diffmk can't handle large files

[lance.larsen]

The standard UNIX tool diffmk cannot handle large files.
When attempting to diffmk two versions of a large file,
the output file was smaller than either of the source files
and the beginning of the output file contained a page full of the line
???? 109 Data Read Error ????  repeated before the first lines
in the source file(s). The text inside the question marks
corresponded to the last line in the truncated output file.

Diffmk uses diff (with the -e option), and ed which has a ~100K
limit on file size.

We've tried substituting ex for ed in the diffmk shell script,
and it doesn't bomb on large files, but it prepends an extra
character in front of lines 2 through n of the output file.
I think that it is garbling the CRLF escape sequence.

Is there an alternate BIG diffmk that can handle files
larger than 100K?  Or is there a way to modify diffmk
to handle larger files?  


Lance F. Larsen
mtunq!lfl


P.S.	I hope that this is not too trivial a question for this forum.
	I did RTFM, and I tried several system administrators, but
	no luck.