1 GNU coreutils NEWS -*- outline -*-
3 * Major changes in release 5.3.0 (2004-03-17) [unstable]
7 ptx now diagnoses invalid values for its --width=N (-w)
8 and --gap-size=N (-g) options.
10 tee now exits when it gets a SIGPIPE signal, as POSIX requires.
11 To get tee's old behavior, use the shell command "(trap '' PIPE; tee)".
15 pwd now works even when run from a working directory whose name
16 is longer than PATH_MAX.
18 cp -pu and mv -u (when copying) now don't bother to update the
19 destination if the resulting time stamp would be no newer than the
20 preexisting time stamp. This saves work in the common case when
21 copying or moving multiple times to the same destination in a file
22 system with a coarse time stamp resolution.
24 dd has new conversions for the conv= option:
26 nocreat do not create the output file
27 excl fail if the output file already exists
28 fdatasync physically write output file data before finishing
29 fsync likewise, but also write metadata
31 dd has new iflag= and oflag= options with the following flags:
33 append append mode (makes sense for output file only)
34 direct use direct I/O for data
35 dsync use synchronized I/O for data
36 sync likewise, but also for metadata
37 nonblock use non-blocking I/O
38 nofollow do not follow symlinks
40 stty now provides support (iutf8) for setting UTF-8 input mode.
42 With stat, a specified format is no longer automatically newline terminated.
43 If you want a newline at the end of your output, append `\n' to the format
46 'df', 'du', and 'ls' now take the default block size from the
47 BLOCKSIZE environment variable if the BLOCK_SIZE, DF_BLOCK_SIZE,
48 DU_BLOCK_SIZE, and LS_BLOCK_SIZE environment variables are not set.
49 Unlike the other variables, though, BLOCKSIZE does not affect
50 values like 'ls -l' sizes that are normally displayed as bytes.
51 This new behavior is for compatibility with BSD.
53 du accepts a new option --files0-from=FILE, where FILE contains a
54 list of NUL-terminated file names.
56 `date -d' and `touch -d' now accept integer counts of seconds since
57 1970 when prefixed by `@'. For example, `@321' represents
58 1970-01-01 00:05:21 UTC.
60 `date -d', `date -f' and `touch -d' now handle fractional time
61 stamps like 2004-02-27 14:19:13.489392193.
63 `date' has a new option --iso-8601=ns that outputs
64 nanosecond-resolution time stamps.
67 * Major changes in release 5.2.1 (2004-03-12) [stable]
71 mv could mistakenly fail to preserve hard links when moving two
72 or more arguments between partitions.
74 `cp --sparse=always F /dev/hdx' no longer tries to use lseek to create
75 holes in the destination.
77 nohup now sets the close-on-exec flag for its copy of the stderr file
78 descriptor. This avoids some nohup-induced hangs. For example, before
79 this change, if you ran `ssh localhost', then `nohup sleep 600 </dev/null &',
80 and then exited that remote shell, the ssh session would hang until the
81 10-minute sleep terminated. With the fixed nohup, the ssh session
82 terminates immediately.
84 `expr' now conforms to POSIX better:
86 Integers like -0 and 00 are now treated as zero.
88 The `|' operator now returns 0, not its first argument, if both
89 arguments are null or zero. E.g., `expr "" \| ""' now returns 0,
92 The `|' and `&' operators now use short-circuit evaluation, e.g.,
93 `expr 1 \| 1 / 0' no longer reports a division by zero.
97 `chown user.group file' now has its traditional meaning even when
98 conforming to POSIX 1003.1-2001, so long as no user has a name
99 containing `.' that happens to equal `user.group'.
102 * Major changes in release 5.2.0 (2004-02-19) [stable]
109 * Major changes in release 5.1.3 (2004-02-08): candidate to become stable 5.2.0
113 `cp -d' now works as required even on systems like OSF V5.1 that
114 declare stat and lstat as `static inline' functions.
116 time stamps output by stat now include actual fractional seconds,
117 when available -- or .0000000 for files without that information.
119 seq no longer infloops when printing 2^31 or more numbers.
120 For reference, seq `echo 2^31|bc` > /dev/null takes about one hour
121 on a 1.6 GHz Athlon 2000 XP. Now it can output 2^53-1 numbers before
124 * Major changes in release 5.1.2 (2004-01-25):
128 rmdir -p exits with status 1 on error; formerly it sometimes exited
129 with status 0 when given more than one argument.
131 nohup now always exits with status 127 when it finds an error,
132 as POSIX requires; formerly it sometimes exited with status 1.
134 Several programs (including cut, date, dd, env, hostname, nl, pr,
135 stty, and tr) now always exit with status 1 when they find an error;
136 formerly they sometimes exited with status 2.
138 factor no longer reports a usage error if stdin has the wrong format.
140 paste no longer infloops on ppc systems (bug introduced in 5.1.1)
143 * Major changes in release 5.1.1 (2004-01-17):
145 ** Configuration option
147 You can select the default level of POSIX conformance at configure-time,
148 e.g., by ./configure DEFAULT_POSIX2_VERSION=199209
152 fold -s works once again on systems with differing sizes for int
153 and size_t (bug introduced in 5.1.0)
157 touch -r now specifies the origin for any relative times in the -d
158 operand, if both options are given. For example, "touch -r FOO -d
159 '-5 seconds' BAR" sets BAR's modification time to be five seconds
162 join: The obsolete options "-j1 FIELD", "-j2 FIELD", and
163 "-o LIST1 LIST2..." are no longer supported on POSIX 1003.1-2001 systems.
164 Portable scripts should use "-1 FIELD", "-2 FIELD", and
165 "-o LIST1,LIST2..." respectively. If join was compiled on a
166 POSIX 1003.1-2001 system, you may enable the old behavior
167 by setting _POSIX2_VERSION=199209 in your environment.
170 * Major changes in release 5.1.0 (2003-12-21):
174 chgrp, chmod, and chown can now process (with -R) hierarchies of virtually
175 unlimited depth. Before, they would fail to operate on any file they
176 encountered with a relative name of length PATH_MAX (often 4096) or longer.
178 chgrp, chmod, chown, and rm accept the new options:
179 --preserve-root, --no-preserve-root (default)
181 chgrp and chown now accept POSIX-mandated -L, -H, and -P options
183 du can now process hierarchies of virtually unlimited depth.
184 Before, du was limited by the user's stack size and it would get a
185 stack overflow error (often a segmentation fault) when applied to
186 a hierarchy of depth around 30,000 or larger.
188 du works even when run from an inaccessible directory
190 du -D now dereferences all symlinks specified on the command line,
191 not just the ones that reference directories
193 du now accepts -P (--no-dereference), for compatibility with du
194 of NetBSD and for consistency with e.g., chown and chgrp
196 du's -H option will soon have the meaning required by POSIX
197 (--dereference-args, aka -D) rather then the current meaning of --si.
198 Now, using -H elicits a warning to that effect.
200 When given -l and similar options, ls now adjusts the output column
201 widths to fit the data, so that output lines are shorter and have
202 columns that line up better. This may adversely affect shell
203 scripts that expect fixed-width columns, but such shell scripts were
204 not portable anyway, even with old GNU ls where the columns became
205 ragged when a datum was too wide.
207 du accepts a new option, -0/--null, to make it produce NUL-terminated
212 printf, seq, tail, and sleep now parse floating-point operands
213 and options in the C locale. POSIX requires this for printf.
215 od -c -w9999999 no longer segfaults
217 csplit no longer reads from freed memory (dumping core on some systems)
219 csplit would mistakenly exhaust virtual memory in some cases
221 ls --width=N (for very large N) is no longer subject to an address
222 arithmetic bug that could result in bounds violations.
224 ls --width=N (with -x or -C) no longer allocates more space
225 (potentially much more) than necessary for a given directory.
227 dd `unblock' and `sync' may now be combined (e.g., dd conv=unblock,sync)
229 * Major changes in release 5.0.91 (2003-09-08):
233 date accepts a new option --rfc-2822, an alias for --rfc-822.
235 split accepts a new option -d or --numeric-suffixes.
237 cp, install, mv, and touch now preserve microsecond resolution on
238 file timestamps, on platforms that have the 'utimes' system call.
239 Unfortunately there is no system call yet to preserve file
240 timestamps to their full nanosecond resolution; microsecond
241 resolution is the best we can do right now.
243 sort now supports the zero byte (NUL) as a field separator; use -t '\0'.
244 The -t '' option, which formerly had no effect, is now an error.
246 sort option order no longer matters for the options -S, -d, -i, -o, and -t.
247 Stronger options override weaker, and incompatible options are diagnosed.
249 `sha1sum --check' now accepts the BSD format for SHA1 message digests
250 in addition to the BSD format for MD5 ones.
252 who -l now means `who --login', not `who --lookup', per POSIX.
253 who's -l option has been eliciting an unconditional warning about
254 this impending change since sh-utils-2.0.12 (April 2002).
258 Mistakenly renaming a file onto itself, e.g., via `mv B b' when `B' is
259 the same directory entry as `b' no longer destroys the directory entry
260 referenced by both `b' and `B'. Note that this would happen only on
261 file systems like VFAT where two different names may refer to the same
262 directory entry, usually due to lower->upper case mapping of file names.
263 Now, the above can happen only on file systems that perform name mapping and
264 that support hard links (stat.st_nlink > 1). This mitigates the problem
265 in two ways: few file systems appear to be affected (hpfs and ntfs are),
266 when the bug is triggered, mv no longer removes the last hard link to a file.
267 *** ATTENTION ***: if you know how to distinguish the following two cases
268 without writing to the file system in question, please let me know:
269 1) B and b refer to the same directory entry on a file system like NTFS
270 (B may well have a link count larger than 1)
271 2) B and b are hard links to the same file
273 stat no longer overruns a buffer for format strings ending in `%'
275 fold -s -wN would infloop for N < 8 with TABs in the input.
276 E.g., this would not terminate: printf 'a\t' | fold -w2 -s
278 `split -a0', although of questionable utility, is accepted once again.
280 `df DIR' used to hang under some conditions on OSF/1 5.1. Now it doesn't.
282 seq's --width (-w) option now works properly even when the endpoint
283 requiring the larger width is negative and smaller than the other endpoint.
285 seq's default step is 1, even if LAST < FIRST.
287 paste no longer mistakenly outputs 0xFF bytes for a nonempty input file
288 without a trailing newline.
290 `tail -n0 -f FILE' and `tail -c0 -f FILE' no longer perform what amounted
291 to a busy wait, rather than sleeping between iterations.
293 tail's long-undocumented --allow-missing option now elicits a warning
296 * Major changes in release 5.0.90 (2003-07-29):
300 sort is now up to 30% more CPU-efficient in some cases
302 `test' is now more compatible with Bash and POSIX:
304 `test -t', `test --help', and `test --version' now silently exit
305 with status 0. To test whether standard output is a terminal, use
306 `test -t 1'. To get help and version info for `test', use
307 `[ --help' and `[ --version'.
309 `test' now exits with status 2 (not 1) if there is an error.
311 wc count field widths now are heuristically adjusted depending on the input
312 size, if known. If only one count is printed, it is guaranteed to
313 be printed without leading spaces.
315 Previously, wc did not align the count fields if POSIXLY_CORRECT was set,
316 but POSIX did not actually require this undesirable behavior, so it
321 kill no longer tries to operate on argv[0] (introduced in 5.0.1)
322 Why wasn't this noticed? Although many tests use kill, none of
323 them made an effort to avoid using the shell's built-in kill.
325 `[' invoked with no arguments no longer evokes a segfault
327 rm without --recursive (aka -r or -R) no longer prompts regarding
328 unwritable directories, as required by POSIX.
330 uniq -c now uses a SPACE, not a TAB between the count and the
331 corresponding line, as required by POSIX.
333 expr now exits with status 2 if the expression is syntactically valid,
334 and with status 3 if an error occurred. POSIX requires this.
336 expr now reports trouble if string comparison fails due to a collation error.
338 split now generates suffixes properly on EBCDIC hosts.
340 split -a0 now works, as POSIX requires.
342 `sort --version' and `sort --help' fail, as they should
343 when their output is redirected to /dev/full.
345 `su --version > /dev/full' now fails, as it should.
347 ** Fewer arbitrary limitations
349 cut requires 97% less memory when very large field numbers or
350 byte offsets are specified.
353 * Major changes in release 5.0.1 (2003-07-15):
356 - new program: `[' (much like `test')
359 - head now accepts --lines=-N (--bytes=-N) to print all but the
360 N lines (bytes) at the end of the file
361 - md5sum --check now accepts the output of the BSD md5sum program, e.g.,
362 MD5 (f) = d41d8cd98f00b204e9800998ecf8427e
363 - date -d DATE can now parse a DATE string like May-23-2003
364 - chown: `.' is no longer recognized as a separator in the OWNER:GROUP
365 specifier on POSIX 1003.1-2001 systems. If chown *was not* compiled
366 on such a system, then it still accepts `.', by default. If chown
367 was compiled on a POSIX 1003.1-2001 system, then you may enable the
368 old behavior by setting _POSIX2_VERSION=199209 in your environment.
369 - chown no longer tries to preserve set-user-ID and set-group-ID bits;
370 on some systems, the chown syscall resets those bits, and previous
371 versions of the chown command would call chmod to restore the original,
372 pre-chown(2) settings, but that behavior is problematic.
373 1) There was a window whereby a malicious user, M, could subvert a
374 chown command run by some other user and operating on files in a
375 directory where M has write access.
376 2) Before (and even now, on systems with chown(2) that doesn't reset
377 those bits), an unwary admin. could use chown unwittingly to create e.g.,
378 a set-user-ID root copy of /bin/sh.
381 - chown --dereference no longer leaks a file descriptor per symlink processed
382 - `du /' once again prints the `/' on the last line
383 - split's --verbose option works once again [broken in 4.5.10 and 5.0]
384 - tail -f is no longer subject to a race condition that could make it
385 delay displaying the last part of a file that had stopped growing. That
386 bug could also make tail -f give an unwarranted `file truncated' warning.
387 - du no longer runs out of file descriptors unnecessarily
388 - df and `readlink --canonicalize' no longer corrupt the heap on
389 non-glibc, non-solaris systems
390 - `env -u UNSET_VARIABLE' no longer dumps core on non-glibc systems
391 - readlink's --canonicalize option now works on systems like Solaris that
392 lack the canonicalize_file_name function but do have resolvepath.
393 - mv now removes `a' in this example on all systems: touch a; ln a b; mv a b
394 This behavior is contrary to POSIX (which requires that the mv command do
395 nothing and exit successfully), but I suspect POSIX will change.
396 - date's %r format directive now honors locale settings
397 - date's `-' (no-pad) format flag now affects the space-padded-by-default
398 conversion specifiers, %e, %k, %l
399 - fmt now diagnoses invalid obsolescent width specifications like `-72x'
400 - fmt now exits nonzero when unable to open an input file
401 - tsort now fails when given an odd number of input tokens,
402 as required by POSIX. Before, it would act as if the final token
403 appeared one additional time.
405 ** Fewer arbitrary limitations
406 - tail's byte and line counts are no longer limited to OFF_T_MAX.
407 Now the limit is UINTMAX_MAX (usually 2^64).
408 - split can now handle --bytes=N and --lines=N with N=2^31 or more.
411 - `kill -t' now prints signal descriptions (rather than `?') on systems
412 like Tru64 with __sys_siglist but no strsignal function.
413 - stat.c now compiles on Ultrix systems
414 - sleep now works on AIX systems that lack support for clock_gettime
415 - rm now works around Darwin6.5's broken readdir function
416 Before `rm -rf DIR' would fail to remove all files in DIR
417 if there were more than 338.
419 * Major changes in release 5.0 (2003-04-02):
420 - false --help now exits nonzero
423 * printf no longer treats \x specially when POSIXLY_CORRECT is set
424 * printf avoids buffer overrun with format ending in a backslash and
425 * printf avoids buffer overrun with incomplete conversion specifier
426 * printf accepts multiple flags in a single conversion specifier
429 * seq no longer requires that a field width be specified
430 * seq no longer fails when given a field width of `0'
431 * seq now accepts ` ' and `'' as valid format flag characters
432 * df now shows a HOSTNAME: prefix for each remote-mounted file system on AIX 5.1
433 * portability tweaks for HP-UX, AIX 5.1, DJGPP
436 * printf no longer segfaults for a negative field width or precision
437 * shred now always enables --exact for non-regular files
438 * du no longer lists hard-linked files more than once
439 * du no longer dumps core on some systems due to `infinite' recursion
440 via nftw's use of the buggy replacement function in getcwd.c
441 * portability patches for a few vendor compilers and 64-bit systems
442 * du -S *really* now works like it did before the change in 4.5.5
445 * du no longer truncates file sizes or sums to fit in 32-bit size_t
446 * work around Linux kernel bug in getcwd (fixed in 2.4.21-pre4), so that pwd
447 now fails if the name of the working directory is so long that getcwd
448 truncates it. Before it would print the truncated name and exit successfully.
449 * `df /some/mount-point' no longer hangs on a GNU libc system when another
450 hard-mounted NFS file system (preceding /some/mount-point in /proc/mounts)
452 * rm -rf now gives an accurate diagnostic when failing to remove a file
453 under certain unusual conditions
454 * mv and `cp --preserve=links' now preserve multiple hard links even under
455 certain unusual conditions where they used to fail
458 * du -S once again works like it did before the change in 4.5.5
459 * stat accepts a new file format, %B, for the size of each block reported by %b
460 * du accepts new option: --apparent-size
461 * du --bytes (-b) works the same way it did in fileutils-3.16 and before
462 * du reports proper sizes for directories (not zero) (broken in 4.5.6 or 4.5.7)
463 * df now always displays under `Filesystem', the device file name
464 corresponding to the listed mount point. Before, for a block- or character-
465 special file command line argument, df would display that argument. E.g.,
466 `df /dev/hda' would list `/dev/hda' as the `Filesystem', rather than say
467 /dev/hda3 (the device on which `/' is mounted), as it does now.
468 * test now works properly when invoked from a set user ID or set group ID
469 context and when testing access to files subject to alternate protection
470 mechanisms. For example, without this change, a set-UID program that invoked
471 `test -w F' (to see if F is writable) could mistakenly report that it *was*
472 writable, even though F was on a read-only file system, or F had an ACL
473 prohibiting write access, or F was marked as immutable.
476 * du would fail with more than one DIR argument when any but the last did not
477 contain a slash (due to a bug in ftw.c)
480 * du no longer segfaults on Solaris systems (fixed heap-corrupting bug in ftw.c)
481 * du --exclude=FILE works once again (this was broken by the rewrite for 4.5.5)
482 * du no longer gets a failed assertion for certain hierarchy lay-outs
483 involving hard-linked directories
484 * `who -r' no longer segfaults when using non-C-locale messages
485 * df now displays a mount point (usually `/') for non-mounted
486 character-special and block files
489 * ls --dired produces correct byte offset for file names containing
490 nonprintable characters in a multibyte locale
491 * du has been rewritten to use a variant of GNU libc's ftw.c
492 * du now counts the space associated with a directory's directory entry,
493 even if it cannot list or chdir into that subdirectory.
494 * du -S now includes the st_size of each entry corresponding to a subdirectory
495 * rm on FreeBSD can once again remove directories from NFS-mounted file systems
496 * ls has a new option --dereference-command-line-symlink-to-dir, which
497 corresponds to the new default behavior when none of -d, -l -F, -H, -L
499 * ls dangling-symlink now prints `dangling-symlink'.
500 Before, it would fail with `no such file or directory'.
501 * ls -s symlink-to-non-dir and ls -i symlink-to-non-dir now print
502 attributes of `symlink', rather than attributes of their referents.
503 * Fix a bug introduced in 4.5.4 that made it so that ls --color would no
504 longer highlight the names of files with the execute bit set when not
505 specified on the command line.
506 * shred's --zero (-z) option no longer gobbles up any following argument.
507 Before, `shred --zero file' would produce `shred: missing file argument',
508 and worse, `shred --zero f1 f2 ...' would appear to work, but would leave
509 the first file untouched.
510 * readlink: new program
511 * cut: new feature: when used to select ranges of byte offsets (as opposed
512 to ranges of fields) and when --output-delimiter=STRING is specified,
513 output STRING between ranges of selected bytes.
514 * rm -r can no longer be tricked into mistakenly reporting a cycle.
515 * when rm detects a directory cycle, it no longer aborts the entire command,
516 but rather merely stops processing the affected command line argument.
519 * cp no longer fails to parse options like this: --preserve=mode,ownership
520 * `ls --color -F symlink-to-dir' works properly
521 * ls is much more efficient on directories with valid dirent.d_type.
522 * stty supports all baud rates defined in linux-2.4.19.
523 * `du symlink-to-dir/' would improperly remove the trailing slash
524 * `du ""' would evoke a bounds violation.
525 * In the unlikely event that running `du /' resulted in `stat ("/", ...)'
526 failing, du would give a diagnostic about `' (empty string) rather than `/'.
527 * printf: a hexadecimal escape sequence has at most two hex. digits, not three.
528 * The following features have been added to the --block-size option
529 and similar environment variables of df, du, and ls.
530 - A leading "'" generates numbers with thousands separators.
532 $ ls -l --block-size="'1" file
533 -rw-rw-r-- 1 eggert src 47,483,707 Sep 24 23:40 file
534 - A size suffix without a leading integer generates a suffix in the output.
536 $ ls -l --block-size="K"
537 -rw-rw-r-- 1 eggert src 46371K Sep 24 23:40 file
538 * ls's --block-size option now affects file sizes in all cases, not
539 just for --block-size=human-readable and --block-size=si. Fractional
540 sizes are now always rounded up, for consistency with df and du.
541 * df now displays the block size using powers of 1000 if the requested
542 block size seems to be a multiple of a power of 1000.
543 * nl no longer gets a segfault when run like this `yes|nl -s%n'
546 * du --dereference-args (-D) no longer fails in certain cases
547 * `ln --target-dir=DIR' no longer fails when given a single argument
550 * `rm -i dir' (without --recursive (-r)) no longer recurses into dir
551 * `tail -c N FILE' now works with files of size >= 4GB
552 * `mkdir -p' can now create very deep (e.g. 40,000-component) directories
553 * rmdir -p dir-with-trailing-slash/ no longer fails
554 * printf now honors the `--' command line delimiter
555 * od's 8-byte formats x8, o8, and u8 now work
556 * tail now accepts fractional seconds for its --sleep-interval=S (-s) option
559 * du and ls now report sizes of symbolic links (before they'd always report 0)
560 * uniq now obeys the LC_COLLATE locale, as per POSIX 1003.1-2001 TC1.
562 ========================================================================
563 Here are the NEWS entries made from fileutils-4.1 until the
564 point at which the packages merged to form the coreutils:
567 * `rm symlink-to-unwritable' doesn't prompt [introduced in 4.1.10]
569 * rm once again gives a reasonable diagnostic when failing to remove a file
570 owned by someone else in a sticky directory [introduced in 4.1.9]
571 * df now rounds all quantities up, as per POSIX.
572 * New ls time style: long-iso, which generates YYYY-MM-DD HH:MM.
573 * Any time style can be preceded by "posix-"; this causes "ls" to
574 use traditional timestamp format when in the POSIX locale.
575 * The default time style is now posix-long-iso instead of posix-iso.
576 Set TIME_STYLE="posix-iso" to revert to the behavior of 4.1.1 thru 4.1.9.
577 * `rm dangling-symlink' doesn't prompt [introduced in 4.1.9]
578 * stat: remove support for --secure/-s option and related %S and %C format specs
579 * stat: rename --link/-l to --dereference/-L.
580 The old options will continue to work for a while.
582 * rm can now remove very deep hierarchies, in spite of any limit on stack size
583 * new programs: link, unlink, and stat
584 * New ls option: --author (for the Hurd).
585 * `touch -c no-such-file' no longer fails, per POSIX
587 * mv no longer mistakenly creates links to preexisting destination files
590 * rm: close a hole that would allow a running rm process to be subverted
592 * New cp option: --copy-contents.
593 * cp -r is now equivalent to cp -R. Use cp -R -L --copy-contents to get the
594 traditional (and rarely desirable) cp -r behavior.
595 * ls now accepts --time-style=+FORMAT, where +FORMAT works like date's format
596 * The obsolete usage `touch [-acm] MMDDhhmm[YY] FILE...' is no longer
597 supported on systems conforming to POSIX 1003.1-2001. Use touch -t instead.
598 * cp and inter-partition mv no longer give a misleading diagnostic in some
601 * cp -r no longer preserves symlinks
602 * The block size notation is now compatible with SI and with IEC 60027-2.
603 For example, --block-size=1MB now means --block-size=1000000,
604 whereas --block-size=1MiB now means --block-size=1048576.
605 A missing `B' (e.g. `1M') has the same meaning as before.
606 A trailing `B' now means decimal, not binary; this is a silent change.
607 The nonstandard `D' suffix (e.g. `1MD') is now obsolescent.
608 * -H or --si now outputs the trailing 'B', for consistency with the above.
609 * Programs now output trailing 'K' (not 'k') to mean 1024, as per IEC 60027-2.
610 * New df, du short option -B is short for --block-size.
611 * You can omit an integer `1' before a block size suffix,
612 e.g. `df -BG' is equivalent to `df -B 1G' and to `df --block-size=1G'.
613 * The following options are now obsolescent, as their names are
614 incompatible with IEC 60027-2:
615 df, du: -m or --megabytes (use -BM or --block-size=1M)
616 df, du, ls: --kilobytes (use --block-size=1K)
618 * df --local no longer lists smbfs file systems whose name starts with //
619 * dd now detects the Linux/tape/lseek bug at run time and warns about it.
621 * ls -R once again outputs a blank line between per-directory groups of files.
622 This was broken by the cycle-detection change in 4.1.1.
623 * dd once again uses `lseek' on character devices like /dev/mem and /dev/kmem.
624 On systems with the linux kernel (at least up to 2.4.16), dd must still
625 resort to emulating `skip=N' behavior using reads on tape devices, because
626 lseek has no effect, yet appears to succeed. This may be a kernel bug.
628 * cp no longer fails when two or more source files are the same;
629 now it just gives a warning and doesn't copy the file the second time.
630 E.g., cp a a d/ produces this:
631 cp: warning: source file `a' specified more than once
632 * chmod would set the wrong bit when given symbolic mode strings like
633 these: g=o, o=g, o=u. E.g., `chmod a=,o=w,ug=o f' would give a mode
634 of --w-r---w- rather than --w--w--w-.
636 * mv (likewise for cp), now fails rather than silently clobbering one of
637 the source files in the following example:
638 rm -rf a b c; mkdir a b c; touch a/f b/f; mv a/f b/f c
639 * ls -R detects directory cycles, per POSIX. It warns and doesn't infloop.
640 * cp's -P option now means the same as --no-dereference, per POSIX.
641 Use --parents to get the old meaning.
642 * When copying with the -H and -L options, cp can preserve logical
643 links between source files with --preserve=links
644 * cp accepts new options:
645 --preserve[={mode,ownership,timestamps,links,all}]
646 --no-preserve={mode,ownership,timestamps,links,all}
647 * cp's -p and --preserve options remain unchanged and are equivalent
648 to `--preserve=mode,ownership,timestamps'
649 * mv and cp accept a new option: --reply={yes,no,query}; provides a consistent
650 mechanism to control whether one is prompted about certain existing
651 destination files. Note that cp's and mv's -f options don't have the
652 same meaning: cp's -f option no longer merely turns off `-i'.
653 * remove portability limitations (e.g., PATH_MAX on the Hurd, fixes for
655 * mv now prompts before overwriting an existing, unwritable destination file
656 when stdin is a tty, unless --force (-f) is specified, as per POSIX.
657 * mv: fix the bug whereby `mv -uf source dest' would delete source,
658 even though it's older than dest.
659 * chown's --from=CURRENT_OWNER:CURRENT_GROUP option now works
660 * cp now ensures that the set-user-ID and set-group-ID bits are cleared for
661 the destination file when when copying and not preserving permissions.
662 * `ln -f --backup k k' gives a clearer diagnostic
663 * ls no longer truncates user names or group names that are longer
665 * ls's new --dereference-command-line option causes it to dereference
666 symbolic links on the command-line only. It is the default unless
667 one of the -d, -F, or -l options are given.
668 * ls -H now means the same as ls --dereference-command-line, as per POSIX.
669 * ls -g now acts like ls -l, except it does not display owner, as per POSIX.
670 * ls -n now implies -l, as per POSIX.
671 * ls can now display dates and times in one of four time styles:
673 - The `full-iso' time style gives full ISO-style time stamps like
674 `2001-05-14 23:45:56.477817180 -0700'.
675 - The 'iso' time style gives ISO-style time stamps like '2001-05-14 '
677 - The 'locale' time style gives locale-dependent time stamps like
678 'touko 14 2001' and 'touko 14 23:45' (in a Finnish locale).
679 - The 'posix-iso' time style gives traditional POSIX-locale
680 time stamps like 'May 14 2001' and 'May 14 23:45' unless the user
681 specifies a non-POSIX locale, in which case it uses ISO-style dates.
684 You can specify a time style with an option like --time-style='iso'
685 or with an environment variable like TIME_STYLE='iso'. GNU Emacs 21
686 and later can parse ISO dates, but older Emacs versions cannot, so
687 if you are using an older version of Emacs outside the default POSIX
688 locale, you may need to set TIME_STYLE="locale".
690 * --full-time is now an alias for "-l --time-style=full-iso".
693 ========================================================================
694 Here are the NEWS entries made from sh-utils-2.0 until the
695 point at which the packages merged to form the coreutils:
698 * date no longer accepts e.g., September 31 in the MMDDhhmm syntax
699 * fix a bug in this package's .m4 files and in configure.ac
701 * nohup's behavior is changed as follows, to conform to POSIX 1003.1-2001:
702 - nohup no longer adjusts scheduling priority; use "nice" for that.
703 - nohup now redirects stderr to stdout, if stderr is not a terminal.
704 - nohup exit status is now 126 if command was found but not invoked,
705 127 if nohup failed or if command was not found.
707 * uname and uptime work better on *BSD systems
708 * pathchk now exits nonzero for a path with a directory component
709 that specifies a non-directory
712 * who accepts new options: --all (-a), --boot (-b), --dead (-d), --login,
713 --process (-p), --runlevel (-r), --short (-s), --time (-t), --users (-u).
714 The -u option now produces POSIX-specified results and is the same as
715 the long option `--users'. --idle is no longer the same as -u.
716 * The following changes apply on systems conforming to POSIX 1003.1-2001,
717 and are required by the new POSIX standard:
718 - `date -I' is no longer supported. Instead, use `date --iso-8601'.
719 - `nice -NUM' is no longer supported. Instead, use `nice -n NUM'.
720 * New 'uname' options -i or --hardware-platform, and -o or --operating-system.
721 'uname -a' now outputs -i and -o information at the end.
722 New uname option --kernel-version is an alias for -v.
723 Uname option --release has been renamed to --kernel-release,
724 and --sysname has been renamed to --kernel-name;
725 the old options will work for a while, but are no longer documented.
726 * 'expr' now uses the LC_COLLATE locale for string comparison, as per POSIX.
727 * 'expr' now requires '+' rather than 'quote' to quote tokens;
728 this removes an incompatibility with POSIX.
729 * date -d 'last friday' would print a date/time that was one hour off
730 (e.g., 23:00 on *thursday* rather than 00:00 of the preceding friday)
731 when run such that the current time and the target date/time fall on
732 opposite sides of a daylight savings time transition.
733 This problem arose only with relative date strings like `last monday'.
734 It was not a problem with strings that include absolute dates.
735 * factor is twice as fast, for large numbers
737 * setting the date now works properly, even when using -u
738 * `date -f - < /dev/null' no longer dumps core
739 * some DOS/Windows portability changes
741 * `date -d DATE' now parses certain relative DATEs correctly
743 * fixed a bug introduced in 2.0h that made many programs fail with a
744 `write error' when invoked with the --version option
746 * all programs fail when printing --help or --version output to a full device
747 * printf exits nonzero upon write failure
748 * yes now detects and terminates upon write failure
749 * date --rfc-822 now always emits day and month names from the `C' locale
750 * portability tweaks for Solaris8, Ultrix, and DOS
752 * date now handles two-digit years with leading zeros correctly.
753 * printf interprets unicode, \uNNNN \UNNNNNNNN, on systems with the
754 required support; from Bruno Haible.
755 * stty's rprnt attribute now works on HPUX 10.20
756 * seq's --equal-width option works more portably
758 * fix build problems with ut_name vs. ut_user
760 * stty: fix long-standing bug that caused test failures on at least HPUX
761 systems when COLUMNS was set to zero
762 * still more portability fixes
763 * unified lib/: now that directory and most of the configuration framework
764 is common between fileutils, textutils, and sh-utils
766 * fix portability problem with sleep vs lib/strtod.c's requirement for -lm
768 * fix portability problems with nanosleep.c and with the new code in sleep.c
770 * Regenerate lib/Makefile.in so that nanosleep.c is distributed.
772 * sleep accepts floating point arguments on command line
773 * sleep's clock continues counting down when sleep is suspended
774 * when a suspended sleep process is resumed, it continues sleeping if
775 there is any time remaining
776 * who once again prints whatever host information it has, even without --lookup
778 ========================================================================
779 For older NEWS entries for the fileutils, textutils, and sh-utils
780 packages, see ./old/*/NEWS.
782 This package began as the union of the following:
783 textutils-2.1, fileutils-4.1.11, sh-utils-2.0.15.