EGA2MF is a utility to generate METAFONT code for
8x14 bitmaps of the type used for EGA screen fonts
on machines of the IBM pc/xt/at class.  The idea
is that you may then produce a font which emulates
the appearance of a video display screen with all
the crudities appropriate for that medium.
There are some trivial changes necessary to make
ega2mf work with some other bitmap type (e.g. 8x16
for the VGA or Hercules Plus).  This is left as an
exercise for the reader.
% 4/25/90: a slightly modified version, vga2mf.c, is
% now available wherein the necessary changes have been
% made for doing VGA 8x16 fonts (at the urging of Dimitri
% Vulis).  A really improved version would have the font-specific
% dimensions abstracted so that a command-line switch could
% determine the font type to be used.
% ``This is left as an exercise for the reader''

EGA2MF expects the name of an input bitmap file and an output
generated code file.  If these are not supplied it will ask.
EGA2MF does not do checking to verify that the input file truly
is a bitmapped font.  You may obtain spectacularly worthless
METAFONT code by allowing EGA2MF to process, for instance, this 
README file.

The METAFONT code generated will expect to have cmbase and the cmtt10
parameter file available.  

An early version of EGA2MF was used to produce the font WNPC10
which uses the ``standard'' character set as found in IBM pc-type
systems in the USA.  This is the same as the code page 437
character set below.  Your copy of EGA2MF __may__ be distributed
with the following code page bitmaps
CP437.EGA  United States
CP850.EGA  Multilingual
CP860.EGA  Portugal
CP863.EGA  Canada-French
CP865.EGA  Norway
CP880DV.EGA  Dimitri Vulus' Cyrillic ``code page 880''
% 4/25/90: there _may_ also be corresponding *.vga files
% suitable for input to vga2mf

For METAFONT tinkerers, the essential works are all at the head
of the file in s small number of magic-number variables which are
used by a macro named crt which draws the individual crt-style dots
for the characters.  I don't have any doubt that someone could
make improvements to the METAFONT code.  The C code can probably
also use some improvement; but given what it does, how much trouble
is it worth if it works at all?

[One such improvement of the algorithim is present in the jis2mf
program distributed as part of jemtex. See [tex.babel.japanese]
for more information. -dh]

EGA2MF works on my machines, under both DOS and UNIX, but is provided
to you in source so that you may verify that it will do nothing
catastrophic _before_ you compile it.  Since it is available to
you free there is, of course, no warranty of any sort whatsoever,
and you use the program entirely at your own risk.

% 91/05/24 Brian {Hamilton Kelly}, checking that the programs would work
% under VAX/VMS, replaced the line reading
  FILE *f1,*f2;
% in each source with these lines
  FILE *f1=(FILE *) NULL;
  FILE *f2=(FILE *) NULL;
% (In C, there's no guarantee that automatic variables will have _any_
% particular value, let alone the (FILE *)NULL that was necessary here
% with the method the program uses to prompt for missing command-line
% arguments.  Perhaps he was just lucky with his Unix & DOS :-)


- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Thomas Ridgeway, Director,
Humanities and Arts Computing Center/NorthWest Computing Support Center
35 Thomson Hall                            H    H    AA    CCCCC  CCCCC
University of Washington, DR-10            H    H   A  A   C      C    
Seattle, WA 98195                          HHHHHH  AAAAAA  C      C    
(206)-543-4218                             H    H  A    A  C      C    
ridgeway@blackbox.hacc.washington.edu      H    H  A    A  CCCCC  CCCCC
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-     Your comments, suggestions and improvements will be welcome    -
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -