Release 1.1.3: 7/23/2003:

Fixed several problems thanks to excellent testing and fixes from
Andrey Mirtchovski (mirtchov@cpsc.ucalgary.ca).  One nasty bug would
rio to abort when using the generated font.  The man page for
cachechar(2) clearly had stated a requirement that didn't make it into
the code.

Another change is to the flags (again!).  The '-r' flag is dropped in
favor of a '-f' flag that generates a '.font' format file suitable for
use with the files that are generated when bdf2subf is invoked with no
flags.  Consequently, the utility subfsize is no longer needed.

Release 1.1.2: 7/14/2003:

Hopefully this will be it for a while.

Based on feedback on the performance of bdf2subf for very large BDF
files, and how this is used in real life, I took several of Andrey
Mirtchovski's (mirtchov@cpsc.ucalgary.ca) suggestions and code fixes
and incorporated them here.  I've added a work-around for the
writememimage slowness (look for the word HACK in the code).  Also
revamped the flags to make it more natural.  Andrey's suggestion to
generate all subfonts in one pass makes sense for very large font
files and I've added it.  I would guess that most of the time
it wouldn't be used as heavily.  PARAMETERS TO THE PROGRAM HAVE
CHANGED. CHECK THE NEW MAN PAGE. Also there is a new utility,
subfsize, contributed by Andrey to ease generation of .font files.

Release 1.1.1: 7/13/2003:

You know what they say about "premature optimization", well, it's true.
In order to reduce the number of calls to memimagedraw, a critical one
was removed that didn't cause a problem when extracting all glyphs in
a file, but when extracting a subrange, the size of the subfont bitmap
was as large as if all the glyphs in the file were being extracted.
This has been fixed.

Release 1.1: 6/15/2003:

Some efficiency fixes have been done.  Beware that handling very large
bdf's (like efonts) in a single step will be slow and the result will
most likely be too large to be useful.  In those cases run bdf2subf
once to discover the contiguous subranges and then run it multiple
times to get the fonts for each selected subrange.

Release 1.0: 9/26/2002:

Bdf2subf has been tested with as many BDF files as I could find on
the Net.  It is pedantic about the syntax of the file.  I've put in
catches for other bogus things like zero length glyphs (what's their
use?  I really want to know.) and other surprises.  A correctly formed
file is handled correctly.  It tolerates empty lines at the end of BDF files
and probably should handle them everywhere, but it doesn't.

to build run:
	mk

see the man page for usage.

Fariborz "Skip" Tavakkolian
fst@centurytel.net
