Richard M. Stallman [Sun, 23 May 1993 23:10:14 +0000 (23:10 +0000)]
(finder-known-keywords): Use i18n, not i14n.
(finder-compile-keywords): Substitute i18n for i14n.
Turn off undo in *finder-scratch*. Ignore file names starting with =.
(finder-mode, finder-current-item): Rename headmark to finder-headmark.
(finder-list-matches, finder-list-keywords): Likewise.
Jim Blandy [Sun, 23 May 1993 22:31:53 +0000 (22:31 +0000)]
Changes for SGI from Matthew J Brown <mjb@doc.ic.ac.uk>.
* m/iris4d.h, m/iris5d.h: Don't use the --cckr CC switch if we're
using GCC.
(NEED_SIOCTL): Move this to...
* s/irix3-3.h (NEED_SIOCTL): ... here; apparently it's not
necessary in irix4-0.h.
* s/irix4-0.h: Remove declaration of getpty; apparently Irix 4.0
already declares this elsewhere.
(PTY_TTY_NAME_SPRINTF): #undef this before re-#defining it.
* m/ibmps2-aix.h, m/pfa50.h, s/hpux.h, s/isc2-2.h, s/linux.h,
s/sco4.h, s/usg5-4.h: Remove HAVE_RENAME; configure guesses that now.
* m/iris4d.h, m/iris5d.h: Don't use the --cckr CC switch if we're
using GCC.
Jim Blandy [Sun, 23 May 1993 22:30:42 +0000 (22:30 +0000)]
Changes for SGI from Matthew J Brown <mjb@doc.ic.ac.uk>.
* m/iris4d.h, m/iris5d.h: Don't use the --cckr CC switch if we're
using GCC.
(NEED_SIOCTL): Move this to...
* s/irix3-3.h (NEED_SIOCTL): ... here; apparently it's not
necessary in irix4-0.h.
* s/irix4-0.h: Remove declaration of getpty; apparently Irix 4.0
already declares this elsewhere.
(PTY_TTY_NAME_SPRINTF): #undef this before re-#defining it.
* m/ibmps2-aix.h, m/pfa50.h, s/hpux.h, s/isc2-2.h, s/linux.h,
s/sco4.h, s/usg5-4.h: Remove HAVE_RENAME; configure guesses that now.
Jim Blandy [Sun, 23 May 1993 21:04:16 +0000 (21:04 +0000)]
* configure.in: Explain that this is an autoconf script, and give
instructions for rebuilding configure from it. Arrange to put
comments in configure explaining this too.
Roland McGrath [Sat, 22 May 1993 22:33:19 +0000 (22:33 +0000)]
(vc-comment-to-change-log): Don't take FILE argument, since
vc-update-change-log doesn't support it anyway.
Don't bind default-directory. Instead pass second arg to file-relative-name.
(vc-update-change-log): Use find-change-log instead of hardcoding.
Jim Blandy [Sat, 22 May 1993 22:09:25 +0000 (22:09 +0000)]
* cl.el (cl-floor, cl-ceiling, cl-truncate, cl-round): Renamed
from floor, ceiling, truncate, and round; the old names conflict
with built-in functions.
Richard M. Stallman [Sat, 22 May 1993 21:25:32 +0000 (21:25 +0000)]
(gud-def): Fix inclusion of the define-key.
(gdb, sdb, dbx): Change bindings from letters to control chars.
(gud-common-init): Add save-excursion.
(gud-display-line): Don't mess with buffer-read-only.
(gud-filter): Set output-after-point *after* deleting old prompt.
Jim Blandy [Sat, 22 May 1993 07:51:26 +0000 (07:51 +0000)]
* keyboard.c (make_ctrl_char): New function.
(read_char): Call it.
(kbd_buffer_store_event): Call it to see if the new character is
the quit character.
Jim Blandy [Fri, 21 May 1993 23:47:10 +0000 (23:47 +0000)]
* systty.h [SYSV_PTYS]: #include <sys/types.h>. Francesco
Potorti` <pot@fly.CNUCE.CNR.IT> says it's necessary on his
machine, and it should be harmless.
Jim Blandy [Thu, 20 May 1993 06:30:20 +0000 (06:30 +0000)]
Some more changes from Michael K. Johnson for Linux.
* s/template.h: Mention that you should #define HAVE_TERMIO or
HAVE_TERMIOS, but not both, and that HAVE_TERMIOS is preferred.
* systty.h (EMACS_SET_TTY_PGRP): Don't assign the return value of
tcsetpgrp to *pgid; it's just a status value.
* config.h.in (HAVE_RANDOM): This shouldn't be defined on Linux
systems using XFree386, and perhaps is inappropriate in general.
* m/intel386.h: #undefine the integer size macros, since the Linux
<values.h> file #defines them itself.
* mem-limits.h (get_lim_data): Linux has the ulimit call; if it
fails, fall back on ULIMIT_BREAK_VALUE.
* process.c: Linux needs the WAITTYPE, etc. definitions.
* unexec.c (hdr, ohdr): Linux has the ordinary `struct exec' type;
no need to use SYSV names.
* s/linux.h: New file.
Jim Blandy [Thu, 20 May 1993 06:29:45 +0000 (06:29 +0000)]
Some time-handling patches from Paul Eggert:
* editfns.c (Fcurrent_time_zone): Take an optional argument specifying
what (absolute) time should be used to determine the current time zone.
Yield just offset and name of time zone, including DST correction.
Yield time zone offset in seconds, not minutes.
(lisp_time_argument, difftm): New functions.
(Fcurrent_time_string): Use lisp_time_argument.
* systime.h (EMACS_CURRENT_TIME_ZONE, EMACS_GET_TZ_OFFSET,
EMACS_GET_TZ_NAMES): Remove.
* config.h.in: Add HAVE_TM_ZONE.
Jim Blandy [Thu, 20 May 1993 06:28:46 +0000 (06:28 +0000)]
Some time-handling patches from Paul Eggert:
* calendar.el (calendar-current-time-zone): New function.
(calendar-time-zone, calendar-standard-time-zone-name,
calendar-daylight-time-zone-name): Use it instead of current-time-zone.
* sendmail.el (mail-do-fcc): Use the same absolute time for both
current-time-string and current-time-zone. Adjust to new format
returned by current-time-zone.
Jim Blandy [Thu, 20 May 1993 02:14:56 +0000 (02:14 +0000)]
Some more changes from Michael K. Johnson for Linux.
* s/template.h: Mention that you should #define HAVE_TERMIO or
HAVE_TERMIOS, but not both, and that HAVE_TERMIOS is preferred.
* systty.h (EMACS_SET_TTY_PGRP): Don't assign the return value of
tcsetpgrp to *pgid; it's just a status value.
* config.h.in (HAVE_RANDOM): This shouldn't be defined on Linux
systems using XFree386, and perhaps is inappropriate in general.
* m/intel386.h: #undefine the integer size macros, since the Linux
<values.h> file #defines them itself.
* mem-limits.h (get_lim_data): Linux has the ulimit call; if it
fails, fall back on ULIMIT_BREAK_VALUE.
* process.c: Linux needs the WAITTYPE, etc. definitions.
* unexec.c (hdr, ohdr): Linux has the ordinary `struct exec' type;
no need to use SYSV names.
* s/linux.h: New file.
Jim Blandy [Wed, 19 May 1993 22:01:35 +0000 (22:01 +0000)]
Some changes from Michael K. Johnson for Linux.
* sysdep.c (sys_siglist): Don't define this if HAVE_SYS_SIGLIST is
#defined. That lets the system provide it, if it has it.
* syssignal.h (sigmask): Only define this if <signal.h> hasn't
given us a definition already.
* syssignal.h (sys_sigpause): Fix argument in prototype.
* sysdep.c (init_signals): The masks are called empty_mask and
full_mask, not signal_empty_mask and signal_full_mask.
(signal_handler_t): Moved ....
* syssignal.h: ... to here.
* systty.h (EMACS_SET_TTY_PGRP): Call tcsetpgrp with the correct
arguments.
* emacs.c (main): Don't try to establish signal handlers for
SIGBUS and SIGSYS unless they're actually #defined.
Jim Blandy [Wed, 19 May 1993 22:01:21 +0000 (22:01 +0000)]
Some changes from Michael K. Johnson for Linux.
* sysdep.c (sys_siglist): Don't define this if HAVE_SYS_SIGLIST is
#defined. That lets the system provide it, if it has it.
* syssignal.h (sigmask): Only define this if <signal.h> hasn't
given us a definition already.
* syssignal.h (sys_sigpause): Fix argument in prototype.
* sysdep.c (init_signals): The masks are called empty_mask and
full_mask, not signal_empty_mask and signal_full_mask.
(signal_handler_t): Moved ....
* syssignal.h: ... to here.
* systty.h (EMACS_SET_TTY_PGRP): Call tcsetpgrp with the correct
arguments.
* emacs.c (main): Don't try to establish signal handlers for
SIGBUS and SIGSYS unless they're actually #defined.
Jim Blandy [Wed, 19 May 1993 19:34:12 +0000 (19:34 +0000)]
* calendar.el (hebrew-calendar-yahrzeit): Correct error from S-P&E
paper in test for Adar I 30 date of death for yahrzeit in a
non-leap year when Shevat 29 must be used.