btoa/atob author(s) -- where are you?

[david at ukma.UUCP]

I'm looking for the authors of the btoa/atob pair of programs
which came with the compress v.4 distribution (The mod.sources
distribution, not the 4.3BSD distribution), and not the first
version of btoa/atob which was slightly different from the
compress v.4 version.

A group of us have been using compress and btoa/atob for news
transmissions across BITNET and have discovered that there are
some subtle bugs in the code.  Sometimes under unknown conditions,
btoa creates a file which atob cannot decode because it thinks
there is a bad checksum involved.  

I'd found one bug in the program involving an ending condition
which ... at the very end of the file under certain conditions
the program (as posted) would make a blank line between the
end of the data and the checksum trailer.  This blank line
(when passed through BITNET) became non-empty and the code
in atob which passed over the blank line failed because
it was no longer empty (i.e. it went from "\n" to "     \n").

I've fixed that bug but there's another one in the program.

I'd like to communicate with the author(s) and see what is known.
The error rate is low enough that we get two to three failures
per week's worth of news (er.. I have no idea how many batches
we have go through per week, bit it might be as many as a thousand(?)).

The header of the program mentions the person who'd been running
the mail.compress list ... I've sent him a bug report on the
one bug I found but that elicited no response.

			advTHANKSance.
-- 
-----   David Herron,  cbosgd!ukma!david, david at UKMA.BITNET, david at ms.uky.csnet
-----           (also "postmaster", "news", "netnews", "uucp", "mmdf", and ...)
-----                             (and the Usenet map maintainer for Kentucky.)
-----            "A ... A carton of Camels by noon, or permanent penguin pate!"