summaryrefslogtreecommitdiff
path: root/contrib/gnu-sort/TODO
diff options
context:
space:
mode:
Diffstat (limited to 'contrib/gnu-sort/TODO')
-rw-r--r--contrib/gnu-sort/TODO93
1 files changed, 93 insertions, 0 deletions
diff --git a/contrib/gnu-sort/TODO b/contrib/gnu-sort/TODO
new file mode 100644
index 000000000000..a102576fb508
--- /dev/null
+++ b/contrib/gnu-sort/TODO
@@ -0,0 +1,93 @@
+Tasks for GNU textutils (listed in no particular order):
+
+ write texinfo documentation for sha1sum
+
+ Something that I would really appreciate is if someone would run the
+ Open Group's VSC-lite test suite against the fileutils and textutils
+ and report the failures.
+
+ http://www.opengroup.org/testing/downloads/vsclite.html
+
+ I've been meaning to do it myself for months, but haven't found the time.
+ There's a bit of set-up required, some of which requires root access, e.g.,
+ to create a few test user accounts and some test groups.
+ ------------------
+
+ uniq: remove support for obsolescent +N syntax
+
+ add tests for od
+ add some endian-aware tests for od
+
+ tac: Set DONT_UNLINK_WHILE_OPEN when necessary.
+
+ tail: add an option so that using -f on N files doesn't monopolize
+ N file descriptors
+
+ tac: add options to help handle boundary cases
+ E.g., options to distinguish DELIM_STRING is
+ - starter (see existing --before option)
+ - terminator (this is what most people expect wrt NEWLINE
+ - separator (this would make `echo -n a:b:c|tac -s:' print `c:b:a')
+
+ tail: support -r option by librarifying tac and using that
+
+ cut: maybe add an option to say `fields are separated by whitespace'.
+ Of course, that isn't really necessary because you can preprocess
+ cut's input with tr to get the same effect:
+
+ echo 'a b c' |tr -s '[:blank:]' | cut -d ' ' -f 2
+
+------------
+
+ From: kwzh@gnu.ai.mit.edu (Karl Heuer)
+ Subject: [textutils-1.22] [sort] feature requests
+ To: textutils-bugs@gnu.ai.mit.edu
+ Date: Thu, 5 Jun 97 13:06:51 -0400
+
+ [...]
+ Another feature that I would sometimes find useful: change -c so that
+ it will report up to N instances of disorder before bailing out, where
+ N defaults to 1 but can be set to infinity or to some finite value by
+ another option. (An "instance of disorder" is two adjacent lines that
+ are malsorted; this does not imply that swapping them or removing one
+ or both would cause the list to be sorted. (1 3 5 7 9 0 2 4 6 8) has
+ just one instance of disorder.)
+
+------------
+
+ Date: Fri, 1 May 1998 20:27:39 -0700 (PDT)
+ From: Paul Rubin <phr@netcom.com>
+ To: gnu@gnu.org
+ Subject: small project suggestion
+
+ Someone should rewrite the "sum" utility to give a choice of
+ different checksum algorithms (it's poorly organized for that now).
+ An experienced programmer could probably do it in a day or so,
+ or it might be a good, self-contained project for someone who is
+ just getting started.
+
+ Algorithms that it should include are:
+ -- the POSIX algorithm
+ -- the BSD algorithm
+ -- CRC32 algorithm (used by pkzip)
+ -- CRC16 (used in TCP/IP)
+ -- possibly other CRC's (like the different CCITT polynomials)
+ -- SHA-1 and MD5 cryptographic hashes (replacing "md5sum").
+ and possibly:
+ -- DSA digital signature based on secret key generated from
+ a passphrase (prompt the user, or read an environment variable).
+
+
+---------------------
+
+comm: add an option-enable check for sortedness of input files
+
+---------------------
+
+uniq: add a more flexible key selection mechanism
+
+---------------------
+
+Charles Randall <crandall@matchlogic.com>
+is working on making sort more suitable and efficient for very
+large sets of input data.