratch* buffer still exists now, the value of
initial-major-mode is called as a function in this buffer.
Normally, this puts this buffer in Lisp Interaction mode.
12. Now the terminal/window-system setup file is loaded. The name of
this file is determined as follows:
1. It starts with the value of term-file-prefix. Normally this is
"term/", but the user can set it to another value. If this
variable is nil, this entire step is skipped.
2. For a window system, the prefix is followed by the name of the
value of the variable window-system, followed by "-win". For X
Windows, this means the name is "term/x-win".
3. For a terminal, the prefix is followed by the value of the
environment variable TERM, for example, "term/vt200". If Emacs
cannot find such a file to load, it retries by stripping off
suffixes that begin with a hyphen. For example, if TERM is
"vt200-40", first Emacs will try "term/vt200-40", and if that
fails it will try "term/vt200".
13. The rest of the command line arguments are handled at this point, in
the order they appear:
"-funcall" (synonyms: "-f", "-e"). The next argument is treated as
the name of a function, which is called with no arguments.
"-load" (synonym: "-l"). The next argument is treated as the name
of a file to load. NOTE: the value of the variable load-path
will be restored after this file is done loading. Thus, trying
to change load-path in such a file will fail!!
"-kill". When this argument is handled, Emacs kills itself.
"+DDD", where D is a digit. DDD is used as a line number to go to.
When the next ordinary file argument on the command line is
handled, Emacs positions point in that file on line DDD.
An ordinary filename. Emacs edits this file using the function
find-file. If the filename is not an "absolute" filename, it is
expanded relative to the value that default-directory in the
*scratch* buffer had at the beginning of step 11. The buffer
visiting this file becomes the current buffer.
"-insert" (synonym: "-i"). The next argument is treated as the name
of a file. This file is inserted in the current buffer.
Initially, this is the *scratch* buffer. Otherwise it is the
buffer visiting the file most recently specified on the command
line. WARNING: the synonym "-i" doesn't work under X11.
Other command line switches may have been created by modifying the
value of the variable command-switch-alist. If X Windows is being
used, all X-specific switches mentioned above in steps 3 and 4 will
be processed. Under X11, they will have already been handled, and
will be ignored now. Under X10, they are being handled for the
first time.
14. The values of the variables term-setup-hook and window-setup-hook
are called as functions with no arguments, in the order listed.
15. If the user did not specify any of the command line options handled
in step 13, and the variable inhibit-startup-message is still nil,
the familiar Emacs startup message is displayed.
73: What are the valid X resource settings (ie., stuff in .Xdefaults file)?
See question 72.
WARNING: Don't pay any attention to the documentation in etc/XDOC or
etc/emacs.1. It's way out of date. {Would someone like to rewrite it?
RMS asked me and I haven't had time yet.}
74: How do I turn down the bell volume in Emacs running under X Windows?
Under Epoch you can do:
(setq epoch::bell-volume 20)
Under normal GNU Emacs you must modify the XTfeep function in
src/x11term.c, and change the number 50 to some other number:
XTfeep ()
{
BLOCK_INPUT_DECLARE ();
#ifdef XDEBUG
fprintf (stderr, "XTfeep\n");
#endif
BLOCK_INPUT ();
XBell (XXdisplay,50); /* change this 50 */
UNBLOCK_INPUT ();
}
75: How do I make Emacs send 8-bit characters to my terminal?
Johan Widen <jw@sics.se> writes:
A patch for emacs-18.55 is available by ftp and mail-server from
sics.se.
Anonymous FTP:
site: sics.se [192.16.123.90]
file: archive/emacs-18.55-8bit-diff
E-mail:
To: mail-server@sics.se
body: send emacs-18.55-8bit-diff
Emacs Lisp programming:
76: What dialect of Lisp is Emacs Lisp?
It's the dialect of Lisp called Emacs Lisp. (No joke!) People also
call it elisp or e-lisp. (NOTE: The term "Elisp" is trademarked by
someone else.)
77: How close is Emacs Lisp to Common Lisp?
Pretty far. GNU Emacs Lisp is case-sensitive, uses dynamic scoping,
doesn't have packages, doesn't have multiple return values, doesn't have
reader macros, etc. For people used to Common Lisp, some of the
functions in Common Lisp that are not in Emacs Lisp by default are
provided in the file lisp/cl.el. There is a Texinfo manual describing
these functions in man/cl.texinfo.
78: How do I execute a piece of Emacs Lisp code?
There are a number of ways to execute (called "evaluate") an Emacs Lisp
"form":
* If you want it evaluated every time you run Emacs, put it in a file
named ".emacs" in your home directory.
* You can type the form in the "*scratch*" buffer, and then type C-j
after it. The result of evaluating the form will be inserted in the
buffer.
* In in Emacs-Lisp mode, typing M-C-x evaluates a top-level form before
or around point.
* Typing "C-x C-e" in any buffer evaluates the Lisp form immediately
before point and prints its value in the echo area.
* Typing M-ESC or M-x eval-expression allows you to type a Lisp form in
the minibuffer which will be evaluated.
* You can use M-x load-file to have Emacs evaluate all the Lisp forms in
a file. (To do this from Lisp use the function "load" instead.)
79: How do I make a set of operations fully local to a region?
Use narrow-to-region inside of save-restriction.
80: How can I highlight a region?
There are ways to get highlighting in GNU Emacs 18.55, but they all
require patching the C code of Emacs and rebuilding. They are also slow
and the highlighting disappears if you scroll or redraw the screen. One
patch is by Kenichi Handa <handa@etl.go.jp>.
You can hightlight regions in a variety of ways in Epoch. GNU Emacs 19
will have everything you need, but won't be out soon.
81: How do I change Emacs's idea of the tab character's length?
Example: (setq default-tab-width 10).
82: What is the difference between (interactive "P") and (interactive "p")?
The value that is a result of "P" can be a list, a symbol, or an
integer; the value that is a result of "p" is always an integer:
Prefix keys typed result of: "P" "p"
nothing nil 1
"M-1" 1 1
"C-u 1" 1 1
"M--" '- -1
"C-u -" '- -1
"C-u" (4) 4
"C-u C-u" (16) 16
Carrying Out Common Tasks:
83: How do I insert ">"'s in the beginning of every line in a buffer?
Type "M-x replace-regexp RET ^ RET > RET". ("replace-regexp" can be
shortened to "repl TAB r".)
To do this only in the region, type "C-x n M-x replace-regexp RET ^ RET
> RET C-x w". (You're going to remember that, right?)
84: How do I insert "_^H" characters before each character in a
paragraph to get an underlined paragraph?
M-x underline-region.
85: How do I repeat a command as many times as possible?
Make a keyboard macro that invokes the command and then type "M-0 C-x
e". WARNING: any messages your command prints in the echo area will be
suppressed.
86: How do I search for an unprintable (8-bit) character that appears
in a buffer as \237?
C-s C-q 2 3 7
(This assumes the value of search-quote-char is 17 (C-q).)
87: How do I tell Emacs to automatically indent a new line to the
indentation of the previous line?
M-x indented-text-mode. (This is a major mode.)
If you have auto-fill mode on (minor mode), you can tell Emacs to prefix
every line with a certain character sequence, the "fill prefix". Type
the prefix at the beginning of a line, position point after it, and then
type "C-x ." (set-fill-prefix) to set the fill prefix. Thereafter,
auto-filling will automatically put the fill prefix at the beginning of
new lines, and M-q (fill-paragraph) will maintain any fill prefix when
refilling the paragraph.
88: How do I make Emacs "typeover" or "overwrite" when I type instead of
always inserting?
M-x overwrite-mode (minor mode).
WARNING: delete-backward-char (usually the delete key) doesn't work
properly in overwrite mode. It deletes the character to the left,
rather than replacing it with a space.
89: How do I show which parenthesis matches the one I'm looking at?
If you're looking at a right parenthesis (or brace or bracket) you can
delete it and reinsert it.
M-C-f (forward-sexp) and M-C-b (backward-sexp) will skip over balanced
parentheses, so you can see which parentheses match. (You can train it
to skip over balanced brackets and braces at the same time by modifying
the syntax table.)
Here is some Emacs Lisp that will make the % key show the matching
parenthese, like in vi. In addition, if the cursor isn't over a
parenthese, it simply inserts a % like normal.
(By an unknown contributor.)
(global-set-key "%" 'match-paren)
(defun match-paren (arg)
"Go to the matching parenthesis if on parenthesis otherwise insert %."
(interactive "p")
(cond ((looking-at "[([{]") (forward-sexp 1) (backward-char))
((looking-at "[])}]") (forward-char) (backward-sexp 1))
(t (self-insert-command (or arg 1)))))
90: How do I make Emacs behave like this: when I go up or down, the cursor
should stay in the same column even if the line is too short?
M-x picture-mode. (This is a minor mode, in theory anyway ...)
91: How do I read news under Emacs?
There are at least three news reading packages that operate inside
Emacs. "rnews" comes with Emacs. "GNUS" and "Gnews" come separately.
{I've never used rnews; could someone write a description? BTW, rnews
will be replaced with GNUS in Emacs 19.}
Both GNUS and Gnews handle reading news over NNTP. I think both can
also read from a local news spool. GNUS also supports reading mail
stored in MH folders or articles saved by GNUS.
GNUS is written (mostly) by Masanobu Umeda. His (?) latest e-mail
address was umerin@tc.nagasaki.go.jp, but I don't think he has an e-mail
address right now. The latest version is GNUS 3.13. There is a
newsgroup for discussion of GNUS called gnu.emacs.gnus. This newsgroup
is gatewayed with the mailing list info-gnus-english to subscribe send
mail to info-gnus-english-request@cis.ohio-state.edu. There is also a
mailing list called info-gnus, which includes discussion in Japanese.
Gnews was written by Matthew Wiener <e-mail address?>. {Could someone
tell me the # of the latest version, and how long it has been since
anyone has heard from Matthew?}. There is a newsgroup for Gnews called
gnu.emacs.gnews.
92: In C mode, can I show just the lines that will be left after #ifdef
commands are handled by the compiler?
M-x hide-ifdef-mode. (This is a minor mode.)
You may have to (load "hideif") first. If you want to do this
regularly, put this in your .emacs file:
(autoload 'hide-ifdef-mode "hideif" nil t)
93: Is there an equivalent to the "." (dot) command of vi?
("." is the redo command in vi. It redoes the last insertion/deletion.)
No, not really.
You can type "C-x ESC" (repeat-complex-command) to reinvoke commands
that used the minibuffer to get arguments. In repeat-complex-command
you can type M-p and M-n to scan through all the different complex
commands you've typed.
To repeat something on each line I recommend using keyboard macros.
94: How do I use emacstool under SunView?
The file etc/SUN-SUPPORT includes the document "Using Emacstool with GNU
Emacs". Also read the man page for emacstool (etc/emacstool.1).
95: How do I get Emacs to display the current line number on the mode line?
There is no "correct" way to constantly display the current line number
on the mode line in Emacs 18. Emacs is not a line-oriented editor, and
really has no idea what "lines" of the buffer are displayed in the
window. It would require a lot of work at the C code level to make
Emacs keep track of this.
Emacs 19 will probably be able to do this, but probably not with great
efficiency.
To find out what line of the buffer you are on right now, do "M-x
what-line". Typing "C-x l" will also tell you what line you are on,
provided the buffer isn't separated into "pages" with C-l characters.
In that case, it will only tell you what line of the current "page" you
are on. WARNING: "C-x l" gives the wrong value when point is at the
beginning of a line.
People have written various kludges to display the current line number
on the mode line. Look in the Lisp Code Directory. (See question 13.)
96: How do I tell Emacs to iconify itself?
You need to modify C source and recompile. Either that or get Epoch
instead. For the interested I have a patch to allow Emacs to iconify
itself.
97: How do I use regexps (regular expressions) in Emacs?
This is documented in the Emacs manual. To read the manual section
online, type "C-h i m emacs RET m regexps RET".
WARNING: Unlike in Unix grep, sed, etc., a complement character set
([^...]) can match a Newline, unless Newline is mentioned as one of the
characters not to match.