Date   : Thu, 21 Dec 1989 13:05:16 EST
From   : "John S. Fisher" <FISHER%RPIECS.BITNET@CUNYVM.CUNY.EDU>
Subject: UUencoding by the RPIECS server

In response to this long debate about the various flavors of uuencode, I've
been researching the whole problem as best I can.  Not being a "Unix-person"
nor in close touch with the Unix community, this has taken longer than it
might otherwise had.  At any rate, the result is the following announcement.

Effective immediately:

(1)  The server at RPIECS uses the grave-accent substitute for blank in
     its uuencoder.  Since this eliminates the trailing blank truncation
     problem, the trailing M character is removed.  I believe that the
     uuencoding variant now in use conforms to the officially recognized
     standard method.

(2)  The option OLDUUE is available on the /PDGET command to explicitly
     request the trailing-M form for those that really need it.  The
     speculation is that no-one will, so this option may be removed in
     the future.

(3)  As an alternate to uuencoding, the so-called xxencoding method can be
     requested by specifying XXENCODE as an option on the /PDGET command.

(4)  The PDGET HELP file will be updated to reflect the changes.

The user-visible change here is that uuencoded files, implicitly or explicitly
requested, are now being sent in the newer format.  (Please note that
checksumming is *not* included.  I have been informed that this is not (yet?)
an official standard variant.)  I hope this is good news for some and no news
for most of you.

On the side of bad news:  In the process of making this change today, three
user requests for uuencoded data slipped in during testing.  They were
incorrectly sent out as xxencoded files.  My apologies to the users affected by
this error.

I am now in the process of packaging the server changes to forward to NDSUVM1.
I suspect the people there will be making the same changes in short order, but
the actual schedule is for them to determine.


/JSFisher