Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!spdcc!m2c!necntc!drilex!dricej
From: dricej@drilex.UUCP (Craig Jackson)
Newsgroups: comp.sources.bugs,news.misc
Subject: Re: sending source code
Message-ID: <303@drilex.UUCP>
Date: Fri, 13-Nov-87 09:37:34 EST
Article-I.D.: drilex.303
Posted: Fri Nov 13 09:37:34 1987
Date-Received: Sun, 15-Nov-87 09:08:04 EST
References: <631@louie.udel.EDU> <332@uvicctr.UUCP> <2566@umn-cs.UUCP> <467@srs.UUCP> <49@48color.UUCP> <939@hao.UCAR.EDU>
Reply-To: dricej@drilex.UUCP (Craig Jackson)
Organization: Data Resources/McGraw-Hill, Lexington, MA
Lines: 23
Summary: Double compression means expansion
Xref: mnetor comp.sources.bugs:456 news.misc:1129

Several articles in this chain have addressed the problems of sending
the output of ARC, etc through Usenet.  Most of them have indicated that the
apparent problem is that the output of these programs is binary, and therefore
the overhead of uuencode or btoa must be incurred.  Some have suggested that
these programs output printable ASCII directly.

Unfortunately, its not the unprintability that matters for these forms, its the
*compression*.  Most Usenet links (except for dedicated links such as the
Arpanet) already use state-of-the-art compression technology to transmit
news.  There's only so much compression that can be achieved on any piece
of information, especially if everybody uses the same algorithm.  In fact,
double compression generally results in expansion.

So the moral is: if you have to send binary, ARC it all you want, but TURN OFF
THE COMPRESSION.

However, I will second Greg's plea: If you have large sources to send, the
regular mails probably have a higher baud-rate.  (In addition, you don't 
impose on other system's hospitality.)
-- 
Craig Jackson
UUCP: {harvard!axiom,linus!axiom,ll-xn}!drilex!dricej
BIX:  cjackson