1 restore djgpp, eventually
3 add unit tests for lib/*.c
4 rewrite lib/ftw.c not to use explicit recursion, and then use nftw in
5 chown, chgrp, chmod, du
7 strip: add an option to specify the program used to strip binaries.
8 suggestion from Karl Berry
11 Address this comment: FIXME: mv's behavior in this case is system-dependent
12 Better still: fix the code so it's *not* system-dependent.
14 implement --target-directory=DIR for install (per texinfo documentation)
16 ls: add --format=FORMAT option that controls how each line is printed.
18 cp --no-preserve=X should not attempt to preserve attribute X
19 reported by Andreas Schwab
21 copy.c: Address the FIXME-maybe comment in copy_internal.
22 And once that's done, add an exclusion so that `cp --link'
23 no longer incurs the overhead of saving src. dev/ino and dest. filename
26 See if we can be consistent about where --verbose sends its output:
27 These all send --verbose output to stdout:
28 head, tail, rm, cp, mv, ln, chmod, chown, chgrp, install, ln
29 These send it to stderr:
33 Write an autoconf test to work around build failure in HPUX's 64-bit mode.
34 See notes in README -- and remove them once there's a work-around.
36 Integrate use of sendfile, suggested here:
37 http://mail.gnu.org/archive/html/bug-fileutils/2003-03/msg00030.html
38 I don't plan to do that, since a few tests demonstrate no significant benefit.
40 Should printf '\0123' print "\n3"?
41 per report from TAKAI Kousuke on Mar 27
42 http://mail.gnu.org/archive/html/bug-coreutils/2003-03/index.html
44 printf: consider adapting builtins/printf.def from bash
46 df: add `--total' option, suggested here http://bugs.debian.org/186007
48 seq: give better diagnostics for invalid formats:
49 e.g. no or too many % directives
50 seq: consider allowing format string to contain no %-directives
52 dd: consider adding an option to suppress `bytes/block read/written'
53 output to stderr. Suggested here:
54 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=165045
56 m4: rename all macros that start with AC_ to start with another prefix
58 resolve RH report on cp -a forwarded by Tim Waugh
60 Martin Michlmayr's patch to provide ls with `--sort directory' option
62 tail: don't use xlseek; it *exits*.
63 Instead, maybe use a macro and return nonzero.
65 add mktemp? Suggested by Nelson Beebe
67 Now that AC_FUNC_LSTAT and AC_FUNC_STAT are in autoconf,
68 remove m4/stat.m4 and m4/lstat.m4.
70 df: alignment problem of `Used' heading with e.g., -mP
71 reported by Karl Berry
73 tr: support nontrivial equivalence classes, e.g. [=e=] with LC_COLLATE=fr_FR
75 fix tail -f to work with named pipes; reported by Ian D. Allen
76 $ mkfifo j; tail -f j & sleep 1; echo x > j
77 ./tail: j: file truncated
78 ./tail: j: cannot seek to offset 0: Illegal seek
80 lib/strftime.c: Since %N is the only format that we need but that
81 glibc's strftime doesn't support, consider using a wrapper that
82 would expand /%(-_)?\d*N/ to the desired string and then pass the
83 resulting string to glibc's strftime.
85 sort: Compress temporary files when doing large external sort/merges.
86 This improves performance when you can compress/uncompress faster than
87 you can read/write, which is common in these days of fast CPUs.
88 suggestion from Charles Randall on 2001-08-10
90 sort: Add an ordering option -R that causes 'sort' to sort according
91 to a random permutation of the correct sort order. Also, add an
92 option --random-seed=SEED that causes 'sort' to use an arbitrary
93 string SEED to select which permutations to use, in a deterministic
94 manner: that is, if you sort a permutation of the same input file
95 with the same --random-seed=SEED option twice, you'll get the same
96 output. The default SEED is chosen at random, and contains enough
97 information to ensure that the output permutation is random.
98 suggestion from Feth AREZKI, Stephan Kasal, and Paul Eggert on 2003-07-17
100 unexpand: [http://www.opengroup.org/onlinepubs/007908799/xcu/unexpand.html]
101 printf 'x\t \t y\n'|unexpand -t 8,9 should print its input, unmodified.
102 printf 'x\t \t y\n'|unexpand -t 5,8 should print "x\ty\n"
104 Let GNU su use the `wheel' group if appropriate.
105 (there are a couple patches, already)
107 sort: Investigate better sorting algorithms; see Knuth vol. 3.
109 We tried list merge sort, but it was about 50% slower than the
110 recursive algorithm currently used by sortlines, and it used more
111 comparisons. We're not sure why this was, as the theory suggests it
112 should do fewer comparisons, so perhaps this should be revisited.
113 List merge sort was implemented in the style of Knuth algorithm
114 5.2.4L, with the optimization suggested by exercise 5.2.4-22. The
115 test case was 140,213,394 bytes, 426,4424 lines, text taken from the
116 GCC 3.3 distribution, sort.c compiled with GCC 2.95.4 and running on
117 Debian 3.0r1 GNU/Linux, 2.4GHz Pentium 4, single pass with no
118 temporary files and plenty of RAM.
120 Since comparisons seem to be the bottleneck, perhaps the best
121 algorithm to try next should be merge insertion. See Knuth section
122 5.3.1, who credits Lester Ford, Jr. and Selmer Johnson, American
123 Mathematical Monthly 66 (1959), 387-389.
126 Document the exit status of each and every program.
127 Suggestion from Dan Jacobson.
128 The sentence or two describing the common case must appear just once,
129 and then it must be made to appear in the 70+ places where it's needed.
131 du: rewrite to use something other than nftw to perform directory traversal.
132 Due to design limitations of nftw, ...
133 using -D makes du dereference each command
134 line argument that is a symlink to a directory.
135 A symlink to non-directory is not dereferenced.