1 If you're interested in helping, here are some tasks that we've considered
2 over the years. Beware: some are quite old and no longer valid. To avoid
3 wasting your time by duplicating work or by working on a task that is no
4 longer pertinent, please search the mailing list and post your intent
5 before embarking on a big project.
7 ==================================================
8 document the following in coreutils.texi:
15 Also document the SELinux changes.
17 comm: add an option, --output-delimiter=STR
18 Files to change: src/comm.c, ChangeLog, NEWS, doc/coreutils.texi,
19 Add a new file, tests/misc/comm (use another file in that directory as
20 a template), to exercise the new option. Suggestion from Dan Jacobson.
23 Now that gnulib supports *printf("%a"), import one of the
24 *printf-posix modules so that printf(1) will support %a even on
25 platforms where the native *printf(3) is deficient.
26 Suggestion from Eric Blake.
28 renice: POSIX utility, needs implementing.
29 suggestion from Karl Berry (among others).
30 Bob Proulx is working on this.
32 install: add an option to specify the program used to strip binaries.
33 suggestion from Karl Berry
36 Address this comment: FIXME: mv's behavior in this case is system-dependent
37 Better still: fix the code so it's *not* system-dependent.
39 ls: add --format=FORMAT option that controls how each line is printed.
41 cp --no-preserve=X should not attempt to preserve attribute X
42 reported by Andreas Schwab
44 copy.c: Address the FIXME-maybe comment in copy_internal.
45 And once that's done, add an exclusion so that `cp --link'
46 no longer incurs the overhead of saving src. dev/ino and dest. filename
49 Write an autoconf test to work around build failure in HPUX's 64-bit mode.
50 See notes in README -- and remove them once there's a work-around.
52 Integrate use of sendfile, suggested here:
53 http://mail.gnu.org/archive/html/bug-fileutils/2003-03/msg00030.html
54 I don't plan to do that, since a few tests demonstrate no significant benefit.
56 Should printf '\0123' print "\n3"?
57 per report from TAKAI Kousuke on Mar 27
58 http://mail.gnu.org/archive/html/bug-coreutils/2003-03/index.html
60 printf: consider adapting builtins/printf.def from bash
62 df: add `--total' option, suggested here http://bugs.debian.org/186007
64 seq: give better diagnostics for invalid formats:
65 e.g. no or too many % directives
66 seq: consider allowing format string to contain no %-directives
68 tail: don't use xlseek; it *exits*.
69 Instead, maybe use a macro and return nonzero.
71 tr: support nontrivial equivalence classes, e.g. [=e=] with LC_COLLATE=fr_FR
73 lib/strftime.c: Since %N is the only format that we need but that
74 glibc's strftime doesn't support, consider using a wrapper that
75 would expand /%(-_)?\d*N/ to the desired string and then pass the
76 resulting string to glibc's strftime.
78 unexpand: [http://www.opengroup.org/onlinepubs/007908799/xcu/unexpand.html]
79 printf 'x\t \t y\n'|unexpand -t 8,9 should print its input, unmodified.
80 printf 'x\t \t y\n'|unexpand -t 5,8 should print "x\ty\n"
82 Let GNU su use the `wheel' group if appropriate.
83 (there are a couple patches, already)
85 sort: Investigate better sorting algorithms; see Knuth vol. 3.
87 We tried list merge sort, but it was about 50% slower than the
88 recursive algorithm currently used by sortlines, and it used more
89 comparisons. We're not sure why this was, as the theory suggests it
90 should do fewer comparisons, so perhaps this should be revisited.
91 List merge sort was implemented in the style of Knuth algorithm
92 5.2.4L, with the optimization suggested by exercise 5.2.4-22. The
93 test case was 140,213,394 bytes, 426,4424 lines, text taken from the
94 GCC 3.3 distribution, sort.c compiled with GCC 2.95.4 and running on
95 Debian 3.0r1 GNU/Linux, 2.4GHz Pentium 4, single pass with no
96 temporary files and plenty of RAM.
98 Since comparisons seem to be the bottleneck, perhaps the best
99 algorithm to try next should be merge insertion. See Knuth section
100 5.3.1, who credits Lester Ford, Jr. and Selmer Johnson, American
101 Mathematical Monthly 66 (1959), 387-389.
103 cp --recursive: perform dir traversals in source and dest hierarchy rather
104 than forming full file names. The latter (current) approach fails
105 unnecessarily when the names become very long.
107 Remove suspicious uses of alloca (ones that may allocate more than
110 Adapt these contribution guidelines for coreutils:
111 http://sources.redhat.com/automake/contribute.html
114 Changes expected to go in, someday.
115 ======================================
117 dd patch from Olivier Delhomme
119 test/mv/*: clean up $other_partition_tmpdir in all cases
121 ls: when both -l and --dereference-command-line-symlink-to-dir are
122 specified, consider whether to let the latter select whether to
123 dereference command line symlinks to directories. Since -l has
124 an implicit --NO-dereference-command-line-symlink-to-dir meaning.
125 Pointed out by Karl Berry.
127 A more efficient version of factor, and possibly one that
128 accepts inputs of size 2^64 and larger.
130 dd: consider adding an option to suppress `bytes/block read/written'
131 output to stderr. Suggested here:
132 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=165045
134 Pending copyright papers:
135 ------------------------
136 getpwnam from Bruce Korb
138 pb (progress bar) from Miika Pekkarinen
140 ------------------------------
142 Remove long-deprecated options. Search case-insensitive for
143 `deprecated' and `remove in '. Automate this.
145 Add a distcheck-time test to ensure that every distributed
146 file is either read-only(indicating generated) or is
147 version-controlled and up to date.
149 Implement Ulrich Drepper's suggestion to use getgrouplist rather than
150 getugroups. This affects both `id' and `setuidgid', but makes a big
151 difference on systems with many users and/or groups, and makes id usable
152 once again on systems where access restrictions make getugroups fail.
153 But first we'll need a run-test (either in an autoconf macro or at
154 run time) to avoid the segfault bug in libc-2.3.2's getgrouplist.
155 In that case, we'd revert to using a new (to-be-written) getgrouplist
156 module that does most of what `id' already does. Or just avoid the
157 buggy use of getgrouplist by never passing it a buffer of length zero.
158 See http://bugzilla.redhat.com/200327
160 remove `%s' notation (now that they're all gone, add a Makefile.maint sc_
161 rule to ensure no new ones are added):
162 grep -E "\`%.{,4}s'" src/*.c
164 remove all uses of the `register' keyword: Done. add a Makefile.maint rule
167 remove or adjust chown's --changes option, since it
168 can't always do what it currently says it does.
170 Adapt tools like wc, tr, fmt, etc. (most of the textutils) to be
171 multibyte aware. The problem is that I want to avoid duplicating
172 significant blocks of logic, yet I also want to incur only minimal
173 (preferably `no') cost when operating in single-byte mode.
175 pr's use of nstrftime can make it malloc a very large (up to SIZE_MAX) buffer
179 Copyright (C) 2002-2008 Free Software Foundation, Inc.
181 This program is free software: you can redistribute it and/or modify
182 it under the terms of the GNU General Public License as published by
183 the Free Software Foundation, either version 3 of the License, or
184 (at your option) any later version.
186 This program is distributed in the hope that it will be useful,
187 but WITHOUT ANY WARRANTY; without even the implied warranty of
188 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
189 GNU General Public License for more details.
191 You should have received a copy of the GNU General Public License
192 along with this program. If not, see <http://www.gnu.org/licenses/>.