OpenBSD-4.6/usr.bin/vi/FAQ
@(#)FAQ 8.13 (Berkeley) 10/14/96
Q: How can I get vi to display my character set?
A: Vi uses the C library routine isprint(3) to determine if a character
is printable, or should be displayed as an octal or hexadecimal value
on the screen. Generally, if vi is displaying printable characters
in octal/hexadecimal forms, your environment is not configured correctly.
Try looking at the man pages that allow you to configure your locale.
For example, to configure an ISO 8859-1 locale under Solaris using csh,
you would do:
setenv LANG C
setenv LC_CTYPE iso_8859_1
Other LC_CTYPE systems/values that I'm told work:
System Value
====== =====
FreeBSD lt_LN.ISO_8859-1
HP-UX 9.X american.iso88591
HP-UX 10.X en_US.iso88591
SunOS 4.X iso_8859_1
SunOS 5.X iso_8859_1
If there's no other solution, you can use the print and noprint edit
options of vi to specify that a specific character is printable or not
printable.
Q: My map won't work!
A: One thing that you should immediately check if a vi map doesn't work
is if depends on the final cursor position after a P or p command.
Historic vi's were inconsistent as to the final position of the cursor,
and, to make matter worse, the final cursor position also depended on
whether the put text came from a named or unnamed buffer! Vi follows
the POSIX 1003.2 standard on this one, and makes this consistent, always
placing the cursor on the first character.
Q: I'm using ksh or csh as my vi edit option shell value, and file
expansions don't work right!
A: The problem may be in your ksh or csh startup files, e.g., .cshrc. Vi
executes the shell to do name expansion, and the shell generally reads
its startup files. If the startup files are not correctly configured
for non-interactive use, e.g., they always echo a prompt to the screen,
vi will be unable to parse the output and things will not work
correctly.
Q: How does the iclower edit option differ from the ignorecase (i.e. ic)
edit option?
A: The difference is that the ignorecase edit option always ignores the
case of letters in the Regular Expression (RE), and the iclower edit
option only ignores the case if there are no upper-case letters in the
RE. If any upper-case letters appear in the Regular Expression, then
it will be treated case-sensitively, as if the ignorecase edit option
was not set.
Q: When I edit binary files, vi appends a <newline> to the last line!
A: This is historic practice for vi, and further, it's required by the
POSIX 1003.2 standard. My intent is to provide a command line and/or
edit option to turn this behavior off when I switch to version 2.0 of
the Berkeley DB package.
Q: My cursor keys don't work when I'm in text input mode!
A: A common problem over slow links is that the set of characters sent by
the cursor keys don't arrive close enough together for vi to understand
that they are a single keystroke, and not separate keystrokes. Try
increasing the value of the escapetime edit option, which will cause
vi to wait longer before deciding that the <escape> character that
starts cursor key sequences doesn't have any characters following it.
Q: When I edit some files, vi seems to hang forever, and I have to kill it.
A: Vi uses flock(2) and fcntl(2) to do file locking. When it attempts to
acquired a lock for a file on an NFS mounted filesystem, it can hang
for a very long (perhaps infinite) period of time. Turning off the
"lock" edit option will keep vi from attempting to acquire any locks
on the files you edit.
Q: When I compile vi I get lots of warnings about pointer assignments
being incompatible!
A: Vi is partially written to support wide characters. When this code
interfaces with the code that doesn't yet support wide characters,
the pointer types clash. This will hopefully be fixed in the near
future, but I've been saying that for awhile, now.
Q: I get jumpy scrolling behavior in the screen!
A: This is almost certainly a problem with the system's terminfo or
termcap information for your terminal. If the terminfo/termcap entry
doesn't have the settable scrolling region capabilities, or the more
powerful scrolling commands, these behaviors can result. Historic
implementations of vi, and some of the vi clones, don't suffer from
this problem because they wrote their own screen support instead of
using the curses library.
The solution is to find a good terminfo or termcap entry for your
terminal, which will fix the problem for all of the applications on
your system, not just vi. Eric Raymond maintains the freely
redistributable termcap/terminfo entries. They can be downloaded
from http://www.ccil.org/~esr/ncurses.html, or you can contact him
at esr@snark.thyrsus.com.
Q: The entire screen repaints on every keystroke!
A: Your system's curses implementation is broken. You should use the
curses implementation provided with vi or a curses replacement such
as ncurses. Eric Raymond is one of the maintainers of the freely
redistributable ncurses package. You can download ncurses from
http://www.ccil.org/~esr/ncurses.html, or you can contact him at
esr@snark.thyrsus.com.
Q: When I use vi on a Sun console (terminal type sun-34) the screen
is occasionally trashed, usually when exiting vi!
A: The Sun console can't handle the 'al' capability of the termcap
entry (the il1 capability of terminfo entries). If you delete that
entry from your terminfo/termcap information everything should work
correctly.
Q: I don't have a version of ctags (or I have ctags, but it doesn't tag
nearly enough things)!
A: There's a version of ctags available on the 4.4BSD-Lite distributions,
as well as the OpenBSD, FreeBSD, NetBSD, Linux and GNU distributions.
Or, you might want to try Exuberant Ctags:
Title: Exuberant Ctags
Version: 1.3
Entered-date: 16JUN96
Description:
A better ctags which generates tags for all possible tag types:
macro definitions, enumerated values (values inside enum{...}),
function and method definitions, enum/struct/union tags, external
function prototypes (optional), typedefs, and variable
declarations. It is far less easily fooled by code containing #if
preprocessor conditional constructs, using a conditional path
selection algorithm to resolve complicated choices, and a
fall-back algorithm when this one fails. Can also be used to print
out a list of selected objects found in source files.
Keywords: ctags, tags, exuberant
Author: darren@sirsi.com (Darren Hiebert)
darren@hiwaay.net (Darren Hiebert)
Maintained-by: darren@sirsi.com (Darren Hiebert)
darren@hiwaay.net (Darren Hiebert)
Primary-site: sunsite.unc.edu /pub/Linux/devel/lang/c
27kB ctags-1.3.tar.gz
Alternate-site: ftp.halcyon.com /local/gvr
27kB ctags-1.3.tar.gz
Original-site:
Platforms: UNIX, MSDOS, WindowsNT, Windows95, OS/2, Amiga
Copying-policy: Public domain
Q: When I update a file I already have open, and use :e to reread it, I
get nul's for the rest of the file!
A: Your system's implementation of mmap(2) has a bug; you will have to
exit vi and re-execute it.
Q: Where can I get cscope?
A: Cscope is available on UNIXWare System V Release 4.0 variants such as
Sun Solaris 2.x (/opt/SUNWspro/bin) and UNIXWare System V Release 4.1.
You can buy version 13.3 source with an unrestricted license for $400
from AT&T Software Solutions by calling +1-800-462-8146. Binary
redistribution of cscope is an additional $1500, one-time flat fee.
For more information, see http://www.unipress.com/att/new/cscope.html.