	       Development log for biblabel and citesub
		   [in reverse chronological order]

[16-Oct-1996]	0.02
		In biblabel.awk, add extra function arguments to
		represent local variables (thanks to the diagnosis of my
		modified awk).  Change all truly global variables to be
		named with an initial capital letter, so that
		locals/globals can readily be distinguished.

		Recognize booktitle, and use it for title abbreviation
		if no title is supplied.

		Comment out all print_debug() calls with a double sharp,
		so that then can easily be restored in the future if
		needed, and until then, cause no extra overhead.

		Change make_unique_new_tag() to receive an argument, and
		return a modified copy of its argument.

		Change test in is_tag_in_use() to not check for file
		difference, thereby catching instances of duplicate
		labels in the current file, and avoiding generation of
		excessively long suffix strings for such duplicates.

		Extend a couple of warning messages with extra
		commentary.


[08-Nov-1994]	0.01
		In biblabel.awk, change fix_author_editor() to
		eliminate generational numbers (II, III, IV, etc.)
		when there is no preceding comma.  Previously, the
		code would reduce "Hunt, III" to "Hunt", but leave
		"Hunt III" as "Hunt III", producing a lastname of III
		instead of Hunt in the citation label.  Use more
		sophisticated code to handle difficult cases like:
			"John Smith Sr and A. Srinivasan"
			"A. Srinivasan and John Smith Sr"
			"A. Srinivasan Sr and A. Srinivasan Jr"
			"Shu-T'ien Xi"

		In biblabel.awk, change title_string() to reduce
		slash (/), \slash, and \emdash to space so that
		Input/Output, Input\slash Output, and
		Input\emdash Output all reduce to IO.

		Add code to citesub.c to check for a substitution file
		on an unseekable file, such as piped stdin, and if so,
		copy the input to a temporary file.  In order to set
		the hash table size to a reasonable value, citesub
		needs to know how many entries there are; it gets that
		number by counting line in the substitution file.


[30-Oct-1994]	0.00
		First working versions of citesub in C and awk, based
		on experience with two earlier unnumbered versions
		that were successfully used for more than a dozen
		large bibliographies, but suffered from performance
		problems.  Total working time for programming and
		documentation: 12 hours.
