Path: utzoo!attcan!uunet!lll-winken!ames!ncar!unmvax!tut.cis.ohio-state.edu!ucbvax!NIEHS.BITNET!ALBRO
From: ALBRO@NIEHS.BITNET
Newsgroups: comp.sys.apple
Subject: Kermit
Message-ID: <8902040034.aa27702@SMOKE.BRL.MIL>
Date: 2 Feb 89 18:54:00 GMT
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The Internet
Lines: 26

This is a clarification of my previous clarification of.......
and hopefully the final clarification.

(1) Previously I referred to a line-feed as a CHR$(11).  It is, of
course, a CHR$(10).

(2) I think I have now done the definitive experiment.  I made a version
of Kermit-65 v.3.84 for the IIgs under DOS 3.3, and made a copy of it
with COPYII+ on a ProDOS disk with ProDOS-8 v.1.4, BASIC.SYSTEM v.1.1.
I then used the DOS 3.3 version to call our Vax cluster, on which I had
a text file known to contain cr,lf combinations (over 600 line feeds,
<ctrl>-J's).  I downloaded the file using Kermit-32 on the Vax, checking
that the default file type was set for ASCII at the Vax end, TEXT at the
Apple end.  The n I booted the ProDOS version of the identical Apple
Kermit and called the Vax again, downloading the same text file the same
way.

(3)  Upon checking the two downloads, I found that line feeds had been
removed in the DOS 3.3 copy, but retained in the ProDOS copy.  So
Kermit will automatically convert cr,lf combinations to cr under DOS
3.3, but not under ProDOS.  This must reflect the fact that a line feed
looks like $0A under ProDOS, but like $8A under DOS 3.3.  So this may
be why some people have trouble EXECing APPLE2-L downloads and others
don't.

ALBRO@NIEHS.BITNET