3 Tell <pinard@iro.umontreal.ca> if you feel like volunteering for any
4 of these ideas, listed more or less in decreasing order of priority.
7 - Update documentation from accumulated mail about it
8 - Changeword without arguments should restore default behavior
9 - Study synclines at the very beginning of each diverted sequence
10 - Make eval work on bignums - the 32 bits limit is artificial
11 From Krste Asanovic <krste@icsi.berkeley.edu>, 1993-03-20
13 * Optimization and clean up
14 - Check for memory leaks and uninitialized reads
15 From Vern Paxson <vern@horse.ee.lbl.gov> on 1993-12-06
16 - Simplify format/ecvt code, together with HAVE_EFGCVT
17 - Finalize the stdarg vs varargs thing
18 - Profile GNU m4 and speed it up
19 From David J. MacKenzie <djm@eng.umd.edu>, 1993-01-20
21 GNU m4 should be sped up by a factor of three for competing
22 with other versions (I think that the lexer is not agressive
23 enough and too often return single characters; obstacks might
24 be a little abused, too).
25 - Have NULs go really undisturbed through GNU m4
26 See `dumpdef' and debugging section, which abuses %s
27 From Thorsten Ohl <ohl@chico.harvard.edu>, 1992-12-21
29 path.c (add_include_directory): Why the '\0' terminator?
31 GNU m4 is lousy regarding NULs in streams (this would require
32 maintaining the string lengths, and avoiding strlen, strcpy,
34 - Clean up the obstack.[ch] code
35 - Use rx.[ch] instead of regex.[ch]
36 From Hal Peterson <hrp@ironwood.cray.com>, 1994-04-22
40 outline-regexp: " *[-+*.] \\|
\f"