XMODEM file padding (was Re: UMODEM on local connections)

DoN Nichols dnichols at ceilidh.beartrack.com
Sun Apr 14 15:34:03 AEST 1991


In article <1991Apr13.014747.7144 at netcom.COM> gandrews at netcom.COM (Greg Andrews) writes:
>In article <1991Apr9.161637.12059 at oswego.oswego.edu> ostroff at oswego.Oswego.EDU (Boyd Ostroff) writes:
>>
>>The only problem I've noticed is that there are often a bunch of NULLs 

	[ ... ]

>If the file size doesn't match up to an exact multiple of 128 bytes, the
>last packet will be only partly full of file data.  The protocol has to
>put SOMEthing into the packet to fill it out, and nulls are often used.

	Also VERY frequently used is the '^Z' character, since it is
interpreted as 'EOT' by text-mangling utilities under CP/M (where the xmodem
protocol originated), and, by extension, MS-DOS.  The packet size was chosen
to be a single sector on then-popular implementations of CP/M.  (I don't
know whether CP/M ever had any other size sectors, alghough probably CP/M-86
which ran on IBM-PC hardware probably used larger sectors, at least as an
option.  At least there are never more than 127 of them, so I can get rid of
them using JOVE, and not have to resort to emacs. :-)

	Kermit is a MUCH nicer protocol, it's a pity that it has so much
overhead :-) (Although two of the newer ones talking to each other are not
bad at all.

-- 
Donald Nichols (DoN.)		| Voice (Days):	(703) 664-1585
D&D Data			| Voice (Eves):	(703) 938-4564
Disclaimer: from here - None	| Email:     <dnichols at ceilidh.beartrack.com>
	--- Black Holes are where God is dividing by zero ---



More information about the Comp.sys.3b1 mailing list