> This is simply  the way the program LISTSERV uses  for UUENCODE works (it
> has its own reasons which were explained  to me once and which I promptly
> forgot  - it  was  designed for  PC  software).
 
I guess that makes a certain kind of sense.  The encoded output stopped
after exactly 128 bytes, but I don't understand how an encoder that
won't handle more than 128 bytes can be said to "work."  Could it be
that the program on CMS reads no more that 128 bytes from each *record*?
 
>  In the  particular case  of transient  MODULEs, there  is a
> rather legitimate case for wanting to  use UUENCODE to retrieve the file,
 
I agree, and, since SPRSHO is such a module (more specifically, a system-
key, transient-area module), I would have thought that the UUENCODER was
ideal for the task.  If, however, it can't read more than 128 bytes from
any one record, then I guess it isn't.
 
> but what about regular MODULE's?
 
Well, maybe not, although there is another special case that could
perhaps use it, namely, modules with no symbol table or RLD data.
Those could be reconstructed, although it would be a nuisance to
have to clip off the first 80 bytes.
                                     John